In performance support, clarity of the support isn’t a luxury — it’s essential. It’s the difference between achieving accurate, successful application of learning or producing mistake-ridden performance that no-one wants or needs.
That’s why early pioneers in the performance support field, like Conrad Gottfredson and Bob Mosher, paid a lot of attention to the importance of structured writing and layout. They knew that consistency in how information is structured and sequenced matters as much as what it says.
In their book Innovative Performance Support, they spent almost eight pages describing how the layout, chunking, and consistency of support materials could make or break usability.
Back then, that discipline had to be learned and applied manually. Every header, bullet and indent was a deliberate design choice — crafted so users could find what they needed quickly and act on it immediately.
Today, (in the context of AI, at least), this aspect of designing performance support is becoming much easier. All that careful thinking about consistency of structure still matters just as much as ever; but the manual work of creating it can now be largely automated. That’s where something called Markdown comes in.
Markdown: A Simple Language for Structured Clarity
Markdown isn’t coding in the traditional sense; it’s actually a plain-text shorthand for expressing structure and layout.
For example, a single “#” makes a heading. Two asterisks make something bold. Lists, sub-lists, and call-outs can all be expressed with simple symbols.
At first glance, it all looks a bit too simple. Surely a bunch of hash tags and asterisks can’t be that useful or important? Well, actually, they can. The simplicity is the point.
Markdown defines something really important when it comes to providing lots of information: an overall content hierarchy, relationships between the levels of the hierarchy and ways to emphasise key pieces of information within that hierarchy.
From a Wall of Words to Structured Chunks
And if you’ve ever opened a document, an intranet page or a chatbot reply that looked like a wall of words, you’ve seen what happens when a content hierarchy (with its inherent structure) is missing.
People mis-read or skim through without really paying full attention or just give up out of frustration.
Applying a structured hierarchy to that wall of words avoids those reactions. A rambling piece of text is broken down into smaller, logically connected units of information.
That level of intentional layout has always been essential. What’s changed now is how Markdown makes it so easy to achieve that.
PerformaGo and Built-In Design Logic
Because with PerformaGo, Markdown is built into the back end. Invisible to you when you are creating a custom GPT. Invisible to the users of your GPT.
However, when the custom GPT you have built for your learners produces a response, Markdown works in the background to ensure the output is properly structured — headings, lists, emphasis, and layout all automatically applied. That means your learners get the benefits of structured writing all the time, every time.
It’s the digital equivalent of having a technical writer sitting quietly in the background, making sure every instruction, step, and example is consistently structured and organised.
And because PerformaGo also draws on the principles of Information Mapping, those Markdown structures align with a deeper information design principles — chunking by purpose, grouping by action, sequencing by need.
In the next Making Learning Stick post, I’ll explain more about what Information Mapping is; how it works; and why Markdown and Mapping are most definitely a match made in heaven.
