One night, just before going to bed, I saw a chat message pop up on my screen: Hey, is your website down? I tried to send someone to it but looks like it’s not there.” I let out a big groan and shot off a chat message to the person who maintains my website. I didn’t have high hopes of reaching her, as her social media posts are from a European beach where she’s meeting up with a long-lost friend this week. I went to bed and slept poorly, knowing that I would have an extra tech troubleshooting task in the morning. But the day is my day off so I braced myself to climb that self-serve tech mountain.

Come the morning, I had a missed call from my web person, and a message asking if I was around. She’d tried to log into my hosting service to fix the problem but the password failed and she’s in the middle of a family day with her husband, two children, and various other factors that will make it hard to get to this situation. Before I even have coffee, I am on my laptop and trying to figure out where to change the “A Records” to the destination she’s given me. I find various places where I can change domain destinations but they don’t look quite right. I start googling “is an A Record the same as a Nameserver” and multiple variations but the results are jargon-filled explanations that don’t help me.

Three interactions, three customer experience fails

On the website of Tsohost, my ISP, I try to open a ticket but got a message encouraging me to use their chat instead. But their chat doesn’t open until 9 AM (no time zone mentioned) so that transaction goes into the Pending list. (And I dread the context switching, where there will be a window of time spent remembering what the problem is, why I’m getting in touch, and how to explain it to them.) Eventually I’ll get there, but after a multi-detour journey. (They’re very helpful once they do answer the chat.)

In the middle of this, my ASUS laptop display turns to a screen filled with “static and snow”. It’s been doing that sporadically for a while now. I’ve downloaded new drivers, twice now, and that doesn’t seem to help. I’ve spent a lot of time on their forums – first, I had to set up an account so I could participate. After posting my question, I got two answers, neither of them particularly helpful, but I was prompted to choose the “right” answer so I picked the least inaccurate one. The suggestions were about updating things and removing things (like I’m about to open up my laptop and start playing around with wires!). It must be a software thing but not a driver thing, and I have no idea, so getting the laptop sent for repair is the best option.

Getting to and navigating the ASUS repair option is another exercise in frustration. The wording is vague and it seems like actually clicking “click here” will wipe my laptop. My partner suggests that I wipe my laptop first, which means I have to ensure that all of my files are backed up. Again, not my first rodeo, and when I had my laptop set up, I asked that any file with data be inside of a folder that gets backed up to the cloud. (Never mind all the alarm around training AI using personal data. I’ll worry about that next week.) But I do take the time to move things off my desktop and make sure that any files I need to work on imminently are on my backup laptop. (What, don’t we all have a backup laptop in case one craps out and you need to keep working?) I work my way through the screens, answering all sorts of questions that require looking through my control panel, clicking all the disclaimers and warnings, and get to the last screen where they want me to upload a proof of purchase. Ostensibly this is to determine whether the repair is covered under warranty (I know it’s not, I just the laptop made usable again.) I go to the site where I had purchased the laptop and download the receipt, because it’s easier than finding the one I downloaded two years ago. I attach and send, and wait for a response.

During all of this, I’ve laboriously copied the laptop’s serial number from a photo I took while balancing the laptop on its side and trying not to drop my phone in the process. But where to paste the information (because this isn’t my first rodeo and I know I’ll end up needing it again, and possibly again and again). Ah, open up a new Word doc and paste it there for the moment. That is, after the cheery messages about something or other that I swat away like flies. (Got It, I click. I have no idea what I just clicked, and really don’t care. I’m carrying as many short-term facts in my brain as will fit right now about whatever new-feature-I’ll-never-use is being promoted.)

A response arrives from ASUS. The laptop is out of warranty. (No shit, Sherlock.) If they’d provided a place to input some text, I could have told them that. So now it’s another go-around with the same interface. They want £45 for shipping, and then the ability to charge the cost of repair to my card without telling me the amount. I get nervous about the idea that they might charge some exorbitant amount, and I bail on the interface. My partner, who manages an electronics exchange shop, offers to navigate that process for me if I’ll just wipe the laptop. He gives me instructions and I set the laptop up for a reset.

Bank interactions are particularly painful

Now, what was I doing again? Ah yes, I need to provide my credit card. My Barclaycard credit card had been declined a couple of days earlier (not wrong password, etc, just out and out declined). I tried to get to the bottom of it because there was no reason to be declined. Went to the banking app and … get a message that the app needs updating. Of course it does.

Sigh. The “Law of Prerequisites” (wherein any given task has one or more associated tasks that must be undertaken before the original task can completed) has escalated to the point where I need to take notes because my short-term memory is overloaded. And I still haven’t fixed my original problem. And now I’m on my third detour, over an hour into my “day off”.

Now, Barclays doesn’t use the usual method of app update. No, they tell you to go find the new app in the app store and install it. So I delete their credit card app and their banking app, go to the app store, download the banking app, and …. the app won’t install on my phone. Do the usual stuff – turn the phone off and on again – but no luck. Guess it’s time to call the bank.

I dread calling the bank. (Barclays one of the big UK banks, and while they try to stay current, they always give off vibes of grandpa trying to “get down” with the grandkids.) The disembodied voice unhelpfully tells me my balance, and then offers some options, one being “do you want to know your balance” and because there’s no option for “your $&!* app won’t download”, it takes me over five minutes just to get through the voice assistant by which time I’m shouting and swearing in frustration. Eventually, a lovely but clueless middle-aged agent tells me they have no idea why the app won’t install on my phone (“do you have over version 5 of Android” – erm, I just told you I have a new phone, and it’s on v13 – and” try cleaning up my memory” – WTF? before I realise that she’s texting their tech group and relaying their responses to me). As I’m now, in fact, locked out of my bank account, I ask her to make a transfer for me and why my credit card was declined. The answer: it must have been a glitch at the venue end because it’s all working as should be. With all due respect, madame, no, nothing is working as it should. The entire system is broken.

Users left with the option of fight or flight

I wish I could say that the self-serve nonsense ended there, that I worked my way back through the Law of Prerequisites to fix the bank situation, then the laptop situation, then the website situation. But the amount of tech in our lives means that fighting with unruly software becomes a weekly, if not daily, occurrence. Corporate greed desire to make even more money for shareholders, at the cost of the sanity of users, is driving even the most stoic of us to self-destruct. During the week, I still had to fight with software that that replicated the shit show customer experience I just described.

• A Microsoft software license subscription issue which took ages because I couldn’t solve the problem online, had no email option, was then confronted with the standard “longer than usual wait times” message for phone, and being directed by their chatbot to a link that led to a non-existent page. And some interface text telling me to contact my IT department for support. (What IT department??) In all fairness, once I went around in circles for about twenty minutes, I finally got to a real person. He was knowledgeable and could untangle the knot of licenses that I was paying for and not using, and got me sorted with exactly what I needed.

• An Air Canada interface that wouldn’t take my input, had no email option, and the standard “longer than usual wait times” message for a phone option with an over 120 minute wait. And after not recognising the tones from my phone, they hung up. Added bonus? They want me to do their UX work for them by asking how I could make the Air Canada experience better. This time I chose flight, wondering if I can find an airline with a less awful experience so I can close my 30+ year user account to simplify my life.

• Another airline I needed to deal with, British Airways, had locked me out of my account. Their interface that wouldn’t take my input, had no email option, and the standard “longer than usual wait times” message simply hung up on me. One less airline account will simplify my life, as I guess I won’t be dealing with British Airways again. (And I did tell the corporate travel agency to book me on a different airline whenever possible.) Definitely flight.

• The Southwark leisure centre’s new website is confused about whether or not I have an account. Logging in doesn’t work, but when trying to set up a new account, the message says my email is already associated with an account. Eventually, my sleuthing pays off and I’m rewarded with access to a degraded interface with a terrible user experience. But at least I can book in for an activity. Well, if I’m willing to fight with their software a few times a week.

SSDD (Same sh*t scenario, different day) in the support (sic) world

The time after all of this is early afternoon, and half of my day off has been spent on these broken self-serve experiences. When my partner asks, “Are you working?” I reply, “Yes, I’m doing unpaid work for the companies that are saving on their operational costs by pushing the burden downstream to their customers.” Don’t think we don’t notice – this “savings” for you has become an unbearable burden for your customers.

Fast forward a few days, and it’s my partner’s turn to waste his day off doing unpaid work for brands, complicated by being in the middle of a switch-over from BT to Sky broadband.

• Sky’s customer service line had an interesting twist meant to create “delight” – a trite term that is the equivalent to offering a child a cookie to make them smile, in compensation for leaving them standing, underdressed, in the snow until they get sick. The hold message provides the usual “high call volumes” and acknowledges that waiting online sucks and allow you to choose the genre of hold music you must endure. My partner’s reaction? Well-played, but what would impress me is if they would hire enough staff to get rid of wait time. Of course, the eventual connection to a person meant they could discuss the problem, but it was not solved. Evidently it’s the fault of BT for cutting off our internet service after Sky informed them to delay for a week.

• I braced myself for the interaction with BT because of the overwhelming dread I felt having to deal with Yet Another Support Bot. Surprisingly, this experience was the least frustrating of all of the interactions to date. They recognised my phone number, sent a text saying they would run tests and inform me of the result, which they actually did. Then, if the problem persisted, I could enter a code word to get someone to call to figure out the problem. Which actually happened. Unfortunately, BT says it’s Sky’s fault that our internet service was not extended because they were never notified, so a good experience still led to a poor result.

Given the time sink and the emotional toll it takes to deal with these brands, the option that would cause the least trauma is to live without internet for a week. For a digital-first household, that’s a pretty telling statement.

No brand loyalty when customer “support” is a farce

Post-sales support is supposed to be a new proof point in a marketing strategy, not an experience that triggers an anxiety attack. If this has become the state of the world, the efforts are falling flat.

When brands have the absolute gall to ask me how to make their experience better, my brain hears: could you do some free User Experience consulting for us because we aren’t investing in that as paid work. To this, I reply with a terse:

1. I’m never going to go through the painful “support” (sic) experience unless I’m desperate and it’s complicated, so give me an option to chat with a human.

2. If you’re going to push people to a chatbot or voice assistant, make those options work for real people, not just a few scenarios dreamed up by your product manager.

3. If I have to interact with a bot, give me an option for “nothing is working, now what do I do”.

4. Don’t expect us to pay any attention to your marketing claims – they mean nothing unless you back it up with good customer support. And we’re on to you.

All the loyalty your marketing department craves will come to nothing if your post-sales experience sucks, because the post-sales phase is a lot longer than the fleeting encounter with the marketing phase. It’s kind of like a bait and switch of a short-and-sweet engagement followed by a long, troubled marriage. Why be loyal to companies that treat you like a housewife trapped in a 1950s bad relationship? If anything, you’re going to constantly be scanning for a better alternative. Customer experience has reached peak farce, and that noise we hear are the rumbles of extreme automation awaiting an explosion of volcanic proportions.

This post was originally published in 2023. It was not updated because the situation has not changed. If anything, the frustrations have gotten worse.

What happens inside the content world seems to be opaque and mysterious. Operating models for content are mostly stuck in the 1980s, taking word processing and spreadsheets as far as possible, but woefully inadequate for a world where content goes far beyond “documents”.

Content is expected to be componentised, re-used, adapted, reconfigured and repackaged, and delivered in multiple ways, into multiple interfaces including AI-enabled chatbots, CMSs, apps, intranets, knowledge bases, support centres, and yes, even PDFs. What can we learn about the need for content operations and why we can’t optimise operations by focusing on the last mile? Let’s get into the topic.

What is “the last mile”?

In supply chain management, there is a concept called “the last mile”. This refers to the final stage – the delivery aspect – of getting a product from its sender to its final destination. In the world of physical goods, the last mile usually covers the movement of goods from a transportation hub to its delivery point.

The goal of last mile delivery is to optimise the process so that the delivery goes as efficiently and cost-effectively as possible. The last mile has nothing to do with the design, manufacturing, packaging, or packing of the merchandise. The delivery aspect is concerned only with what happens once the goods leave the warehouse.

Let’s use the example of ordering food from a restaurant. The food is to be delivered by one of the ubiquitous food delivery services that pick up food orders from the restaurant and drop them off at customers’ homes. There is a strong business need to operationalise the process so that the businesses keep their competitive advantage.

The delivery companies start optimising their processes. They develop an app to track their orders. The drivers use scooters instead of bicycles. They study how to shave minutes from the process, all to make the customer happy. This is creating a strong operating model for the last mile.

The problem with optimising the last mile

In the restaurant scenario, what happens if the kitchen isn’t optimised for faster food production, if the restaurant has a couple of pots, a few wooden spoons, and a bunch of kitchen staff running around, bumping into each other like a scene from the television show, Kitchen Nightmares? No matter how much the delivery services optimise their processes, the customers won’t get their food any faster if the kitchen isn’t organised for efficient production. When the operating model of the kitchen is poor, it becomes the weak link in the supply chain.

In the content world, the last mile is when delivery-ready content is delivered into a delivery platform – there are several flavours of delivery platforms from CMS to digital experience platforms, it doesn’t really matter which one for purposes of this discussion – and awaits publication. This is the delivery mechanism that takes content from the point-of-offer to the consumer of that content.

The entire industry is set up to think about optimising that last mile. The optimisation continues, but unfortunately, optimising the production process upstream, in the content kitchen, so to speak, is not on the industry’s radar. This is more aptly called delivery optimisation, and is on the rise as conversational AI gains traction.

Content operations starts before the last mile

The restaurant scenario is an apt metaphor for what happens in the content industry. There is a lot of work done to optimise the delivery mechanisms – the last mile, using publication-ready content, akin to taking the ready-to-go content and delivering it to the right audiences. These delivery platforms manipulate the content in the ways that the business needs it to do. There is post-publication analysis and improvement of the delivery mechanism. But it’s not actually optimising the operating model for the production of content.

So who is in charge of optimising the production of content upstream, in the “kitchen”? Ask those responsible for the management of content delivery about optimising the delivery model for content production, and they’ll tell you that’s not their responsibility. They are responsible for optimising the delivery of the publication-ready content.

Ask them what it takes to get content produced, and they’ll know shockingly little about the supply chain before content hits the last mile. It’s not dissimilar to the delivery drivers you see waiting outside of a busy restaurant, waiting for their orders to be ready. What happens inside the kitchen is of no matter to them; they spring into action only once the packaged food is ready for delivery.

Who’s fixing the “content kitchen”?

Ask the product team who they think is responsible for optimising content production upstream, and you’re likely to get an “I don’t know” shrug or a guess at “whoever’s in charge of content”. Ask the person in charge of content, and you are likely to be told that’s a tech problem. The problem is that whoever is responsible for content may be willing, but not particularly able, to fix the problem.

So what happens in those content kitchens? Content production remains stuck in the 1980s, and content teams struggle to deliver content that keeps pace with product development, with the ability to scale in a resource-efficient way, with the need to produce regulatory-compliant content, and to reduce time to market.

The question begs to be asked: if the technologists don’t understand enough about content production processes and governance to implement a more efficient operating model, and the content people don’t understand enough about technology to know what is possible to gain more efficiency in their production, then who does? This is often the critical gap that leaves an organisation vulnerable, not just in the area of content production, but in other areas as well.

Sometimes, organisations bring in big-name consultants, usually with disappointing results. One product director told me that the single biggest pain point (and I believe he said exponentially problematic) in a prominent UK government project – it was the COVID-19 Test and Trace app – was content production.

The Accenture consultant assigned to the project either would not or could not recommend an operating model that would accommodate content production in multiple languages, insisting that Confluence could do the job. (It blatantly couldn’t, and any content person worth their salt would know that.) How much this cost the public in time, resources, money, reputation, and effectiveness wasn’t measured, but the best guess is in the many millions of pounds.

The skills needed to figure out content operations spans a number of disciplines. There’s the business analysis, content analysis, process mapping, governance and change management, and technology aspects – as well as someone knowledgeable about the challenges and potential efficiencies to take the lead.

When an organisation is fortunate enough to have a content operations strategist or content engineer on hand, that’s a step in the right direction – but there are precious few of those to go around. Convincing management and the other gatekeepers is often the largest challenge of all, as they’ve heavily invested in the last mile, and don’t want to hear that they now also need to invest in fixing their content kitchen problems.

This problem is exacerbated with generative AI, which management is convinced can do everything, from research and analysis to writing drafts to stakeholder management to managing review cycles to accuracy and compliance checks. The last mile has become a chat or voice interface, with an opaque mechanism behind the delivery of content.

Is fixing the content production process worth it?

The biggest challenge for organisations is to figure out whether the exercise is worth it. It’s said that companies measure what they value, and this is a good time to start measuring what happens upstream. So much waste is built into most content production systems that it’s only once all the waste has been identified, Lean for Services style, that an estimation can be made of how much time and effort can be saved and what benefits can accrue.

Given the mounting evidence that the state of the content in your repository matters immensely to the accuracy of the content that gets delivered via generative AI, as SEO is being replaced by GEO and AEO, and as the rise in user expectation for quality answers, the calculation becomes quite clear. It’s no longer enough to deliver “good enough” content and move on. Now, curating what’s in the repository is as important as maintaining the quality of the original content.

It should be noted that adopting an operating model that optimises content creation generally does not affect the existing technology stack. The delivery platform continues to do what it does best: publish content for consumption by the user. A content operations stack replaces the word processing and/or spreadsheet and/or note-taking software that’s held together by things like Slack, Trello or Jira, and email. The finished content is still delivered to wherever it’s supposed to go, but likely with a lot more accuracy and automation.

Content production is not for the faint-hearted

The content production process is far more complicated than visible to anyone outside the content team. It’s not uncommon for a relatively refined process to have dozens of steps, including multiple editing, review, and approval loops, before being sent off, often by email or copy-and-pasted into a CMS, knowledge base, or other delivery platform. However, the relative cost of content production is generally not in the creation of content. If content truly worked as a supply chain, where content is created once, delivered, and forgotten, then calculating the cost would be quite straightforward. However, that is rarely the case – at least outside of marketing content.

Content operates on a lifecycle, where it is revised, versioned, localised, and so on. In a single lifecycle, it’s not unusual to be able to reduce content production time by well over 50%. Once a piece of content has been slightly modified for, say, a handful of new products, then for a new version of each of those products, a single change now means a change to each one of those ten pieces of content. And the following year, there may be ten more variations.

Tracking those variants through spreadsheets only goes so far, aside from being a slow and error-prone process. The biggest pain points and time wasters turn out to be things that are complete surprises to practitioners outside of content. Once the content lifecycle is subjected to any additional stress, such as adding a language variant, processes can slow to a crawl.

What kind of improvements can you get from content operations?

Here are some metrics from my own work with clients and/or colleagues who kindly shared with me:

• A government department that produced a particular type of content for about 50 audiences. Changing the operating model by moving content operations upstream to the author, calculated a reduction in the number of steps by 49%, production time by 65%, overall production cost by 66%, and length of time for a publishing cycle from 12-18 months to 4.5 months – meaning that they could deliver updated, accurate content to the CMS over 60% sooner.

• A start-up producing software that got embedded in medical devices needed to deliver content with a tight audit trail. They created specifications, code, code descriptions, and so on. By adopting an operating model that tagged all of the content with precise metadata, this company of a CEO and two developers were able to use a “build” command to combine all of the content and generate over 200 sets of content for various clients and uses, including online topics, embedded content, and PDF documents.

• A company producing complicated software – many modules with many functions per module – used a technology that isolated UI strings and labels in a library that could be auto-inserted into the interface and into the instructions on how to use the software that went into online help, training materials, and user guides. The savings in the upkeep of four outputs was reduced by 75%, while the always-accurate content ensured that the users could stay productive.

• A company that changed their operating model from a common method of using word processing software to create topics and spreadsheets to track where the content was used across multiple outputs. They upgraded to an authoring system that allowed them to create topic, auto-track re-used content, update that content in a single place and propagate the update throughout the content corpus with a single click, and deliver updated content to multiple outputs – a CMS, and to a publishing partner to go to print – with minimal effort. They eliminated the 48 tracking spreadsheets completely, significantly reduced production throughput time, and found that instead of expanding the team to deal with growth, the existing team could handle all the content and still have capacity to handle scale.

It may be tempting to say that these are outliers, exceptions to the rule. The truth is that there is too little data yet to make that call. All of the success stories presented at conferences and discussed between professionals seem to show over a 50% improvement, but we have to assume that there are failure stories that don’t get factored into the mix. Anecdotally, however, the organisations that go all-in reap significant benefits. 

Where does the damage happen when it comes to content operations?

If the organisations that commit to the process reap significant benefits of content operations, where do things go wrong, and how does it damage the operating model? At the risk of making sweeping generalisations, let’s look at some of the situations that hobble content operations or do outright damage.

• Gaps in knowledge of staff outside of the content team. The lack of understanding of the rudimentary processes, problematic complications, and governance tensions that content people face on a daily basis is legendary. So when an organisation consults software developers, content management integrators, data scientists, or other technologists about what content people need to operate efficiently, it’s not surprising that they can’t help. They are used to dealing with the delivery side of things – once the content has actually been produced and finalised. When consulted, they tend to recommend software that’s inadequate for content developers to improve their operating model. The reaction can range from well-intentioned but uninformed to downright hostility and everything in between. The alternatives that are offered are often advantageous to those in charge of the delivery side, but damaging to the content people whose time and cost are never factored into the overall cost of operations.

• Gaps in knowledge of staff inside of the content team. Content people – writers, editors, content designers, content managers – are so used to coping with whatever tools they are given that they just limp along. They are provided with word processing tools that were designed for “documents” such as business correspondence or reports. The right workflow tools, or even any workflow tools, are not provided, so spreadsheets are used as slow, manual, error-prone processes. In the final example in the section above, the conversations were quite typical and very enlightening. The department manager and team gave us a short list of “must haves” and a “wish list” that they would like, if affordable. They were surprised to discover that there were commercial solutions that could satisfy both their must haves plus the items on the wish list and a few more items that they didn’t realise they needed until we explained it to them. These conversations happen all too often; management within the content area traditionally have expertise on the editorial side of content, but not on the technical side.

• Gaps in a content operations strategy. If you don’t know where you’re going, any route will do, goes an old saying. A good content strategy will have a section on content operations – operational efficiencies – and how this affects the velocity of content production and, in turn, how this affects the product itself. The background to the strategy should include a current-state operational baseline and a future-state operating model. The strategy should address processes, technologies, and governance. One calculation commonly overlooked is the consequence of content production. One organisation had over a dozen CMSs, and each time a new one was adopted, the product owner would do an assessment of the cost-benefit ratio. However, they neglected to include in the calculation the impact on the content people expected to maintain content in each of those systems. It was impractical to ask the central team of content designers, some staff and some contractors, to learn multiple systems, with multiple logins, with no centralised repository, workflow, or reporting functions. Given that the system is set up once, but the content is produced day in and day out, this oversight can range from highly inconvenient to downright volatile, depending on the needed speed and real-time accuracy of the content being published.

The idea that today an organisation can optimise content production by focusing on the last mile is no longer adequate, bordering on naïve. Using outdated technologies meant for casual business use, with governance models meant for documents, combined with outdated processes  that don’t allow for management of components, the addition of metadata, or efficient delivery of content to the distribution systems is becoming (if it’s not already) a critical bottleneck that flows downstream to the final consumption point.

Note that this post has been updated since it first appeared on Content, Seriously Consulting Ltd. in 2022.

Content professionals involved in pushing the boundaries of content capabilities have spent years talking about intelligent content. We’ve structured it, tagged it, modelled it, and developed taxonomies for it. And yet, for many organisations, content still get locked into PDFs and stored in Digital Asset Management systems as BLObs (Binary Large Objects) or pushed out, with little or no semantic markup, to content management systems with little more than heading and body text fields.

This is exacerbated by the problem of practitioners collapsing content and data into a single entity, trying to manage content the same way as data, ignoring the complexities and nuance that separates the two concepts. That’s an argument for another time, but pertinent to this topic, I will discuss content as contextualised, human-usable data, and information as contextualised content plus data. No matter how you define and categorise the data–content–information continuum, we have to admit that information needs to perform way in more complicated ways than it used to.

Let’s face it, since the popularisation of generative AI (GenAI), the entire paradigm for content findability, retrieval, and delivery has changed.

• The average user never looked for information by poring over a multi-page PDF downloaded to their phone. Now, they can get a small snippet of information served to them from whichever site has the most AI-friendly content.

• The traditional skip-skim-scan method of looking at web pages has gone into high gear. As more people are relying on the AI summaries generated by search engines, they aren’t reading through long pages of text to find whatever nugget of information they need.

• There is a sharp rise of organisations using AI-enabled chatbots to retrieve an accurate, concise answer from a knowledge base, whether for use by customer support agents or by end users. This means that the information needs to be written, structured, and marked up in ways that makes it more retrievable.

In other words, information needs to be kinetic: responding when called by other systems.

Kinesis: a quick physics lesson

The definition of kinesis is movement in response to a stimulus. Potential energy is stored energy; kinetic energy is energy in motion. Applied to information, kinesis is content and data that is retrievable in response to a call by users via some sort of system. The system could be anything from a web page to a chatbot to a voice assistant. You ask your voice assistant what the temperature is in whatever city you’re travelling to; the system goes out to a weather service’s data repository and retrieves the relevant data; the voice assistant injects the temperature data point into a piece of content (in this case, a sentence), turns the text into voice, and reads out the information. That’s kinesis in action.

The history behind kinetic information

Industry 4.0 laid the groundwork for the shift from static to kinetic information, with the idea of automation and data exchange in manufacturing technologies. There are four underlying principles of Industry 4.0 which allow factories to run themselves: smart manufacturing, cyber-physical systems, and the Internet of Things (IoT). Those principles are:

• Interoperability — machines, devices, sensors and people that connect and communicate with one another

• Information transparency — the systems create a virtual copy of the physical world through sensor data in order to contextualise information

• Technical assistance — both the ability of the systems to support humans in making decisions and solving problems and the ability to assist humans with tasks that are too difficult or unsafe for humans

• Decentralized decision-making — the ability of cyber-physical systems to make simple decisions on their own and become as autonomous as possible.

What is not addressed in this explanation is the role played by information. Seen through an information lens, we can bee the enabling function of what became known as Information 4.0. I’ve added, in bold, the phrases that are implicit to the principles.

• Interoperability — machines, devices, sensors and people, and information that connect and communicate with one another

• Information transparency — the systems create a virtual copy of the physical world through sensor data in order to contextualise information

• Technical assistance — both the ability of the systems to support humans with targeted information in making decisions and solving problems and the ability to assist humans, through the use of content, with tasks that are too difficult or unsafe for humans

• Decentralized decision-making — the ability of cyber-physical systems to make simple decisions, guided by information provided, on their own and become as autonomous as possible

A similar situation exists with the Internet of Services, or Service 4.0, where disruptive technology concepts are applied to value chains. A similar exercise in reading between the lines of their principles, we can see the role played by information.

• Big Data and Analytics. Deeper insight using information into customer behaviour, preferences, and pathways

• Bionic Computing. Natural interaction, by proving information, with virtual agents, digital devices, and services

• Ubiquitous Connectivity and IoT. Ongoing connections (on-the-spot service provision, remote monitoring)

• Cloud Computing. Manage large data and content volumes in open systems and provide services on demand

• Cognitive Computing. Simulate human thought processes, provide intelligent virtual help

• Smart Devices. Ecosystem of apps and cloud services serving up content using high-performance devices

• Robotic Process Automation. Replace humans for rule-based work processes, comprehensively documented

• Virtualisation. Remove reliance on specific hardware/software for operational flexibility

• Augmented Reality. Provide needed information on demand (manuals, pricing, alerts)

The increasing importance of kinetic information

This is where kinetic information comes into play. Kinetic information challenges the primacy of the document as the unit of knowledge transfer. Kinetic information picks up where Information 4.0 left off. It’s no longer only about smart factories and IoT. Kinetic information is needed across the enterprise where the contexts vary but the underlying principles are very constant:

• Marketing departments want to personalise their content for multiple audiences in multiple markets.

• Technical writers need to deliver product knowledge in multiple deliverables for multiple products in multiple product lines for multiple verticals in multiple markets.

• Customer support agents, under time pressure, need to retrieve answers to a wide range of questions from a range of customers in a range of locations while assured it’s reliable information.

• Employees need to retrieve information from their intranets, from Standard Operations Procedures (SOP) to backgrounders for business analysis to how to file an expense report, and be confident that they have the right information for their particular circumstances.

The move to ubiquitous computing, where content is made to appear any time and anywhere, using any device, in any location, and in any format, absolutely needs kinetic information to succeed.

Microcontent content as the new unit of information

One of the more consequential shifts in Information 4.0 thinking is the move away from the document and toward microcontent as the smallest unit of content that can still retain its validity in context. Microcontent carries one fact or idea. It has a unique purpose. Microcontent is, by design, ready to be assembled, delivered, and consumed independently of the document that once contained it.

Microcontent goes by several names. (Yes, our industry has a naming convention problem.) Data, content components, microfacts, bite-sized content, snackable content, atomic content, snippets, nuggets. If I were to define microcontent, it would be small, modular units of information that have self-contained meaning and can be used independently. Microcontent is also volatile and state-dependent, which means that it contains enough semantics to be context-ready for real-time, continuous delivery.

This matters enormously for the delivery channels that have come to define how users now seek information. AI-enabled chatbots and query interfaces have changed user expectations around information retrieval. Users increasingly expect real-time, precise answers to specific questions rather than a link to a page where the answer might be. A well-formed piece of microcontent functions as a direct response to a user intent in a way that an editorially structured document doesn’t.

The traditional method of creating new content specifically for each new channel: chatbot responses, dedicated help widget copy, channel-specific FAQs, UI strings locked into software interfaces is not sustainable. A more durable strategy is to restructure a single source of content as microcontent, marked up in machine-readable formats for query-response delivery channels and capable of being called by downstream systems as needed.

What makes information kinetic

To be called kinetic, information objects or units need to have the following properties, arranged into a convenient-to-remember acronym, PODIUMS:

• Profiled: personalised and automatically directed to the right audience according to context.

• Offered: rather than simply stored, content objects are assembled and delivered when called by downstream systems, not just when a user happens to search.

• Dynamic: updated automatically in response to new data, usage metrics, and content needs.

• Independent: usable across contexts without editorial rework or reformatting.

• Ubiquitous: always ready for use, online, searchable, and findable wherever a user might look.

• Molecular: stored as independently-usable modular units of microcontent ready to be delivered as Content as a Service (CaaS).

• Spontaneous: information assembly triggered by context in real time, delivered without human intervention.

The layering of AI across every part of the content ecosystem, from authoring to delivery to conversational interfaces, has changed what systems can do and what users expect. AI doesn’t retrieve documents, though it may cite sources. AI generates responses, and those responses are only as good as the source content. It’s become increasingly clear that unstructured, document-bound content is poorly suited to AI-mediated delivery. Content that is modular, metadata-rich, context-aware, and independently valid is.

Organisations can no longer avoid investing in structured content, information architecture, and semantic metadata. The need to have a robust infrastructure that determines how information performs in an increasingly AI-mediated world is here, now. For organisations navigating digital transformation, content operations maturity, or AI deployment, kinetic information is a useful lens for evaluating readiness. The narrative goes from which complement of content is available to whether content has the attributes required to behave intelligently at the point of use.

In other words, the work of making content kinetic is the work required to make it AI-ready. This ties back to the Content Integrity Model, that balances the strategic, editorial, operational, and infrastructure aspects of content production in order to produce AI-ready, kinetic information.

As an aside, if you recognise your own need to understand more about how kinetic information works, how to formulate a strategy for it, or how to implement it, a relatively new professional association has been formed for exactly that purpose. An umbrella home for the various professions that come together to solve real-world business problems related to content. the Kinetic Council serves to fill that gap for professionals looking for unbiased answers from industry experts.

Content has become a critical component of how organisations deliver value, differentiate themselves, and support increasingly complex digital ecosystems. Yet many organisations still treat content as a tactical deliverable rather than a strategic asset with long-term implications for customer experience, operational efficiency, and organisational growth. As products evolve, markets expand, and expectations for personalisation rise, it is no longer enough to focus solely on editorial correctness or isolated improvements. Ensuring that content consistently performs its role requires a more integrated view, one that considers the entire lifecycle of how content is conceived, created, managed, and sustained.

The Content Integrity Model provides that holistic perspective. It reframes content work across four interconnected dimensions - strategic, editorial, operational, and infrastructure - to help organisations understand not only what content should achieve, but how the underlying systems and decisions support or undermine that intent. Each dimension contributes a necessary part of the overall ecosystem, and strength in one area cannot compensate for weaknesses in another. By approaching content integrity as a system rather than a series of independent directions, organisations can build content environments that are more adaptable, more scalable, and better aligned with long-term organisational goals.

Exploring the Strategic Dimension of the Content Integrity Model

The Content Integrity Model positions content not as a by-product of operations, but as a deliberate contributor to consumer value. Strategically, content integrity sits where organisational priorities intersect with audience expectations and operational capacity. It influences how content is envisioned, planned, governed, created, and maintained, and not just how it reads on the page.

In many organisations, discussions about content integrity begin and end with editorial correctness. While accuracy is necessary, treating content integrity purely as an editorial issue obscures its broader purpose. Content integrity is not just about getting the words right; it is about making intentional choices that allow content to deliver value as organisations expand, evolve, and introduce new offerings.

Content Integrity Begins with Value for Consumers

At its core, content integrity is an organisational decision about the value content should deliver. Organisations that treat their content as a valuable asset are choosing to reach particular strategic goals. For example, they may use content to outperform competitors by providing clearer, more usable, and more supportive content experiences. That value manifests in several ways:

• Minimising friction for people completing tasks

• Simplifying complex products or services

• Supporting users across the full journey rather than only at conversion points

Consider a simple example: onboarding. When users struggle to get started, they disengage. Competitors that provide clearer prompts, guided steps, or contextual help keep users engaged. The difference lies not in the underlying product but in the surrounding content experience.

Elevating consumer value is a strategic investment. It signals that clarity, usability, and relevance are competitive differentiators worth resourcing.

Content Must Align with Organisational Direction

Content does not exist independently of corporate priorities. Strategic content decisions must align with, and anticipate, broader business direction. This typically requires understanding:

• Corporate strategy and expansion plans

• Competitive dynamics and positioning

• Internal capabilities, resourcing, and operational constraints

For instance, expanding into new markets immediately raises questions about localisation, language support, cultural nuance, and scalable governance. Similarly, if competitors pivot to AI-supported help channels, organisations must determine whether they will follow, differentiate, or take a selective strategic path.

Conscious decision-making is essential. Content integrity suffers when choices are made reactively, based on trends or internal pressures rather than on strategic clarity.

Strategic Content Emerges from Business Strategy

Decisions about what content to create, who it serves, how it is delivered, and how it is managed stem directly from organisational strategy. Business goals, marketing direction, and desired impact shape the content footprint.

Strategically significant questions include:

• Which markets need content, and in how many languages or variants?

• How deeply will localisation extend to cultural differences and market norms?

• What content formats and types will be required to meet user needs and remain competitive?

• How will the organisation maintain accuracy and coherence as content volume expands?

For example, deciding to produce short-form video for frontline workers is not just an editorial choice. It has implications for tools, workflows, capabilities, and distribution. It requires alignment across strategy, operations, and content systems.

When strategy is thoughtful and explicit, content operations run more smoothly. When strategy is unclear, teams compensate through workarounds that create long-term inefficiencies.

Operational implications include:

• Designing lifecycle models that support long-term maintenance

• Managing content corpora to maintain accuracy at scale

• Ensuring infrastructure enables strategic execution

Content Integrity Functions as a System

The Content Integrity Model operates across four interconnected dimensions. Weakness in any one dimension compromises the rest. High editorial standards cannot offset poor strategic direction. Strong systems cannot fix unclear objectives. Integrity results from coherence, treating content as a system rather than a series of isolated tasks.

Research, Not Assumptions, Drives Strategy

Delivering content that reliably adds value hinges on research. Without it, organisations default to assumptions grounded in internal convenience. Audience research helps reveal:

• Misunderstood expectations about location, language, or preferred format

• Cultural norms that shape engagement and behaviour

• Differences across markets, regions, and platforms

Research prevents reactive fixes by enabling informed decisions from the outset. What succeeds in one market may fall flat in another, due to language nuance, cultural expectations, or consumer behaviour patterns. In some regions, interactions centre on in‑person negotiation; in others, automation and speed dominate. Strategic content design requires acknowledging and responding to these differences.

Maintaining Accuracy at Scale Is a Structural Challenge

Every organisation wants to grow, yet growth introduces significant complexity into content maintenance. The goal of content accuracy at scale is oversimplified and somewhat naïve. Even within one language, regional variations matter. Content written for the UK does not automatically work for audiences in the US, Canada, Australia, New Zealand, India, or other countries where English is used. Multiply that by languages, platforms, and product lines, and the challenge grows exponentially.

Without strategic management, growth leads to fragmentation:

• Multiple versions of similar content proliferate unchecked

• Terminology diverges across channels and teams

• Maintenance costs escalate without being resourced

Accuracy is often treated as a publishing issue, but it is fundamentally a maintenance issue. Interfaces evolve, products change, and information ages quickly. What was correct six months ago may be confusing today.

At scale:

• A single system change can affect dozens of content variants

• Updates must cascade through languages and markets

• Manual audits become slow, costly, and error‑prone

• Periodic reviews allow outdated content to linger and erode trust

Content is displayed everywhere, unlike code that lives in a central location. Each instance requires validation. Content integrity means acknowledging that maintenance must be continuous, not occasional.

Content Integrity as a Strategic Capability

Content integrity is not about perfection. It is about deliberate planning, alignment, and sustainability. Organisations that treat content integrity as a strategic capability:

• Make intentional choices about where content delivers value

• Prioritise research over assumptions

• Plan for scale, maintenance, and long-term coherence

In doing so, content stops being an operational liability and becomes a strategic asset that strengthens organisational outcomes.

Exploring the Editorial Dimension of the Content Integrity Model

The Editorial dimension of the Content Integrity Model focuses on delivering value to end users. The investment in quality must serve a function; otherwise, organisations would not commit resources to it.

Historically, editorial quality followed the familiar 4‑C model: correct, complete, clear, and concise. These principles worked when content centred around printed materials and early digital documents. But today, content production spans multiple destinations, diverse audiences, personalised experiences, and large‑scale accuracy. The old model is no longer sufficient. Content now requires richer metadata, increased adaptability, and more technical sophistication. This is what some describe as content that behaves kinetically.

Editorial Integrity Requires More Than Accuracy

Modern content quality depends on balancing six editorial qualities that collectively define editorial integrity:

Relevant – Does the content meet the user’s expectations at the specific point of delivery? For example, a page titled “How to Renew Your Subscription” should provide actionable steps, not generic information on subscriptions.

Accurate – Are all details correct and actionable? Pricing, specifications, steps, and context must be reliable and free of omissions or misleading patterns.

Informative – Does the content provide the complete information needed for users to take action or make informed decisions? Does it address nuances, edge cases, or directions for users with atypical scenarios?

Timely – Is the version shown the most recent? Is the content up to date? Users searching for policy guidance or product info must trust they are viewing current instructions, rather than outdated versions that mislead or confuse.

Engaging – Is the tone appropriate for the audience? Technical terminology may be essential for specialists but obstructive for general audiences or users with lower literacy levels.

Standards‑compliant – Content must align with applicable standards, including:

• Plain Language requirements

• Accessibility guidelines (including alt text, colour contrast, and multimodal formats)

• Metadata practices that enable automation and personalisation

• Schema standards that improve searchability and contextual retrieval

• Localisation and translation standards

• Semantic frameworks such as SKOS, RDF, and OWL

• Learning content standards that support interoperability and reusability

These elements often fall outside traditional style guides, which typically emphasise grammar and terminology. As content creation shifts toward automation and AI‑supported production, these standards become foundational to quality.

Editorial Quality Supports Scale and Sustainability

Ensuring editorial integrity is essential in environments where content must be reused across varied outputs: web, mobile, product interfaces, signage, and more. Simultaneous publication and multi‑market distribution demand structures that preserve accuracy and efficiency. Editorial quality can no longer be incidental; it must be systematically connected to broader strategy and operations.

The Operational Dimension of the Content Integrity Model

Previous discussions covered the strategic and editorial aspects of content integrity. The Operational dimension builds on those foundations and focuses on the systems and processes that enable content creators to deliver consistent, high-quality content at scale.

Operational Enablement Converts Intent into Impact

Operational integrity is often underestimated as a tactical issue. In reality, it is a foundational enabler. Its value is internal: empowering the people who create and manage content.

Operational content integrity ensures that content has the characteristics it needs to function effectively:

• Structure – Content should be created using semantic structures that support reuse.

• Semantics – Metadata enables systems to identify and deliver content automatically, improving both automation and human findability.

• Context – Content signals help AI interpret meaning and respond accurately.

• Interoperability – Standardised content enables automated delivery to downstream systems.

Without operational efficiency, the maintenance burden grows rapidly. Copy‑and‑paste reuse creates multiple uncontrolled versions, generating content debt that accumulates over time.

Governance Provides Stability and Speed

Governance is often perceived as a constraint, but within operational integrity it is an accelerator. When governance frameworks are weak or ambiguous, teams under pressure make isolated decisions that prioritise speed over long-term coherence.

Effective governance:

• Establishes shared expectations for quality and reuse

• Clarifies where speed is appropriate and where integrity must take precedence

• Reduces ambiguity and misalignment across teams and markets

Governance is not about imposing unnecessary restrictions. It is about ensuring that decisions can be made quickly and consistently, allowing practitioners to work with confidence.

Technical Standards Enable Automation

Many operational decisions depend on systems and standards outside the content team’s direct control. Technical standards determine how content is stored, structured, and exchanged. These choices affect automation, interoperability, and scalability.

Machine‑readable content enables:

• Structured reuse

• Automated publishing

• Integration with translation, search, and AI systems

• More efficient knowledge management and retrieval

When standards are missing, teams resort to copying, exporting, and manual transformations, introducing inconsistency and inefficiency. Governance ensures technical decisions align with operational realities.

Operations Translate Strategy into Scalable Practice

Content operations reflect strategic intent. Organisations intending to scale globally or rely on AI-driven interfaces cannot depend on manual processes. Operational models must support:

• Rapid delivery with minimal manual handling

• Risk mitigation through consistent, updated content

• AI readiness through structured, machine‑readable content

• Alignment with organisational priorities

Operations, editorial practices, and infrastructure must work together to achieve content integrity.

The Infrastructure Dimension of the Content Integrity Model

Infrastructure is the foundation of the Content Integrity Model. It enables strategy, supports editorial quality, and underpins operational efficiency. Without robust infrastructure, even the best-designed content models eventually buckle under the weight of manual effort.

Infrastructure Is the Often‑Invisible Limitation

Content teams can only operate as efficiently as their systems allow. Disconnected tools, outdated platforms, or systems lacking semantic capabilities slow down the entire content value chain.

Many organisations rely on general productivity tools, assuming they are “good enough.” These tools are useful for everyday business tasks but insufficient for content that needs to be structured, versioned, governed, transformed, and delivered across multiple channels. In such environments, generic tools hinder progress rather than support it.

Infrastructure Must Support Operational Integrity

Three core capabilities define infrastructure that supports modern content operations:

1. Enabling Semantic Enrichment at the Point of Authoring

Content creators understand context best at the moment of creation. That is the ideal time to apply metadata that enables personalised delivery. Without tools that allow this, teams rely on comments, instructions, or downstream guesswork that introduces variability. Strong infrastructure also requires centralised, governed semantic assets such as knowledge graphs and metadata frameworks.

2. Streamlining Authoring Workflows

Automation must occur across the entire content lifecycle, not only at publication. While many organisations automate downstream delivery, few tools support automated drafting, reviewing, and preparing content upstream. This gap forces manual copying, pasting, and reformatting, which introduces errors and slows production.

3. Enabling Interoperability for Omnichannel Delivery

True content maturity depends on interoperability: the ability for content to move across systems with semantic clarity. Structured, machine‑readable content can be assembled dynamically and delivered consistently across touchpoints. This improves translation workflows, data integration, and content aggregation.

Content as a Value Chain

A major barrier to content infrastructure maturity is the failure to recognise content as a value chain. Like other organisational assets, content passes through stages—authoring, review, semantic enrichment, transformation, and delivery. Each stage adds value by improving quality, speed, reusability, and overall return on investment.

Infrastructure as an Enabler of Organisational Growth

Infrastructure is not merely operational; it is strategic. Growth depends on scalable content processes. Robust infrastructure supports:

• Scalability – through structured, reusable content

• Operational efficiency – via automation that reduces manual labour

• Quality assurance – through consistent semantic practices

• AI readiness – through machine-readable structures

Without this foundation, expansion becomes unsustainable.

Infrastructure Enables the Entire Model

The four dimensions of the Content Integrity Model—strategic, editorial, operational, and infrastructure—function together. If infrastructure does not support creators, editors, product teams, and downstream systems, the model degrades. Manual work increases, reuse diminishes, AI systems struggle, and content debt grows.

A strong infrastructure prevents this decline. It enables content to function as a reliable, scalable asset that supports long-term organisational objectives.

A Framework for Process Maturity

Content integrity is achieved through these interconnected dimensions of strategic direction, editorial practices, operational efficiency, all supported by a robust infrastructure. Content integrity emerges only when all four dimensions work together, each reinforcing the others to create an ecosystem that is both resilient and scalable. Organisations that treat content integrity as a holistic capability are better positioned to deliver meaningful value to their audiences, adapt to change, and maintain trust over time. When decisions are intentional and grounded in research, when quality standards extend beyond correctness, when operations reduce content debt instead of compounding it, and when infrastructure enables rather than constrains, content becomes its own value stream rather than a cost centre.

As organisations grow, the volume, complexity, and expectations placed on content will only increase. Meeting those expectations demands more than individual skill or well‑intentioned teams; it requires a mature ecosystem that supports clarity, accuracy, and continuity at scale. The Content Integrity Model provides a practical framework for building that ecosystem. By treating content as a strategic asset, aligning people and systems around shared principles, and planning for long‑term sustainability, organisations can elevate content from an operational burden to a driver of competitive advantage and consumer value.

This article has been made into an ebook for your convenience. You can download it directly from my Substack article.

Content Integrity Model Content Seriously

5.28MB ∙ PDF file

Download

Download

Accessibility guidelines for digital content turned 10 in 2022 – the WCAG (Web Content Accessibility Guidelines) became an ISO standard in 2012. The start of the W3C initiative was in 1995, with version 1.0 published in 1999. Yes, accessibility standards have been around that long! Yet my industry experience tells me that many development teams still don’t think of accessibility until way too late in the product roadmap. The product managers discover that they will need to spend an inordinate amount of time retrofitting their digital products or upgrading their code to meet accessibility standards.

Accessibility is a legal requirement for both organisations public and private sectors across North America and the UK and Europe. It’s not a surprise, then, that lawsuits against companies for accessibility shortfalls rose 64% in the first half of 2021. As the saying goes, you can pay for a lawsuit that you’ll definitely lose and then fix the problem, or you can just fix the problem.

W3C Web Accessibility Initiative https://www.w3.org/WAI/standards-guidelines/wcag/

Increased accessibility lawsuits (Wall Street Journal) https://www.wsj.com/articles/lawsuits-over-digital-accessibility-for-people-with-disabilities-are-rising-11626369056

The EU Accessibility Act (EEA) applies to businesses with customers in the European Union (EU). Enforcement of the EAA is now underway. https://www.levelaccess.com/resources/the-complete-guide-to-european-accessibility-act-eaa-compliance/

If you include plain language as part of accessibility, you can look to the new ISO standard to guide you.
https://www.iso.org/standard/78907.html

How to make a product accessible

Companies who think that fixing their accessibility problems simply means checking against colour contrast and colour-blindness charts and adding alt-text fields to their digital assets will be sorely disappointed. Aside from the easily-fixable misuse of alt text fields for citations and inside jokes, there are ninety other guidelines that also need attention.

Rather than fixing individual accessibility problems as they arise, it’s best to work out an accessibility strategy that looks at a product or service holistically. This approach lets you embed accessibility from inside out: from the code to the design to interactions, to reach the “ignored minority” – 1.85 billion disabled people in the US alone. Here are five aspects to consider when creating your accessibility strategy.

Misuse of alt text (Wall Street Journal) https://www.wsj.com/articles/misuse-of-twitters-alt-text-feature-draws-criticism-from-accessibility-advocates-11657879200?cx_testId=3&cx_testVariant=cx_2&cx_artPos=1&mod=WTRN#cxrecs_s

The ignored minority of people with disabilities (Svetla Kouznetsova on Twitter) https://x.com/svknyc/status/1549430300555575296

1. Choose your expertise with care

Consulting with accessibility experts is more than paying lip service to your strategy. The depth of knowledge that accessibility experts, particularly those who have had to navigate those accessibility issues daily, can greatly shortcut your road to accessibility compliance. Let’s look at two examples of deeper knowledge of how to do accessibility with integrity.

• Digital designer Chiara Angori discusses the nuance involved in communicating through sign language, and how easy it is fall into the “UX theater” of assuming that having used a resource for a short time is a substitute for the experience of someone who lives it daily. https://medium.com/doctolib/how-learning-a-sign-language-made-me-a-better-designer-dc42e564bfe1

• Accessibility consultant Svetlana Kouznetsova, who is also deaf, explains the difference in the usefulness of edited captioning over auto-captioning. The related business benefits are aspects off accessibility compliance that companies often don’t consider.
  https://svknyc.com/social-media-accessibility/

Engaging an accessibility expert who has lived the experience is an investment that pays for itself in record time. Creating an accessible user experience from the first release onward creates credibility with customers, who can be quick to take to social media to complain about accessibility shortcomings. 

2. Start with the basics

First, know your guidelines. The W3C has categorised them into four areas, creating the acronym POUR:

Perceive

Information and user interface components must be presented to users in ways that users can perceive. This includes providing alt-text on images, captions on audio or video, and related accessibility alternatives such as sign language. There are guidelines around colour contrasts for low vision and for colour blindness, and usefulness of the graphics. Perception also means structuring content to be read out in a logical sequence, that tables are usable, and that text is presented in a way that makes it easy to read. Dyslexia is an example of a condition that can be helped with the use of a particular font, adequate spacing, and so on.

Operable

This area, mostly for developers, has everything to do with the ability to use commands from a keyboard.  This doesn’t preclude the use of a mouse, but keyboard shortcuts and other mousing alternatives should be available. Operable also covers providing enough time for users to read and use the content. There are guidelines around authentication, interruptions, and time-outs, guidelines around how to disable visual presentations that could cause seizures, guidelines around the ability to navigate and find content, and guidelines around input modalities – for example, how large a clickable area needs to be so that people with disabilities can easily click in the target area. 

Understandable

This area is squarely aimed at text-based content. Accessibility guidelines prescribe that abbreviations be expandable and have a way to show the meaning of unusual words. The guidelines also dictate that the language level be aimed at an average reading level. The need for consistency is stressed: navigation, identification, and ways to avoid and correct mistakes.

Robust

The term robust includes extensibility, scalability, and the expectation that the content can be used across a wide range of current and future interfaces, including assistive technologies.

W3C Quick Reference https://www.w3.org/WAI/WCAG21/quickref/#to 

3. Know which standards affect you

ISO accessibility standards cover a lot of ground, from built environments to cognitive accessibility to digital products to ergonomics. There may be more than one standard that affects your organisation or even your digital presence.

It should be noted that the guidelines apply no matter what type device the content is displayed on – from website, tablet, and mobile displays to kiosks and other devices that can handle spoken commands. They also apply to a range of disabilities that include impaired vision or hearing, limited movement and dexterity, as well as cognitive and learning disabilities.

 Cognitive accessibility https://www.boia.org/blog/what-is-cognitive-accessibility

Accessibility standards apply to the following categories of content, assets, and interactions:

Website content

Though website content isn’t a strict category, the corporate website is such a fundamental asset that it’s worth mentioning. Website content falls under the W3C’s WCAG (Web Content Accessibility Guidelines) initiative. The W3C updated WCAG 2.1 in 2018, with an anticipated major revision to WCAG 3.0 to include emerging technologies and display mechanisms, from wearables to IoT to using the metaverse.

W3C Accessibility Guidelines 2.1 https://www.w3.org/TR/WCAG21/

W3C Accessibility Guidelines 3.0 https://www.w3.org/TR/wcag-3.0/

Interactivity

Your digital product likely falls under the UAAG 2.0 guidelines, which covers what are known as “user agents” – plug-ins, web-based app, or video and audio players. The guidelines ensure that people with various types of disabilities can access the content through the plug-ins or apps. This ranges from being able to control settings to accommodate different types of disabilities to avoidance of flashing that could cause seizures to better control time-based media.

W3C User Agent Accessibility Guidelines https://www.w3.org/WAI/standards-guidelines/uaag/

Tools

This category covers software developed for use and administration by humans. This could be an intranet, software used by a call centre, wearables, IoT (Internet of Things) systems, and so on. There are guidelines for authoring tools, to ensure that users can work with the tools.

Windows-based code has long been accessibility-compliant by design, particularly around programmatically accessible user interfaces. The most popular CMSes (content management systems) have accessibility built in, both for those who use the CMS, and for the output of the CMS. Some even include accessibility checks.

Cognitive accessibility

https://www.boia.org/blog/what-is-cognitive-accessibility

Accessible ICT products and services

https://www.iso.org/standard/70913.html

ATAG (Authoring Tool Accessibility Guidelines)

https://docs.microsoft.com/en-us/accessibility-tools-docs/

Developing for Web Accessibility

https://www.w3.org/WAI/tips/developing/

4. Implement and test

Implementing accessibility has been a bit of a struggle. Of course, it’s best to think about accessibility from the start, and do an iterative develop-and-text process throughout your development cycle. An organisation following that process would be at a relatively high level on the digital accessibility maturity model.

The reality is that most organisations fall below the threshold that would have continuous compliance. The temptation to prioritise cool over compliant, to take dodgy shortcuts rather than build in safeguards, to make public apologies rather than take preventive measures, are all too frequent.

Accessibility starts at the top, with management directives to incorporate accessibility practices across the board, from code to design to content, and from audits to ongoing testing. Management needs to back up their commitment through a strong governance model, with a change management component that builds in the time and resources to allow for the needed activities.

5. Monitor and educate

Accessibility is not a project but a programme. As long as content and digital assets are produced, software is developed and systems implemented, existing technologies deployed and new technologies emerging, there is an ongoing need to stay on top of accessibility compliance.

At some point, the logical progression is to have a dedicated resource – this may be a part-time, full-time, or outsourced role – who knows and understands the organisation and its accessibility challenges. This person’s role is not only to monitor the state of affairs, but also to educate new hires and update existing staff about new regulations and guidelines. Accessibility affects so many aspects of an organisation that a module on accessibility should be part of onboarding, sitting alongside the modules on company mission, ethics, and cyber-security.

We don’t do our clients any favours by ignoring accessibility requirements – or letting them postpone the implementation. It is in our best interest to inform and educate them about the regulations and offer to provide them with a plan to reach a state of compliance. While the client takes the ultimate decision on whether they want to implement accessibility now or later, we owe it to them to point out the benefits of compliance, and also the risks and potential repercussions.

This article was originally published in 2022 and updated in 2026.

That’s become a less common scenario in today’s business environments. Content production is circular, and works on a lifecycle model of creation, orchestration, and delivery, repeated many times over. This creates a very different set of tensions and complexities than exist in a supply chain.

Working within content lifecycle means recognising that the business of producing content follows a standardised, repeatable process. The sub-processes may be subject to variations – the lifecycle for marketing campaigns is very different than for product content, and even within product content, there are variations within the production cycle – yet the overall process must remain stable and predictable.

The content lifecycle describes an organic system, and is system-agnostic. The phases of the lifecycle take into account the production of content from cradle to grave, from the planning of the need for content, through the base lifecycle and all of its subsequent iterations, through to when that content is retired. Those iterations can run from a single cycle to dozens or hundreds. Content can survive many sprints, many iterations, and many products.

Using the analogy of constructing a house, an architect creates a blueprint that specify not only the structural dimensions, but also the heating, ventilation, and plumbing. These plans are what the builders work from. Similarly, a content strategist creates a blueprint by which designers, writers, and developers can build a successful model for delivering content. A strategist is similar to an architect in that way, designing for the system by which content producers will manage content throughout the lifecycle. There are many types of strategies for many circumstances and many content genres, trying to systemise content production for many different purposes. However, there are commonalities across the board. Much as every house has some sort of floor, roof, and entry way, every content lifecycle has four general phases:

• At the beginning is the strategy phase, where you decide what type of content is needed, or whether content is needed at all, and what happens at each stage of the complete lifecycle.

• Then, there is an acquisition phase, in which the content is collected, either by creating it or getting some another source. Most editorial functions falls into this phase, ensuring that content will make sense to the person ultimately consuming the content. Translations and localised content fits into this phase of the lifecycle.

• The third phase is orchestration, which involves ensuring that a framework exists that allows content to become kinetic. In this phase is schema adherence, aggregation of content components, data inclusion, semantic enrichment, enhancement for AI, and a multitude of technical aspects that help content conform to pertinent rules and be processed by the tools that enable content kinesis - that is, able to respond to calls from downstream systems.

• The fourth phase is delivery, where content is ready for retrieval by systems to be consumed by AI, search engines, and ultimately by people. Measurement of delivery effectiveness also determines what happens next: is it the end of the line for this content, or does the content go through the next iteration?

A content lifecycle is likely assisted or enforced through the use of technology, but it is actually conceptual. A content lifecycle governs content production, whether the lifecycle is explicit – codified knowledge used to govern adherence, or it can be implicit – a tacit understanding of how the lifecycle works and reinforced by behavioural patterns.

Lifecycle variants

While all content has a lifecycle, not all lifecycles are created equal. An organisation may have many content genres, each one with its own lifecycle. For example, marketing content works on a very different rhythm than product content. (By product content, I mean all of the content artefacts that go into the product lifecycle, such as engineering specifications, instructions for use, standard operating procedures, troubleshooting and maintenance instructions, API documentation, release notes, production descriptions and feature lists, warranties and disclaimers, subscription renewals and return policies, disposal instructions, user support content, chatbot conversations, and a host of other content that gets produced during the product lifecycle.)

Within product content, several types of content can come together, each with a different lifecycle. Content, such as a compliance disclaimer, might get used once and then be reviewed on a regular basis. Other content gets aggregated from multiple sources for presentation as an integrated unit—for example, product descriptions sent by vendors, pricing from an ERP system, and a publishing cycle with multiple dependencies, from promotional schedules to geo-boundaries. In the world of technical content, an entirely different set of tensions inform the content lifecycle. Conditional processing, re-use maps, and publishing pipelines are of paramount importance.

The introduction of generative AI has had a noticeable effect on the content lifecycle. While the wider industry is still grappling with how to use AI effectively within the content lifecycle, new twists to content lifecycles are being implemented by forward-thinking practitioners in the content space. Technical writers seem to be particularly innovative when it comes to incorporating AI into their processes, which in turn affects the overall lifecycle - mostly for the better.

The content lifecycle is the foundation of every content strategy, whether it’s for an editorial, multimodal, or technical strategy. The lifecycle affects how a content ecosystem is built to facilitate the production and delivery of content. A content lifecycle eases the process of developing a strategy and becomes the backbone that holds together the operating model for content production.

Because artificial Intelligence is an area that is so wide and so varied, We focused on the aspects that interested digital leaders. During the day, I made notes on what people were discussing and then put together a mind map of sorts to create a visual that would cover the topics of interest to the digital leaders in the room. Everyone seemed to somehow be interested in AI as it related to content: content generation, content production, content management, and that’s what was covered. This is what the starting point looked like.

Traditional AI

The first thing to note is that, in general, we seem to make a distinction between two eras of AI: traditional AI and generative AI. It may seem odd to call AI “traditional” but considering that AI has been around since the 1950s, there is a long and progressively sophisticated history of use of AI in tools meant to improve and/or automate content. Some common uses are in the following areas:

• Translation management. There’s been machine translation and translation memories since the 1950s, and neural machine translation has been around for at least a decade now.  It’s quite a mature industry now. This year, all of the major platforms are marketing their platforms as incorporating Ai in one way or another.

• Content optimisation. We’ve had things like spelling and grammar checkers for a long time now. The first AI-enabled content optimisation application used by multinationals was launched in the early 2000s. The app added to the usual spelling and grammar checking with tone and voice checking, disambiguation of meaning, and sentiment analysis.

• Image recognition. I remember how exciting it was when my friend, a Software Engineering Program Director at my alma mater, showed me how image recognition worked and some of the projects that he was involved with at the university. Today, we use this research to recognise, recall, and manage digital assets. Image recognition began in the 1950s and AI-enabled auto-tagging really gained ground in the mid-2010s.

• Data analysis. Tool-enabled data analysis goes back for a long time, back to the 1950s, and become more sophisticated with the application of artificial intelligence. For example, academic papers discussing the advances in how medical data is analysed with the help of AI is mind-boggling. Robotic process automation (RPA) was coined in 2012, and combining RPA with AI moves data analysis to another level.

Generative AI

The year 2022 was the break-out year for generative AI. Despite being around since the 1960s – the first LLM was Eliza, which powered the world’s first chatbot, there was little innovation until the mid-2010s, when natural language processing was widely adopted. So what changed? ChatGPT and DALL-E 2 were given simple, easy-to-use interfaces that democratised access to them. Anyone with an interest could interact and see instant, tangible results. There have been some basic uses of GenAI that we could have anticipated, such as replacing writers in order to produce rather mundane long-form content. Equally, there have been some extremely creative uses that have led to net new value.

Given that AI doesn’t actually “create” but “generates” through reformulating and re-interpreting existing content. It took a while until the limitations were understood. Generative AI was widely touted as being the miracle solution to all content problems. The “tell the LLM a few vague sentences and you’ll get award-winning text-based content and images” phase was short-lived. The word salads, the hallucinations, the unrealistic body appendages, and the invented references dulled the shine on the new toy. However, as the technical folks and the tech-inclined content people experimented and the dust began to settle, there was some clarity around what worked, what didn’t, and how you could improve the results produced.

• Formulaic content. This type of content was generated even before AI came on the scene, but generative takes it up a notch. Reporting of seismic activity (an earthquake of [magnitude] on the Richter scale was reported at [hh:mm] on [date] at [locale]…), summarising sports results, or any other content that follows recognisable patterns can be reliably generated and is a prime application of generative AI.

• Creative content. Generative AI is not great at producing long-form content that needs some sort of analysis or depth. Analysis and creativity is still in the power of the human brain.  However, the creative uses are good as enablers, like having a personal assistant but one that never tires, sits by your side, and ruthlessly edits your text. It can be a muse, generating (mostly) relevant points that can be structured and fleshed out into useful content.

• Optimisation apps and AI companions. I’ve grouped them together because there is some sort of ongoing interaction, despite those interactions being vastly different. Not all LLMs are created equal, and some have been developed with particular functions in mind. AI companions could mean apps such as Microsoft Copilot, which acts as an assistant to content developers to optimise content quality – what used to be called an “assistant” – as well as assistance with drafting of emails and texts – a sort of predictive completion. This could include voice assistants such as Amazon’s Alexa or Apple’s Siri.  Another type of AI companion, using a very different type of LLM, carries on conversations as a “synthetic friend”. These companions can chat, play games, roleplay, or even give advice.

• Image generation. Between last year and this year, the quality of the generated images has gotten exponentially better. Previously, a generated image of a person might yield someone with three hands or one hand with twelve fingers. Today, you’ll get a much more realistic look. Of course, the output is still generated from existing styles, whether that’s steampunk to Rembrandt to Hockney – anything that the LLM was trained on.

• Data analysis. AI already allows large volumes of complex data to be speedily crunched. What Generative AI brings to the table is the ability to generate more and varied insights. The LLM’s understanding of language and patterns can produce results that could take ages for a person to realise. As well, generative AI contributes to predictive analytics – academic papers about cancer predictions through data analysis explain the importance of these developments – that have far-reaching implications. New aspects of process automation are also available. Rote tasks could be automated before generative AI; the ability to easily articulate more sophisticated tasks expands its repertoire. A simple example is the generation of Excel formulas and VBA code.

By the time we finished our discussion, the graphic looked something like this (but messier).

Discussing the universal issues

The other discussion points that came up had to do with universal aspects of AI across the board. These are not related to function or efficacy, but to social implications arising from the rapid adoption of artificial intelligence, particularly now that generative AI can be used by a wide range of people, many of whom have little or no training in setting appropriate parameters when generating content. These eight discussion points are not a comprehensive list but a good starting point for thoughts during your own pursuits using generative AI.

• Information integrity. When newspapers were queried about articles referenced by ChatGPT which turned out to be non-existent articles, this known problem became a hot topic. What are the sources being used to train the models? How accurate are the sources? Are they traceable? Can citations be verified for authenticity?
  ChatGPT is making up fake Guardian articles. Here’s how we’re responding – The Guardian
  Large Language Models Don’t “Hallucinate” – Better Programming
  Different ways of training LLMs – Towards Data Science

• Ethics. There is little to no transparency about what data is collected and how it is being used. Not only is data gathering exploding, so is data brokering. The uproar about the change in Zoom policies is just the tip of the iceberg. What privacy measures are in place? Can they be identified if data from multiple sources is concatenated? How good is the data quality? How transparent is the data gathering? Is it being brokered to reputable entities? Is there an ethicist overseeing data collection?
  Artificial Intelligence: examples of ethical dilemmas – UNESCO
  AI Ethics And AI Law Clarifying What In Fact Is Trustworthy AI – Forbes
  Great promise but potential for peril – Harvard Gazette

• Bias. Because the models are being trained on historical information, inherent cognitive bias creeps in, getting replicated in the LLMs. Are people being denied social, financial, employment, and other services because of bias? Is software for government, law enforcement, and social welfare being examined for bias? How likely are people to be misidentified by, for example, facial recognition? Are population segments most affected by biases involved in the testing of bias?
  Generative AI models are encoding biases and negative stereotypes in their users – Science Daily
  Eliminating bias in AI may be impossible – a computer scientist explains how to tame it instead – The Conversation
  Ageism, sexism, classism and more: 7 examples of bias in AI-generated images – The Conversation

• Model quality. As models are trained on information that includes some generated by the LLM itself, we can see temporal degradation. In addition, some agent provocateurs have undertaken model or data poisoning to thwart training of models using their intellectual property in an attempt to protect creator copyright. How do you know whether the model you use as a consumer has degraded? Are you being informed about any data poisoning that may skew your reliance on generative AI for data?
  Why Machine Learning Models Degrade in Production – Towards Data Science
  91% of ML Models degrade in time – NannyML
  Emerging cyber-security threats in 2023 from AI to quantum to data poisoning – CSO
  This new data poisoning tool lets artists fight back against generative AI – MIT Technology Review

• Social. Social implications is a vast topic on its own. We’ve seen the application of AI in countries such as China (overt) and the USA (covert), creating a surveillance society. Since the time that Cambridge Analytica manipulated first voters through micro-targeting, there has been an explosion of the use of AI to influence election outcomes. How can we protect ourselves and the vulnerable from being exploited? Where are the checks and balances that preserve our social structures?
  The Global Expansion of AI Surveillance – Carnegie Endowment for International Peace
  AI Cameras Took Over One Small American Town and Now They’re Everywhere – 404 Media Podcast
  The AI-Surveillance System Symbiosis in China – Center for Strategic and International Studies
  Artificial Intelligence Has the Power to Destroy or Save Democracy – Council on Foreign Relations

• Sustainability. The carbon footprint of AI is substantial. Generative AI takes massive amounts of computing power, which is driving both the expansion of data centres and the energy consumption to keep the servers cooled. From the extraction of raw materials to  with the production of the supercomputers, the construction and powering of the data centres, and the ongoing costs from training to models to customer usage, there is a large environmental cost, at a time when industry is pushing the envelope of how much the earth can take before reaching a critical tipping point. When will standards for measuring AI’s carbon footprint emerge? How can we measure our own carbon footprint when using generative AI? What is the real cost of generating those dozens of images for one-time use on social media?
  We’re getting a better idea of AI’s Carbon Footprint – MIT Technology Review
  The Carbon Footprint of Artificial Intelligence – Communications of the ACM
  A Computer Scientist Breaks Down Generative AI’s Hefty Carbon Footprint – Scientific American
  AI’s Growing Carbon Footprint – Columbia Climate School

Balancing positive outcomes with cautions

AI can have negative outcomes, as we have already seen in dozens of high-profile scandals. AI can be a tool for bad actors to spread misinformation, but misinformation can also be spread “without malice” – when a model answers a question with a best guess which, in turn, gets lots of engagement, which leads existing reporting structures (that base accuracy on engagement levels) to boost the inaccurate information.

There is an obvious lack of moral judgement of an LLM, which can create some dubious results (see ethics, social, and bias, above). People who barely understand technology, let alone AI, are now targets. Scammers are looking to AI for new ways to separate marks from their money. Deepfakes are being used to manipulate, humiliate, or titillate. Voice manipulation has been used for everything from monetising actors’ voices without compensation to investment fraud to attempting to subvert troops during war.

Artificial intimacy is another aspect that can lead to misinformation. It ranges from the relatively benign “I’m sorry, I’ll change my answer to agree with you” to becoming complicit in the downward spiral during a mental health crisis.

We owe it to ourselves to ask questions such as: How will we protect the marginalised and vulnerable members of society? How are regulations being updated to factor in developments in AI? Can regulations be enforced, and would the impetus be there to enforce them? How can regulations be enforced if detection is not possible? How can we protect ourselves and be protected from the increasingly sophisticated emotional manipulation seeking to influence our decisions?
Deepfake Technology: Assessing Security Risk – American University
Can AI therapy solve the mental health crisis? – New Statesman Podcast
‘It could have me read porn’: Stephen Fry shocked by AI cloning of his voice in documentary – The Guardian
The Dark Side of Generative AI: Automating Inequality by Design – California Management Review, Berkeley University

Let’s end of a hopeful note. There are many positive outcomes from use of generative AI. It would not be fair to discuss so many cautions without covering some of the ways that AI is being used for good. AI-powered robots are teaching English to preschoolers. Advanced data analysis is leapfrogging medical findings. Work productivity and opportunities for net new value abound. Using AI when monitoring, say, network traffic, can detect and  deflect threats. The ability of AI to connect events that could be overlooked by humans can help us make better decisions. AI has the potential to find errors in formulas and solve complex problems, whether averting danger to patients, business forecasting, or modelling climate change.

That said, while AI has the potential to help with the analysis and implementation, the impetus to do so still lies with humans. There may always be bad actors, but there are always good actors, as well, who bring ideas to the table along with the ethics to ensure that as the tides rise, we are raising all boats, not just the yachts.
Can Artificial Intelligence Help See Caner in New, and Better, Ways? – National Cancer Institute
Generative AI could transform work; boosting productivity and democratising innovation – Oxford Martin School
How can AI support diversity, equity and inclusion? – World Economic Forum

Note that links to articles were all verified 2023-11-05. Many other sources were consulted before and during the creation of this article, and only a few chosen in the interest of not overwhelming readers with information.

The post How AI and Generative AI are affected by, or effect, content – a summary from a digital leaders’ event appeared first on Content, Seriously Consulting Ltd..

Outside of the marketing area – or perhaps more appropriately, the wider content area, including marketing but far greater reaching – is where practitioners such as Noz Urbina ply their craft. Noz, founder of the OmnichannelX conference, defines omnichannel as the “unification of engagement and communication strategies so that they complement each other – rather than run in parallel – to give the audience what they really need.” In other words, the end goal is to deliver relevant content to end users no matter where they are in the user journey.

Producing content for omnichannel delivery

Achieving the goal of a great user experience through omnichannel techniques depends on automation. To use a rather clumsy metaphor, it’s a bit like having a servant who, upon hearing your demand, runs to a well-stocked room and fetches it for you. It doesn’t matter what you ask for or where you expect to have delivered, it’s just there. In a digital world, that means having a rich bank of content that is tagged up and ready to call, just waiting for the demand of the end user.

Recently, while booking travel, I thought about the omnichannel implications on my customer journey. The general stages were searching, booking, managing, and checking in. Airlines have come to understand which pieces of information are needed at what point in the customer journey. After I had booked my flight, I was particularly interested in knowing details of my luggage allowance. There was no point in presenting that to me at the pre-booking stage of the customer journey. However, in the post-booking stage of managing my journey, the airline offered me a link where I could consult the luggage allowance for my particular fare.

Omnichannel vs multichannel

Compare the delivery of luggage allowance content – omnichannel – to delivery of the boarding pass – multichannel. At check-in time, I can choose to get my boarding pass as a PDF attachment to an email, which I could print, as a QR code on my phone, at an airport kiosk, printed on thermal paper, or at the check-in desk, where the boarding pass is issued on cardboard stock. The information is exactly the same – personal information and a scannable barcode – but served up looking a bit different, with responsive design and adaptive content, according to the device being used to display or print the information.

How content operations contributes to omnichannel

If content operations is about working with efficiency, much of it is about automating the process as well as the outcome. It stands to reason that a key way to keep pace with the demands of an omnichannel strategy is to pair it with a content strategy. An omnichannel strategy plans for the details of delivery; a content strategy plans for the production of content to fill those needs. In other words, the content strategy needs to account for which stage of the user journey a person is at, what content they’ll likely need at that stage, and how to make the content relevant to a particular user segment or even down to an individual. 

The strategy consists of user research, persona development, user journeys, domain models, taxonomy development, and content typing, which leads to an overall content model. The strategy is operationalised by breaking up the content into reconfigurable units, making them semantically rich in both structure and descriptive metadata, and ensuring that they can be combined and recombined in ways that meet the needs of content consumers, no matter which stage of their journey, which type of device they are using, or where they might be located.

Figuring out omnichannel is complex but does not need to be complicated. It’s a matter of looking for information in the right places. If you’re in omnichannel for retail or marketing, a simple search will turn up dozens of search results for you. Outside of the marketing arena, you’ll want to look at resources such as these:

• The OmnichannelX conference, which covers a range of topics about omnichannel delivery

• The Intelligent Content Primer, from Ann Rockley, Charles Cooper, and Scott Abel

The ContentTECH conference, a little more geared to marketing but still valuable to practitioners outside of marketing

The post Figuring out omnichannel appeared first on Content, Seriously Consulting Ltd..

Sources of resisting content operations

The resistance falls into a few general scenarios.

• The management with budget on the line for the investment are unfamiliar with the need for the technology. They insist that they have a CMS that they are convinced is enough of a solution to solve their content dilemmas. They would rather accept a partial solution than invest in tooling that goes hand-in-hand with governance and process changes.

• Management talks to procurement, who decide that the new tooling is above some corporate level of spending, and they block the investment. Sometimes, the block is an outright rejection based on cost; other times, they delay the investment until they seek confirmation from outside sources.

• Management or procurement consults the CMS developers, data scientists, computer engineers, or other technologists who are involved at the delivery end of content but only a passing knowledge of content production processes upstream. The technologists decide that the tooling isn’t necessary, based on a superficial assessment.

• Less often, content producers themselves are resistant to change. They feel that they’ve managed to optimise their processes using a hodge-podge of tools such as word processors, spreadsheets, process ticketing, and code hosting software, and they don’t want to go from moderate pain back to dire pain.

Assumption-based vs evidence-based decisions 

The common factor in the various types of resistance is the method used to make their decisions. None of these groups consult the content production department to determine what the real pain points are. No one sends in a business analyst to document processes. There is no Gemba audit or 5S exercise to look for the multiple types of waste throughout the content production processes. The decision not to investigate is because the management and/or procurement and/or developers make assumptions about what is involved in the production process. 

The assumptions made are generally way off base. A few of the more dramatic statements I’ve heard as reasons to ignore content operations include:

• Why would you need [that] feature? I mean, how often would you use it – once every couple of months? (How about multiple times a day!)

• They write, get it approved, and move it to the CMS. What’s to improve? (Instead of 3 steps, try over 70 steps with multiple forks, loops, and iterations.)

• The writers we have here wouldn’t be smart enough to use more sophisticated tools so they’ll have to use what they have. (I was so taken aback that I didn’t even know how to respond.)

• We’ve decided it would take an extra sprint to connect the authoring tool to the production system, and we don’t see the value. (Contrasting the cost of an extra sprint with multiple years of productivity seems like a superficial comparison.)

• I’ve worked in this field for 15 years, and I’ve never heard of the tooling you’re describing; sounds like voodoo to me. (Seeing as the tooling had been around for 20 years at that point, I couldn’t help him with his substantial knowledge gap.)

If the decisions not to proceed were evidence-based, I could respect that. But not getting the facts and basing the decisions on gross assumptions? That’s just bad business.

The cost of doing nothing

While the cost of adopting new software and the related change management is relatively straight-forward to calculate, no one does a head-to-head comparison of before and after processes. In other words, the cost of “doing nothing” is not calculated against the cost of adopting new tooling or processes. In the comparison grid, the “Keep the Status Quo” should be calculated, to determine the actual costs incurred on an ongoing basis.

Each organisation is different, of course, but it is not unusual to get anywhere from 40% to 80% improvement in productivity. Here are some case studies that either I’ve worked on, that colleagues have worked on, or have been presented at conferences by clients.

• A financial services company where improvements on a single piece of content reduced the overall effort from 1-4 hours each of 12 people – writers, translators, subject-matter expert approvers, legal, and developer time – over a 2-week period to 5 minutes of 1 person’s time. No matter which hourly rate you use to calculate costs, the savings were impressively massive.

• A company that produces training guides for multiple trades across multiple states in a couple of languages, had a detailed copy-and-paste method over some 12 outputs, tracking that content over almost 50 spreadsheets. Instead of hiring a second manager to add needed capacity, they invested in fit-for-purpose tooling and changed the operating model for their content production, thereby eliminating their use of spreadsheets and automating the multichannel publishing from a single source – without hiring a second manager and freeing up the existing manager to do more value-add work.

• Calculations of content production waste for a large healthcare equipment manufacturer uncovered numerous areas where efficiencies could be gained, with customer-facing content re-use hovering around 80% and internal-facing content (maintenance technicians and customer support) over 60% for an average savings of 70% in the production cycle. More importantly, the error reduction in customer documentation dropped dramatically, improving customer trust and reducing support calls.

• Measuring waste in a government department demonstrated that investing in production-grade software to improve content production would reduce the cost of maintaining content by over 60%, and assist with personalised delivery of content – a problem that had plagued several teams over several years. The cost of doing nothing also resulted in dozens of temporary contractors being brought in to help meet urgent publishing deadlines, costing taxpayers well over £2M across two years.

Caveats about quantifying content efforts

The temptation to throw some resources at the problem sometimes ends badly. The importance of employing the skills of someone who has domain knowledge, has kept up with the range of potential tooling, and understands which tooling is important and which tools to choose in particular situations cannot be emphasised enough.

• Sometimes, the best outcomes aren’t identified because the analyst or consultant doing the discovery does not understand enough about the content production domain to ask the right questions. Or they ask only about the delivery end and not the actual work environment, because they don’t understand the nooks and crannies where the most efficiency could be gained.

• Often, the recommendations are based on using the same old, same old tooling as before. If you’re not tired of hearing the metaphor: when you ask people how they’d get across the Atlantic, those who don’t know that planes exist will describe a boat. When you describe an airplane to them, they will continue to describe different types of boats, but not an airplane. In calculating potential content production efficiencies, a true comparison is not done because of substantial gaps in the knowledge of staff, technologists, and the analyst documenting it all.

• The other situation that is quite common is professional hegemony. The company, while quite willing to invest in some technologies, simply don’t see the value of investing in tech for content operations, and as their opinions dominate – think of the last time that a content designer was asked about the suitability of tooling for the, say, data scientists vs how often content designers are coerced into using tooling meant for code because the developers’ opinions were that the tools they use for code should work for content.

As the bottlenecks increase, the cost of production increases, and the inability to respond to customer demands for information, I predict that ignoring the cost of doing nothing will no longer be acceptable. Organisations will be forced to look at their content operations in order to meet the demands of their customer experiences.

The post The cost of not improving content operations can be steep appeared first on Content, Seriously Consulting Ltd..

Content Ops is about production efficiencies, looking at end to end processes, eliminating the many non value add activities and then streamlining the value add activities. So what makes content operations foundational to omni channel, let’s look at the five ways that content operations supports your Omni channel needs.

The first way is to create content and components so that the CMS can mix and match them, like Lego blocks to manage different messages as needed.

The second way is to create content with predictable structures called schemas. This helps you under your CMS understand how to process the content and present it automatically.

The third way is to use semantics, different forms of metadata so that the CMS can automate the delivery of the right content to the right audience.

The fourth way is to specify which content is important to which stage of user journey so that the CMS can automatically show the right content to the right audience at the right time in their user journey.

And the fifth way is to combine all of these techniques to create a continuous delivery pipeline without continuous intervention.

The post Five ways that ContentOps powers omnichannel [Video] appeared first on Content, Seriously Consulting Ltd..

First, let’s agree on  terms

Before we go any further, we do need to clarify a few basic terms around what types of content assets a CMS actually manages. Note that because of naming conventions across disciplines, some of the descriptions have been modified to have more relevance to practitioners in both the digital asset and text-based content communities.

Digital assets

Digital assets are a type of content, typically of formats such as images, audio, video, or graphics. A Digital Asset can also refer to a text-based object such as a document or a PDF. A digital asset, sometimes called a Binary Large Object (BLOB), is delivered in its final form and used “as is” by any system that publishes it.

Text-based content assets

The term “content” often refers to text-based assets, but should be called by a more specific term. Text-based assets differ from BLObs because they have more complexity to them. A user interface (UI) label can be used on its own, or used as part of a UI string, and either the label or the string can be incorporated into a larger chunk, such as a topic or an article. Any of these can be pulled into one or more systems for publishing, or the content can be part of a BLOB, such as a non-readable PDF, and then published in the same way as a digital asset.

Metadata

Is metadata content? It could be, not not always. Some metadata is text-based, such as meta descriptions or alt text, though some may exist in data form. Metadata is used by the various systems involved to selectively route, filter, and publish digital assets or text-based content. The three types of metadata – structural, descriptive, and administrative – are used for purposes such as personalisation, multichannel and omnichannel publishing, and for establishing an order or sequence for display to the content audience.

Data

Data is in a category of its own. It’s not content because, without context, it is of limited use to a content consumer. Reference data can be used, combining data with text-based content, to enrich the context that increases human comprehension. For example, the physical dimensions of a piece of hardware are data points, but when what those data points mean are explained in a data sheet, the data sheet itself is content.

Second, let’s talk content management systems

Next, let’s look at the types of content management systems that are likely to come into play. In the bigger picture, almost any type of system that ingests, stores, and delivers content could come under the umbrella of a CMS. There are several types of systems that could co-exist within a content ecosystem that handle different content types. They all fit the description of a CMS, so in this article, we will avoid the generic term and assign different names to each system type. 

Digital Experience Platforms

A DXP is the new name for the old concept called a [Web] Content Management System (WCMS). Despite decades of saying that this “manages” content, this type of system focuses on content delivery. The publication-ready content – a finalised photo, video, audio, graphic, or text-based asset – is moved, manually or through an API, into the system, where the DXP decides when and how to display the asset, and to whom. The DXP may include other post-publication functions such as performance analytics.

Digital Asset Management

A DAM system, much like a WCMS, focuses on the delivery side, but stores and manages digital assets. There are, however, a couple of fundamental differences between a DAM and a WCMS.

• A content creator, such as a photographer, videographer, podcaster, or illustrator, does not do the actual work in the DAM, but in a tool with specialty functions. A photographer may do their photo editing in Adobe Photoshop; a videographer may use Final Cut Pro; an illustrator may work in Adobe Illustrator; a podcaster may edit in Audacity. Once there is a publication-ready version, the assets are moved into the DAM. (The drafts may be moved into the DAM for storage, but the work on the files themselves are not done in the DAM.)

• The DAM likely sits between the work environment and the DXP. In other words, when it’s time to use the publication-ready assets, they are pulled by the DXP, which then routes the assets to be displayed in the appropriate place, at the appropriate time, to the appropriate audience.

Product Information Management 

A Product Information Management (PIM) system – also known by Product Data Management (PDM), Product Resource Management (PRM), Product Catalogue Management (PCM) – stores product information, such as product name, product description, images, and data attributes such as dimensions, weight, colours, and so on. There are many functions associated with a PIM, but for the purpose of this article, think of a PIM as the same category as a DAM; the PIM is the single source of truth for product information, and helps the DXP to publish the right information to the right audiences at the right times.

Content Operations Management

Content Operations Management (COM) is a class of systems that help manage text-based content during the creation and approval process, as well as communicate with a DAM to connect to digital assets. Content creators are notorious for repurposing whatever tools are at hand, even when the long-term effects to content may be detrimental. The bulk of authors and editors will use bog-standard tools such as word processors or spreadsheets, not because they are COM tools – these are tools meant for casual business use, not for production-level content – but because they are part of the standard office package. In other words: they use these sub-par tools because they’re there.

On the other hand, there are built-for-purpose tools that allow for more sophisticated manipulation of content, such as a Component Content Management System (CCMS), a Content Operations Platform (COP), a Help Authoring Tool (HAT), or Learning Management System (LMS). And when there are multiple language variants, add in a Translation Management System (TMS). We should note that these tools have not only a full range of features meant for production efficiency, they usually have some form of workflow function for navigating and documenting the review and approval processes, as well as a robust version control that will satisfy the most stringent regulatory requirements.

We can draw a parallel between COM tools and the tools used by the professionals creating digital assets. For example, a photographer works in Creative Cloud, then moves the assets to the DAM. The text-based content creator works in a COM tool, then moves the content to a repository. Sometimes a DXP will have an authoring tool built in; these tools are generally considered “authoring lite” because of their limitations. (That premise is an entire discussion on its own, but let’s not get distracted.)

Connecting the tools that process content

Here’s a simplified visualisation of the way these tools might connect:

• Work environments. These COM tools (including PIMs) are where the actual management of content happens. These will differ by the type of content being created. Photographers will have photo manipulation software. Videographers will have video manipulation software. Podcasters will have audio manipulation software. Authors will have software that manipulates text-based content in sophisticated ways.

• Enrichment environment. This term is a catch-all term for “things that you do to content to make it better” and covers enriched semantics such as metadata, or content quality optimisation through a quality checker. They are often integrated with the COM tools.

• Processing environment. This is where some sort of compilation or build process takes place. The way this works in text-based content processes is discussed a little later in this article.

• Delivered environment. This is one or more repositories where publication-ready content gets stored. For text-based content, this could be a content delivery platform with specialised functions to boost the effectiveness of the DXP. For digital assets, this could be a DAM, again with specialised functions specific to audio, visual, or video processing.

• Publishing environments. A publishing environment is an umbrella term for one of many environments that pull content on an as-needed basis to publish it for use by a desired audience. These vary according to the complexity of publishing needs. Generally, there will be a DXP that feeds a website. Or an app. Or a voice assistant. Or an artificial intelligence model. Or a wearable. Or a piece of equipment with an LED display. The list goes on, including content pulled by a different system to be incorporated into a larger body of content.

To recap, work environments for different types of content get enriched with metadata. The content gets processed from its draft states and delivered, in publication-ready form, into a repository where the content is stored, ready to be pulled by whatever systems need it.

In this scenario, a DAM is in the category of “Delivered environment”, where publication-ready rich media content, such as videos, audio, photos, and illustrations, await being called by the publishing environment. Some text-based copy does get stored in the DAM; the metadata and asset descriptions are two such examples. (Another potentially distracting topic that really needs its own discussion.) However, this is the exception to the rule when it comes to effectively managing content.

When content needs more than a DAM

So after looking at the different content systems, let’s return to our question. Does it make sense to manage content in a DAM? Let’s break this down by how much demand there is during the management – authoring, editing, reviews and approvals, versioning, and language variants – of content. 

When content professionals – copywriters, content designers, technical communicators, marketing communicators, documentalists, and so on – work on content, the level of content manipulation they need can vary wildly.

A DAM can likely handle the simplest need – content directly related to a digital asset, such as the asset name or a short description. Think of a painting that has a name, a short description, and some metadata attributes – perhaps size, medium used, painting style, artist name, historical period. All of this is stable information.

The next level of manipulation is for long-form content, such as an article for a static publication on a website. The author may split out the title into a separate form field, add a summary for an aggregation page or a Search Engine Results Page (SERP), and some metadata that routes the article into the correct category. This requires basic functionality that could be done inside a DXP, though the interface is likely lacking. And when that happens, authors will find a workaround, usually with whatever tools are at hand. In this case, a DAM is not the logical replacement, as its content “work environment” functions are likely less than that of a DXP.

The next level of manipulation is for content that gets written in components and often has a high re-use ratio. An example would be the content of a company that produces software that gets used by multiple clients or has multiple freemium levels. There could be different names that need to be changed, or different feature sets for different clients, or all sorts of different combinations of content that get produced for users, customer support, YouTube videos, and so on.

This scenario is a very common use case. Content gets created in one of two ways:

• Copy and paste. The author copies the source content in one document, deletes what is not needed, and modifies the text to that particular variant. Each time the content needs updating – anyone who has worked in an environment knows how often the product team will change its mind about product names, feature sets, or other product minutiae on a regular basis – the author opens each of the documents and updates them separately. If a mistake is found partway through the editing process, the author goes back through each document and makes the corrections. (In one project, instructions for three types of credit card machines resulted in a dozen documents, and took weeks to synchronise the content.) As the author works through the content, they have to hold lots of details in their short-term memories, and the likelihood of error is high, especially after an interruption.

• Tranclusions. The other way is to work in a production-grade authoring environment – not word processing tools, spreadsheets, or other tools developed for casual business use – where all common content can be transcluded, or re-used by reference, so that if an error needs correcting, it can be corrected a single time in the source component. Any variant using that source component will automatically have the error corrected. An extension of that principle is that any volatile information, such as product names, feature names, and so on, can be referenced in the source component. As the author assembles the variants – let’s say that the Basic version of a product has a slightly different set of features than the Pro version – they assemble a virtual set of content for each version of the product, with the variables still intact. There’s no need for the author to do anything else except indicate, when they generate the “build” to process all the variants, which variants they want to deliver. Want to deliver both variants at once? Done in minutes. (A recent example is how one government ministry that generates nine different variants of cybersecurity policies in under two minutes.)

There is no DAM on the market, at least to my knowledge, that can rival the basic word processor, let alone handle the type of production efficiency that a COM system is designed to handle. That’s because a DAM’s strength is not semantic structures, editing productivity, or any of the other needs of the editorial process. Add a localisation process to the mix, where content needs to be exported from the content ecosystem into a Translation Management System (TMS), and you’ve now introduced a level of complexity that would require so much custom development that it would make more sense to use purpose-built tools rather than try to adapt a tool never meant for that purpose.

We can take a lesson from DXP systems that wanted to become COM systems: when the features needed by authors are missing, authors will simply revert to the tools with the closest feature set. Recognising that a DAM is a delivery environment, where publication-ready content goes to be stored, we resist the temptation to treat it as a working environment for content. That goes for photo, video, audio, image and, yes, text-based content, too.

Why do product managers need to understand content operations?

Download our 10 essential content operations tips for product managers

I want to learn about the benefits of ConetentOps –>

The post Should you manage content in a DAM appeared first on Content, Seriously Consulting Ltd..

Content management systems actually don’t manage content

Let’s start with the CMS (Content Management System). Here’s a case where the S doesn’t actually M the C. If we look back at the history of these systems, we get a better understanding of how this misnomer came about. In the mid to late 1990s, organisations started building websites. They may have been rudimentary compared to today’s standards – just HTML and some hyperlinks, really. The sites were built based on the assumption that they were static, but of course, it soon became apparent that content continually needed updating: a date here, a typo there, a new piece of information.

I was one of those people who built basic sites for my entrepreneur friends, and the constant stream of phone calls and emails asking for “just a little change” started to get overwhelming, and I started referring site owners to webmasters (remember that job title?) to maintain.

The webmasters had that same annoyance with constant requests for changes, and they started building some basic interfaces that would allow a site owner to make content updates themselves. These interfaces were never meant for actually managing content production; they were meant for managing content publication – simply for putting the finalised copy into place for presentation.

In their defence, the concept of “content management” could seem logical to a profession where content seemed to spring into being, fully formed, ready for publication. But this is the same profession that decided that, in Excel, the opposite of “Hide” was not “Show” but “Unhide”. In other words, semantic accuracy is not the forte of technologists.

Where content management really happens

This set the stage for the next two decades of misnamed content technologies. We already had “word processing” software such as WordPerfect and Microsoft Word. As that class of software got more sophisticated, software such as FrameMaker were more accurately called “documenting processing” software.

This is the real place where content gets managed. Content gets written, re-written, marked up, moved around, moved back, moved back again, commented, edited, and eventually finalised as “good enough”.

The Web became the gateway to componentised content, where HAT (Help Authoring Tool) software let content developers, specifically technical communicators, create content in stand-alone modules (topics), smaller components (references) and even smaller components (variables), and then deliver the content into either online help systems as a series of topics, or aggregate the topics into larger units, usually some sort of manual.

Anyone who understood how a HAT worked soon realised that the software could be used to speed up the creation of any genre of content that traditionally got managed in cumbersome ways. Proposals, reports, resumes – anything that has standard boilerplate combined with changeable chunks of content, and the need to change little bits of content, like changing a company name in each proposal.

Yet, this incredibly powerful tool for all sorts of business content never took off outside of online help. Why? You guessed it, the name “Help Authoring Tool” sealed the fate of this software. The lack of imagination or understanding of software buyers about the potential of the tool solidified the perceptions of its limitations by what it was called. I say this with a bit of smugness, as a recent search on “Proposal Management Tools” turned up a system with functionality and an interface that looks exceedingly similar to a HAT.

The difference between a CMS and a CCMS

Now that a lot of product content is produced in components, the situation becomes even more confusing. The term, CMS, though firmly embedded in the lexicon of content technology, actually does not manage content – at least, not the way that content developers need to be able to manipulate text during the writing process.

I say text because we wouldn’t expect photographers or videographers to manage their content in a CMS; we know that they need software that allows them to do functions that are outside of the capability of the CMS, and then put the publication-ready work into a special CMS, a DAM (Digital Asset Management) system.

What seems to escape CMS vendors is that content developers creating text-based content need special functions, too. That’s why they tend to work outside of the CMS, and reserve the “S that doesn’t M the C” for copy-and-pasting publication-ready content.

Then, a game-changing application came along that really, truly let content developers manage content. Not just manage content, but manage it in highly sophisticated ways. The learning curve was steep, but once content developers learned how to use the software, they realised how accurately, easily, and efficiently they could manage content. Feedback such as “I could never go back to working in Word” and “our team of two is doing the work of six, with time left over” was quite common.

This time, the name of the software is dead accurate: a CCMS (Component Content Management System). Finally, a system for content developers that can actually manage the content components!

When a name limits software adoption

Where this gets interesting is the reaction of stakeholders in the industry when it comes to adopting the software. The name of the software definitely affects the perception and ultimately its adoption. For each audience, the conversations are oddly similar with similarly detrimental outcomes. 

Conversations with content developers go something like this:

• Content person: How can I intersperse universal and product-specific guidance, managed from a single source, and show users only the content that’s relevant from them based on the product they’ve bought?

• Me: Oh, there’s already software out there that can single sourcing and apply semantics for multiple outputs. It can take care of this problem and do more things that you hadn’t realised that needed fixing – like making the source easier to manage.

• Content person: Oh, really? Well, we can’t use that software for what we’re doing. According to the website, that software is to create content that’s different from what we produce.

• Me: Does it really matter what the name of the software is, as long as it has the features that you need to solve your problem? 

• Content person: It looks complicated – we want all the power without the steep learning curve.

• Me: *deep sigh*

The vendors don’t help themselves with how they position their brands. The conversation with vendors goes something like this:

• Me: I have a client that could really use your software to multichannel publish multi-variant content for a bunch of websites.

• Vendor: We do exactly that for technical documentation and product manuals.

• Me: There is no documentation involved, let alone technical manuals. You need to show how you can switch up content in these specific ways and how much more efficient that is.

• Vendor: Sure, your use case sounds completely doable with our software.

• Me: To them it will feel very off-label. How will you get their imaginations to make the leap from technical documentation to marketing content on desktop and mobile sites?

• Vendor: Ummmm, but our demo is set up for technical documentation. We can show them that.

• Me: *deep sigh*

The conversation with the technologists is a little different, but they act as firm gatekeepers

• Me: The missing piece in your content ecosystem is an authoring environment. What you need is some sort of CCMS (Component Content Management System).

• Techie: We have a CMS.

• Me: But you don’t have a CCMS. It’s not the same thing.

• Techie: I told you, we already have a CMS, and if it’s not that, the content people don’t need it. We are the technologists here.

• Me: *deep sigh*

(There is an entire subplot that demonstrates the parallel universes of the Web-based technologies – fill in a few form fields, add some metadata, and you’re good to go – and the XML-based technologies – needs far more planning but has far more power – but that’s a whole other article.)

Software by function or by genre?

Vendors think narrow – they’d probably say focused – thereby locking themselves into a particular content genre. This is a little like marketing Final Cut Pro as film editing software for a single genre – say, only for science fiction or comedies, or saying that Microsoft Word can be used to write articles and poetry but not screenplays.

Because vendors aren’t thinking of all the ways that their technologies could be used, the CCMS has remained largely in the domain of “technical documentation”. Even the term “documentation” rather than “content” signifies an aging way of working: producing manuals, using PDFs, and so on.

The same naming problem applies, perhaps to a lesser extent, for other classes of software. Consider PIM (Product Information Management) systems. They are good at managing data but not as good at managing content. When I asked a vendor how their six-figure software managed multiple products with the same description, the answer was a matter-of-fact “just copy and paste”.

Digital asset management systems, on the other hand, do what they says on the package: they manage objects (in the content world, those would be the equivalent to documents). Translation management systems also do what they say on the label: they manage the translation of content between languages. Again, when the software is named for what it does rather than for which genre, it’s less confusing for customers and easier to market by vendors.

If I were to keep going, I could create a long list – but this is an article, not a book. And speaking of books, a book has been written on the topic of how content management systems have failed authors. Rick Yagodich wrote Author Experience: Bridging the gap between people and technology in content management back in 2014, and not much has changed since then.

The challenge to vendors and adopters

The challenge to would-be adopters of technology is to think outside of the box. Much like we use medication “off-label” – the way that botox was originally used to treat strabismus (crossed eyes) but is now a beauty-industry staple for wrinkle-control, we should be looking at the suitability of the software to solve the problems at hand.

After all, I’ve seen misguided attempts to use Confluence as an authoring system, a subpar Translation Management System as a CCMS, and don’t get me started on off-label uses of GitHub for many things related to content. It’s not that I’m anti-GitHub; I just realise its limitations.

When other technologies could make a difference to the operating models of content production, there is a strange resistance. Actually, perhaps the resistance is not so strange. The reason that GitHub and Confluence get used off-label is because the techies, though they are unfamiliar with how content works, are familiar with those software packages. It’s easy for them to take their limited understanding about content production and match it to features that they think will get the job done.

However, the assumptions about the complexities of content generally fall short of reality. It’s time to involve the content people, to get an understanding of what they really need, and then look further afield for software that actually does what content people need it to do.

The challenge to vendors is to stop limiting their potential market by defining a too-narrow set of content genres that the software can handle. The business benefits are eventually user-facing – better user experience, faster time to market, and so on – but the operational benefits get ignored in the process. The business benefits are hobbled unless the operational side is activated. The same functionality that allows for publishing variants of technical manuals is the same functionality that allows for creating easy variants of training material, marketing proposals, sales contracts, and many other non-technical documentation.

There are single-purpose systems for proposal management, contract management, and so on, but limiting a system for production-grade content to a single genre does not do the software justice. Take a page from conversational clouds. These AI-driven chatbots could apply to multiple industries, and are described for what they can do across many verticals. 

As the complexity for producing and delivering content increases, so does the need for powerful tooling upstream from delivery. It’s time to alleviate the pain of managing content, not just at the delivery end but throughout the content lifecycle. The tools are there, and the market is certainly there; the challenge is getting the offers and benefits to align.

Why do product managers need to understand content operations?

Download the 10 essential content operations tips for product managers

I want to learn about the benefits of ConetentOps –>

• Telegram

• LinkedIn

• Mail

ABOUT THE AUTHOR

Rahel Anne Bailie, Content, Seriously

Rahel is a results-driven, seasoned consultant with extensive experience in digital transformation. She has a strong track record of delivering end-to-end content systems in the context of digital strategy projects, often in environments with complex content delivery requirements. A professional who delivers the hard truths and sometimes difficult prescriptions that help organisations leverage their content as a business asset.

More by Rahel…

The post Why you need an off-label prescription to alleviate the pain of managing your content appeared first on Content, Seriously Consulting Ltd..

Jeff sees a whole generation of work that has over-scoped the execution of personalization. The assumption was thinking that they can personalise all the things that they need to do for all the stuff. They then under-scope or descope the very things that would make their personalisation efforts successful, such as experience debt and information sciences. He posits that figuring out the personalisation problem is actually harder than figuring out the answers. He has developed some simple methods to getting the right personalization outcomes, and it all stems from how to get started. Returning to the Amazon example, the failure is not of personalisation itself, it’s a failure of classification. If you were being retargeted for toilet paper, a consumable, it might be a little more credible as a cadence in your reengagement technique. But a toilet seat is a household durable, not a consumable. So, Jeff observes, this is more of a taxonomy fail combined with an over-narrowing of a result.

Listen to the full discussion to understand the need for progressive personalization

Published: 12 Jan 2022
Duration: 54 min
Guest: Jeff MacIntyre
Host: Rahel Bailie

Listen to the full discussion –>

My own experience shows just how long we have to go before this type of personalisation really works. Clicking on an ad for clothing in the wrong size range doesn’t stop the ad from being shown to me ad nauseum. It was also bemusing to me to be targeted with products for middle-class African women for at least six months. It could have been mobile ad targeting – my mobile phone is often in close proximity to the mobile phone of my (Afro-EU) partner – but when I got the African-print head wraps (I have baby-fine hair that lays as flat as a pancake), the ads were simply casting a wide net and hoping for the best, but with little effectiveness.

Calling out the personalisation gap

This is called overfitting, according to Jeff, who notes that this is a race to see what the easiest way is to target a user. However, it breeds mistrust and, ultimately, failure. Instead, designing the personalisation to look at intention during looking versus buying, say, a head wrap, could back off the retargeting, and then re-engage you. It’s designing with data, which tends to get overlooked as part of the strategy.

I asked Jeff about the differences in how marketing teams and product teams considered personalisation. There is a difference between pre-sales and post-sales personalisation efforts. However, there are much more dramatic differences by business sector. Ecommerce is 10 to 15 years ahead of many other sectors in terms of understanding techniques, practices, data hygiene, and data quality. Even then, you need to consider whether you are operating in a sales context, a marketing context, or employee engagement context. These things are very determinant of what success looks like and the techniques you would undertake.

Organisations often try to personalise, whether that be ad retargeting, doing omnichannel delivery for different stages of a customer journey, or localising product content for different markets, by grasping at straws and trying things out without internal expertise. Jeff calls this the personalisation gap.

IA before AI

The personalisation gap is not about the technology; it’s about the expertise – a lack of experience in personalisation activities, and lack of subject matter experts. Take, for example, not addressing long-standing metadata debt in the organisation. Instead, there becomes a rather treacherous combination of overthinking the solution by deciding to go some big-bang route using artificial intelligence (AI). But having poor vocabularies unconnected to different systems of record doesn’t help reconcile the shortcomings of the current state of your Information Architecture (IA). Jeff’s mantra of “IA before AI” is because those systems must be connected at a certain level – that’s why it’s called connected experience. That means having multiple systems of record being able to string a single customer experience across them, which means being data-driven.

The idea that an organisation can deprioritise their taxonomy work and only use machine learning and AI on data an unstructured content is folly. I’m using the term in the context of British architecture, where some wealthy person would construct a building for decoration, with elaborate classical pretensions. They might go picnic at the folly, but otherwise, it was a quite useless construct. So we have these follies in personalisation where organisations are chasing the shiny things and ignore building a solid foundation. Follies won’t result in systems all magically working together. There needs to be a solid foundation laid before anyone builds on it, and taxonomy is part of that foundation.

Three on-ramps to get to personalisation

One of Jeff’s contributions to the profession is the creation of three different methods for putting effective personalisation into play. These on-ramps are meant to be contained explorations or pilots that will help validate the ROI, the business benefit, and the user benefit, before tackling things such as normalising all of the taxonomies or connecting all the subsystems.

Discussing personalisation as connected experience brings together the various ways that people talk about personalisation, which means different things to different people. For some people, talking about personalisation actually means automation. In other cases, people are thinking about altering the user experience, or it could mean customisation of content delivery. There is an on ramp for each one of those situations, to which Jeff has given musical terms. These aren’t meant to be sequenced but used as stand-alone methods. Different organisations will adopt different methods, depending on their maturity level at any given time.

• The Solo – automation of some of the customer journey, creating a flow that leads to tangible benefits.

• The Sonata – customisation through audience segmentation and analytics, to deliver custom views of content for particular audiences.

• Symphony – classic personalisation in the form of content or product recommendations.

Jeff discusses these three on-ramps in detail on the webinar, where you can also find some slides about each of the on-ramps.

Connecting personalisation to content operations

Jeff emphasizes that the best outcomes from using these on-ramps is shining a light on content, assets, and elements, and showing their contribution to advancing a user. It has to remedy a user need more promptly or driving lift on conversion, so a business benefit is extremely important in for us to get an idea of content performance. In the end, you need to have a performance attribution revenue attribution element to any of the on-ramps.

The old adage of “you do not manage what you do not measure” is key when implementing personalisation. Measurement could be through Return on Investment (ROI) or Internal Rate of Return (IRR). At the risk of being overly reductionist, ROI is spending to earn money, while IRR is spending to save money. To measure well, this means building up operational efficiencies. This fits more in the IRR category, but organisations have traditionally been reluctant to think about improvement to business processes. Yet to measure the effectiveness of personalisation efforts, an organisation needs to consider organisational change for operational efficiencies.

For the full benefit of the explanations of the three on-ramps, elaboration on the benefit of operational improvements, you’re going to want to listen to the webinar. You can find it on the Let’s Talk ContentOps show, Simplifying Personalization, where it is available as an on-demand webinar.

Why do product managers need to understand content operations?

Download our 10 essential content operations tips for product managers

I want to learn about the benefits of ConetentOps –>

Learn more about personalisation:

• Listen to my conversation with Jeff McIntyre

• Learn about the first principals to ContentOperations

• Learn the basics of Content Operations

• Telegram

• LinkedIn

• Mail

ABOUT THE AUTHOR

Rahel Anne Bailie, Content, Seriously

Rahel is a results-driven, seasoned consultant with extensive experience in digital transformation. She has a strong track record of delivering end-to-end content systems in the context of digital strategy projects, often in environments with complex content delivery requirements. A professional who delivers the hard truths and sometimes difficult prescriptions that help organisations leverage their content as a business asset.

More by Rahel…

The post Personalisation as a Connected Experience: An Interview with Jeff MacIntyre appeared first on Content, Seriously Consulting Ltd..

Standardised method of inquiry

The Five Pillars framework adapts basic management consulting processes to content turn-arounds. Familiar processes create as much confidence as the outcomes, which in turn, helps with buy-in from management. The vast majority of organisations – and, dare I say, the management consultants sent in to fix a content problem – don’t understand the inner workings of content, the pain points that content producers face, the efficiencies that can be garnered, or how to use content to further business goals. The Five Pillars approach is aimed at remediating that basic flaw of other inquiry processes.

The five pillars of content strategy

The Five Pillars framework is a cornerstone of the inquiry process, and helps shape the outcome in an efficient and effective way. The goal is to gather enough information about how the organisation currently works and their aspirations for a future state, and be able to reframe that information in a way that can be fed back to the client for validation, and suitable recommendations in each of the areas. The outcome is generally a summary that organises the information into logical buckets – in other words, the five pillars – so that management gets a clear picture of what’s happening in their organisation, with a series of recommendations.

The four phases of the content strategy process

The overall process consists of four phases: discovery, gap analysis, and recommendations, and a roadmap.

Discovery

During the discovery phase, determine what business problem the organisation is trying to solve, or what business goal they aspire to achieve. Using the Five Pillars framework helps you ask a range of questions to tease out hidden needs, and extrapolate the salient points. The outcome is information that frames the questions for the next phase: the gap analysis.

Gap analysis

During the gap analysis, you will determine what the organisation is doing now – the current state – and what business value the organisation wants their content to deliver – the future state – and what the gap is between these states. That will tell you what’s missing, and what they need to succeed.

The questions that get asked during discovery and the gap analysis can range wildly. Using the Five Pillars framework will help focus your questions. This approach gives you a way to structure your interviews, and to synthesise the results in a way that addresses the five umbrella areas of any organisation.

By the time you’ve developed the strategy, you’ll want to have asked questions about each of the five pillars. This will help you understand:

• the business problem

• the current and future states, so that you can calculate the gap

• conditions that might affect how you put together the roadmap, such as budget, governance, or infrastructure constraints.

Recommendations and roadmap

From the discovery and gap analysis, there should be enough evidence gathered for you to pull together a series of recommendations for how to move forward, and for a high-level roadmap to get the organisation to where they want to be. The roadmap is the synthesis of the work done in previous phases. There are no cookie-cutter businesses, and no cookie-cutter roadmaps, and each set of recommendations is highly situational. The roadmap will be an extension of the recommendations, and provide the organisation with a way forward. The roadmap should show how the strategy addresses each of the recommendations from the gap analysis, what steps the organisation needs to take, and in which order.

Using the five pillars framework

The five pillars framework is an overview, a tool to add to your toolkit for the part of the discovery phase and gap analysis related to the questions posed during interviews. Within the activities applicable to each pillar, you will most certainly employ the familiar techniques of conducting a content inventory and audit, carrying out a qualitative analysis, looking at personas, analysing metadata, and so on.

The questions included in the download are generic and meant as a starting point for deeper investigation. Your questions will certainly be deeper and more focused in each of the areas. The framework helps you ensure that the questions you ask of an organisation are balanced across the five areas and the four stages.

However, data does not tell the whole story; presenting that data doesn’t mean that the audience will necessarily understand what the data means. The role of context is provided by content. In a 2018 study discussed in Science Daily, between half and three-quarters of the study group had trouble interpreting statistical data, particularly when presented as probabilities instead of natural frequencies. Content fills the gap by adding context. That context completes the story by filling in the gaps of “what does that data actually mean?” to the audience segment who are not data scientists or statistics enthusiasts or not experts in a particular field.

Frontiers. “Why don’t we understand statistics? Fixed mindsets may be to blame.” ScienceDaily. ScienceDaily, 12 October 2018.

Data tells only part of the story

For all the importance of content to be delivered along with data, content has been largely ignored in the data arena. The manipulation of content to turn copy into meaningful context bears little resemblance to the processes for bringing meaning to data. Editing data means changing and possibly normalising a data point in a database cell. Editing content means checking the accuracy, the consistency of the language, the spelling and grammar, and most importantly, the context. Sandwiching a sentence between two other sentences is not a neutral act; it can enhance the telling of the data story or backfire terribly and not only detract from the story but offend your audiences in the process.

Content and data are maintained in very different ways

Content is a very separate discipline, and although the automated delivery of content dovetails with the need for automated data delivery, that’s there the similarity ends. Yet there is a very strong temptation to treat content like data, confining content to cells in a database, moving it around in static chunks like so much boxed cargo. The mechanisms meant for processing data limit content in so many ways. The complexity and nuance are dampened; the contexts are limited; its potential is hobbled. The editing process becomes cumbersome and error-prone; content bloat occurs as copies are pasted into multiple database cells; and the overhead of content maintenance becomes unwieldy. To allow content to operate at its full potential, it needs to use its own standards and its own semantics, which ultimately enables its ability to interoperate.

How we produce content affects how much value it accrues

The end-to-end method of producing content that can co-exist alongside data is best described by breaking down the production process into three distinct areas.

• The first area is the authoring and editing environment. This is where authors can manipulate content by combining content components into larger content objects; these objects use one of the content standards appropriate for authoring systems. These systems allow authors to transclude content fragments for use across many content objects, do batch checking for many levels of content quality, and add semantics to content objects for outcomes such as personalisation, aggregation by keywords, or automated delivery and presentation by a CMS.

• The second area is a processing environment. Once the content is considered final, the authoring and editing environment sends the content to be processed. Processing is similar to doing a software build; the processing builds publication-ready strings, resolving all of the transclusions and cross-references, and auto-application of more semantics. The build also transforms the content into whichever formats are needed by downstream systems. That could mean HTML for a website, specialised XML for an enterprise system, strings destined for a bot or AI model, or even a PDF destined for a print publication.

• The third area is a delivery environment. The “built” content is then treated as content objects that are ready for use by other systems. The content is held “on offer” in a repository or database, and is pulled by other management systems, as needed. This is where the content is delivered to, though it is not the final destination. It is at this point that would be appropriate for a database to pull content, for use alongside data. This ensures that a database stores only publication-ready strings, and allows authors to retain the powerful features needed to efficiently produce the needed content.

Intelligent content is valuable content

These content principles are known in the content industry as “intelligent content”, which states that content must be structurally rich and semantically categorised, to make content automatically discoverable, re-usable, reconfigurable, and adaptive. The demand for intelligent content is being driven by the offer of Content as a Service. This has opened the door for both content standards and applied semantics as ways of automatically delivering content alongside data for more targeted contexts.

Related:

To learn more about this topic, you can watch my presentation at ENDORSE.

The Quest for Content in Context: Using Standards and Semantics for Interoperability

ENDORSE: The European Data Conference on Reference Data and Semantics

Information operations in a digital environment

Content production is evolving, and with each change, we look for new ways to represent those changes. Despite not liking the term quad for its lack of semantics, I realise the usefulness of the quad style of infographic. It’s easy to visualise information in groups of four, or divisions by four. In that spirit, I’m introducing a new quad for content.

The need for a new quad

If there’s a common theme among clients or employers, it’s an overwhelming lack of understanding of how complex it is to produce content. They don’t understand the strategy aspect, the operational aspect, or the potential that they’re missing out on. They think they understand what it takes to write and edit content, but get into the topic a little deeper, and they don’t know anything about the complexity. And don’t get me started on the lack of knowledge about localisation. Or personalisation. Or operations.

There’s an assumption that “writing and editing” is the sum of content production. The implicit end of that sentence is “unlike producing code, which is complicated and makes the magic happen”. Pffft. The nuanced complexity of content leaves the complications of code in the dust.

Trying to show the depth of content and the connection between strategy, editorial, operations, and growth took a lot of thought and fiddling around with graphic representations. For better or worse, I decided that a quad in the form of a Venn diagram overlaid by a diamond worked best.

A quad for all seasons

OK, a bad pun on A Man for All Seasons. Notice that I’m calling this a content quad. It’s not a content strategy quad – content strategy is only one aspect of the graphic. In this quad, we consider the entire content ecosystem. Content is so much more than what is represented in any single quadrant. It’s not just web content or just marketing content or just editorial or just…well, you get the picture.

Interpreting the infographic

A good start is with the centre diamond. The four outcomes are general, not intended to be attached to any one of the quadrants. Having said that, the position of each of the outcomes has more relevance to the nearest quadrant.

The strategy

The strategy aspect is exactly that – planning out what to do to make the content ecosystem work. The strategy is not the implementation nor is it the ongoing doing of the work. A content strategy in itself is only useful if the strategy gets implemented. Commissioning a content strategy only to put it into a drawer is not helpful. In fact, it’s more frustrating for the content professionals it’s supposed to help because you’ve held out hope and then watched it die through management inertia.

If a strategy is a plan, then what are you planning for? The outcome of the strategy is some sort of operational model for content. Another way of putting this is that you are turning content chaos into a functioning system. The idea is to support the production of content, whether that is through better editorial processes that increase content quality, or through better technical underpinnings to optimise production. Layered over this is governance, which affects both the editorial and technical sides, and other things such as change management, training, and standards.

Operations

To clarify, strategy is planning for how your copy will get created and managed. Content operations is the manifestation of the strategy. The Operational Leaders podcast host, Terrance O’Malley, refers to this as “the business that runs the business”. No matter what industry vertical your organisation is in, there are operational models in place to make the business run smoothly. Finance has an operational model for accounting; HR has an operational model for recruitment; Administration has an operational model that controls inventory, maintenance, and other things that make a workplace run smoothly. The content operations model gives you a system to be able to create content, day in and day out, in an efficient way. One example is content modelling. It is part of the strategy, as content modelling helps set up the infrastructure. Using the content model to create content on an ongoing basis is part of operations.

Editorial

At the top is the editorial quadrant. This is the area that most content producers are familiar with. Either we have started out as a content developer of content genre or another – anything from business communications, marketing communications, technical communications, ecommerce, journalism, or even fiction – or perhaps we’ve come to our jobs through translation and localisation, or some production control such as supply chain management. No matter how we came to the industry, we recognise that at the core of what we do is good copy. It’s not enough to produce bad content more efficiently. The consumers of our content will judge us on the quality of what we’re serving up for consumption. Those consumers are largely potential or existing customers. The forgotten consumer segment is often those who need good content the most – the developers and technical audiences who are dependent upon clear, precise instructions to set up and configure equipment, implement APIs, and administer the systems that support operations. The impetus to create good content is the same, no matter who the audience. If your message cannot be understood, all the strategy and operational support you can throw at it won’t help. Investing in good quality copy is the biggest service you can provide to your end users.

Enablement

The final quadrant is labelled Enablement, as information enablement is the key to adding value, whether that is enabling publishing at scale, producing at pace, managing risk, expanding scope, or enabling personalisation. Enabling growth and personalisation are the two most-discussed aspects of content at the moment. Personalisation, in particular, is today’s holy grail in terms of delivering value of a product or service. By its very nature, personalisation and scale go hand in hand. Personalisation for multiple audiences means multiple content variants. Multiple audiences, industries, locations, languages means a steep increase in content volume, and that means more operational horsepower is needed to manage it. There is no single way to manage that type of content, just as there is no single way that a store arranges its merchandise. But you can get that a store of any substantial size has an inventory management system to keep track of stock. Supporting growth is an integral part of enhancing value of the products and services provided by our organisations.

The intersections

Of course, each quadrant does not exist in a silo. They overlap in multiple ways, some of which are impossible to show in a 2D graphic. However, I’ve called out the strongest intersections between adjacent quadrants. Finding your own (dare I say personalised?) intersections may spark some aha moments specific to your organisation. You may have NLP as a factor, or machine learning, or chatbots. You may need to support an artificial intelligence model, augmented reality, or otherwise connected products. Any of these may spark additional ideas.

Neither beginning nor end

This is by no means the first content-related quad that’s come along, and it certainly won’t be the last. I’ve been in the business of producing content since 1989, and have watched the evolution of content production change at a steady pace. Each time it seems that there may be a plateau, technologies change, and with it, the complexities of producing content. Even the things that seem constant – principles of comprehension, reading patterns, attention span, cognitive loads, learning methods – have not remained static. This decade, we’re poised to take the next step up along our journey toward industry maturity.

A couple of theories have bounced around about why content operations, ContentOps for short, is gaining traction. One of those theories is concept formation – that is, when people learn to sort specific experiences into general rules or classes. Because concepts such as DevOps and DesignOps have become familiar terms, it’s a short mental stretch to consider what ContentOps might mean. Another theory is that the need for operational models has grown along with the size of our content corpora. It’s when enough organisations feel the pain of trying to manage a large body of content that they start to look for ways to operationalise their processes.

For many of the practitioners among us, the last ten years has been focused on content strategy. But a content strategy is worthless if an organisation says thank-you-very-much and pops the carefully crafted, elegant strategy into a drawer to gather dust. It’s worth clarifying some of the concepts and reframing them to make sense of content operations.

Content strategy vs content operations

Content strategy is the formation of a plan that codifies a repeatable system that governs the management of content throughout the entire content lifecycle. In other words, why are we creating a content strategy? To operationalise content processes. Some strategists operationalise social media posts; some operationalise marketing material; some operationalise technical documentation; others operationalise product content.

If there is no implementation phase, then the strategy is just that: an unimplemented plan. The content strategy needs to connect to content operations, whether that’s planning for more efficiency on the marketing or editorial side, or it’s designing a technical content ecosystem to increase delivery efficiencies. Sometimes, it’s all of the above. In organisations that can squander time and effort – 13 people in 6 departments over 3 months in a recent benchmarking exercise – content operations is not a luxury; it’s critical to a competitive edge.

Precursor to content operations

It’s worth mentioning here that content operations has been practiced by technical communication communities for decades – they didn’t call it by that name. It was usually called “survival in the face of more content with less budget”. Even before proper systems for managing content existed, operational models were being developed by organisations with large content corpora. At one multinational company, we created a bank of “chapters” that were stored centrally, and drawn out to be mixed-and-matched according to the size and complexity of the product. Of course, version control was rudimentary, and editing was still done manually across the 80+ writers to create a single voice.

Early content operations

In the 1990s, I had my first aha moment while listening to a practitioner discuss how they managed training content for a suite of enterprise account software. They created a series of training modules, one per product: one for accounts receivable, one for accounts payable, and so on. They set up parallel structures for other language translations, such as French, German, and several other languages. When a client bought a series of products, the trainers structured a custom training package according to what they’d bought, in English, and then specified the output language, and the training material was delivered electronically to a local shop for printing. Pretty heady stuff for the late 1990s!

Now, we have our choice of content management systems into which we can paste publication-ready copy to be maintained. Or better yet, our publication-ready content can be sent via API into the Web CMS, for even more operational efficiency. The technology stack looks impressive. (Take a look at the Semantic Web Stack on Wikipedia for an example.) But a web CMS only the tip of the iceberg when it comes to how we actually manage content production.

ContentOps for technical content

In very large organisations, the technical communication groups may have implemented some form of ContentOps. It makes sense, as the function of technical communication often involves managing massive amounts of content with astounding amounts of complexity. They’ve taken on ContentOps out of sheer survival. In fact, there is a subdomain called DocOps that has a strong following in this area, of super-efficient processing of technical documentation.

With the appearance of generative AI, the potential complexity of content delivery has increased. Content delivery platforms can ingest and enrich content, and practitioners have become familiar with concepts such as enriched semantics and Graph RAG. Best practices for content operations are starting to emerge - and what was old is new again, as the concepts of intelligent content, popularised by Ann Rockley in 2015, are more important than ever.

ContentOps for marketing content

These efficiencies tend not to have migrated yet to the marketing side of the house, where a typical technology stack – and I use the word in a very tongue-in-cheek way – involves Google Docs or Microsoft Word, email for sign-offs, sticky notes on a computer monitor, and a spreadsheet to track it all. And we are so used to being system-poor and process-deficient that no one in our organisations are even recognising the need for proper processes to be handled within a robust system to support content operations. Larger organisations have adopted creative operations, defined by Atlassian as “a system used to streamline creative workflows, improve collaboration, and optimise team output”. The systems used to manage creative ops are focused on workflow of digital assets, however, rather than on managing content as components for assembly and re-use.

ContentOps for all

This is where opportunity comes in, to leverage the potential for content operations at a very fundamental level. Now that organisations recognise the need for an operational model for content, there’s also the potential to pool resources to find ways to work smarter instead of harder. Ventilating the silos between marcom and techcomm for their mutual benefit leverages the potential for efficiency and effectiveness for organisations as a whole. And when that connection happens, it’s a beautiful thing.

Since the original post, some five years ago, the industry has undergone massive changes. Generative AI has had particular impact in the content industry. The disruption has been profound. The complexity of operational efficiencies have multiplied, making the techniques of the previous decade look like child’s play in comparison.

What and why of information operations

Most of what we do as practitioners is in response to a business problem, or perhaps a business aspiration. In the case of information operations, it is often a double-barrelled approach to achieving a particular outcome. These outcomes can be stated in several ways; not every organisation may choose to go down the same path.

Some ways of expressing the what and why:

• We want to get information out there faster, so we need to automate aspects of content production.

• Our data is changing more often than ever, and the team managing the data (sometimes still in spreadsheets) can’t keep up with the multitude of places that the data has to be kept current.

• We need to commoditise corporate knowledge – perhaps with real-time publication – so that we have less of a bottleneck in our call centres, providing market-appropriate, personalised information to call centre agents.

• We don’t have enough resources to handle all the manual production steps, so we want to build more efficiencly into production.

• We want to deliver a personalised customer experience, which means creating more specific content, marrying it with specific data, and delivering it to the right audience segment at the right time.

• We want to stop entering data points into our content manually when we know there’s a way to automate that function.

• We want to democratise our knowledge by making it available across our client base on a self-serve basis.

The common themes here are efficiency on the creation side, and automation on the delivery side.

For some organisations, the impetus may be to provide a better user experience. For other organisations, the impetus may be to control internal costs. Either way, success depends on the working on both sides, and meeting in the middle.

Building on the DIWK framework of knowledge management

Looking at a traditional knowledge management pyramid, the base is data, building up to information, and resulting in knowledge and wisdom. While this is a useful model within knowledge management to understand the cognitive transformation of raw material (data) into meaningful conclusions, the model needs to be updated to be helpful in explaining the mechanics of delivering information.

To understand the operational side of DWIK, we need to stand the pyramid on its head. First of all, content is a missing element from the entire equation. It may be co-opted into the category of data, but content is inherently different than data and needs its own category. And secondly, in information operations, we cannot strive to create wisdom. Information can be combined and delivered in ways that create a form of knowledge; it’s up to the consumer of that information to contextualise their findings into something more.

The information operations pyramid

The word “operations” is the key to the interpretation of the practice. The explanation of information operations is firmly rooted in the production and delivery sides of the equation, as an operational model that leads to a range of potential outcomes that are varied as the organisations that incorporate the practice.

Let’s look at each layer of the inverted pyramid separately.

• Data. For InfoOps to work, data needs to operationalised. For purposes of explaining the need behind information operations, the data sets must be maintained in a usable state to be available on demand, to enable dynamic delivery of data. The data can be used to create analyses, data visualisations, or simply to populate data points within content objects. An simple example is “Our company was founded in [year] and has [number] employees. Revenues have grown at a steady rate over the past [number] of years [trend over time graph].” Those data points can be dynamically updated on an ongoing basis, with no intervention from the content developer. There is a lot more to data operations that indirectly benefit information operations, which you can have explained by people much closer to data than me.
  Data operations

• Content. For InfoOps to work, content needs to be operationalised. Operationalising content is very different from operationalising data. What differentiates content from data is context. Content is human-usable, contextualised data. My usual example is to tell me whether the data point “12” is good or bad. You don’t have enough context to make an informed decision. If it’s a £12 pencil, it’s probably not a good investment; if it’s £12 for a case of pencils, you’re probably getting a better deal. If 12 is a dozen, then you might think eggs, while a baker’s dozen is actually 13. If I tell you that 12 = month, then generally, 12 equals December. There are some explanations about content operations freely available, including this one.
  Content Operations

• Information. If content is contextualised data, then you can consider information to be contextualised content. Unfortunately the term “information operations” seems to have been adopted in the USA to mean “support for military operations”. Outside of that very specific definition, operationalising information can take many forms. Here are two examples.

  • In the first example, we look at embedding data parameters into a piece of content, as follows. “How many countries are in the world? The UN recognises [number] countries and territories. The USA officially recognizes fewer than [number] nations. Ultimately, the most recognised answer is that there are [number] countries in the world.” The numbers on their own would be meaningless without the context provided by the text surrounding them. Automating the insertion of data into the content in real time allows us to update the data points from time to time without intervention by the writer.

  • In the second example, the data does the heavy lifting. An investment app I use shows me an image with a dynamically-generated, up-to-the-minute analysis of a financial investment, supplemented by enough content to explain what the chart means. This is to allow users to understand the ROI they are getting on their investments. The data is most important, with content playing a supporting role.

You can read about commoditising information in a few places, though one of the most accessible articles I found insists on calling content “data” all the way through it.

Commoditisation of Information

Output of information operations

Information is the input; knowledge is the outcome. It is fitting that knowledge is at the top of my InfoOps pyramid, as the combined outcome of a number of content and data inputs can result in many different combination of results. In the earlier example of 12 = December, a European resident may conclude that it might be a nice time to travel in Thailand, whereas Thai business owners may conclude that they need to know how to brave an onslaught of European tourists or stock up on merchandise for their souvenir shops.

When it comes to enhancing a user experience with that information, some of the process can be automated – recommendation engines deliver different information to different audiences all the time, sometimes with hilariously inappropriate results – and other aspects require the recipient to consider their own contexts. We can provide information, but the rest is up to the information consumers to draw their own conclusions, and make knowledge out of the information.

Benefitting from information operations

Striving for operational efficiencies has increased in complexity of execution. The addition of AI into the mix has introduced great opportunity - but also with the corollary risk and responsibility. In the vein of “it takes a village”, in this case, it takes a content ecosystem to make your information operations happen. And that means developing a strong strategy, in conjunction with the data, knowledge management, and AI teams to implement that strategy that works within the larger ecosystem.

================

If your organisation is interested in what benefits a strong operational strategy can bring to your processes, feel free to get in touch.