The way users want to access information may be undergoing a seismic shift. Some have suggested technical communication is rapidly moving towards a question-and-answer (Q&A) paradigm.
As user behaviour moves away from reading and towards asking, we discuss how chatbots, AI, and Q&A-driven interfaces are reshaping the expectations and responsibilities of technical communicators.
Takeaway quote:
“It’s not about crafting a list of 20 questions. It’s about building a robust knowledge system that can handle any question and deliver meaningful answers.”
Resources
- Michael Andrews and Ardis Ramey on Q&A communication paradigms
- UK Government Digital Service post: FAQs: why we don’t have them
- Statistics on voice and search behaviour
Transcript
This is the Cherryleaf podcast.
Hello. My name is Ellis Pratt. I’m one of the directors at Cherryleaf. We are a technical writing services and training company in the UK.
And today’s episode on a rather cool June morning … makes a pleasant change from some of the hot weather that we’ve had, and that’s predicted to be here …. is, are we moving to a question-and-answer world?
And this has been prompted inspired by a post by Michael Andrews on LinkedIn and a follow up by Ardis Romney with her thoughts.
It was an interesting topic, and we have done a blog post on this documenting our thoughts.
I thought it would be a good topic for the podcast and also, we can combine it with talking about the results from our recent survey into the use of AI in technical communication
See whether the trends that were reported in that indicate whether this is happening, this trend, if it happens towards the Q and A world is happening today, whether it might happen in the future.
Technical communication has for many years been rooted in instructional information instructional guides, resources where users could go when they get stuck when they need information on how to do something or how to resolve a problem.
And one of the challenges lots of discussions in technical communication is how can we make that information available to users. How can they find it when they need it. And the model that we have is essentially that. users search for what they need, they find it and then they read it.
And on the provider side the writer side the goal has been to make the content easy to find as well as easy to understand and easy to act upon that it’s actionable.
And all of that still has its place.
But it may be that the way that users want to get to that information access that information could be undergoing a significant and seismic shift.
And on LinkedIn, Michael Andrews suggested that technical communication is moving towards a new paradigm. excuse me for using the word paradigm, a new world that’s a question and answer-based world.
And Ardis Ramey added her comments to Michael’s post. According to Andrews and Ramey chatbots are shifting the emphasis from reading to asking.
The argument is that reading requires effort time focus and often some previous knowledge about the topic
Asking, in comparison, is fast. It’s intuitive it’s accessible. And in this new model users don’t need to know where the answer lives. They just need to know what they want to ask. Let me quote from Michael Andrews post.
Online content testing centred on this user as to whether users noticed information and messages and how they understood them ebooks might tally which passages were most frequently highlighted Already a range of apps are available that allow users to load URLs or PDFs of content and ask questions of the material.
These tools support a more active relationship with the topics than the passive reading offered by static pages They allow users to look at content more broadly and selectively.
With chatbots the reader’s agenda becomes more important than the authors. We can gather insights into what readers are curious about and how similar those interests are to what the authors have written. It’s not just whether the content covers an issue but how the content addresses it. The value of content will increasingly be defined by how effectively it shows its relevance to users chatbot conversations.
And Ardis Ramey in her reply said:
Users preferring a questioning interface over a reading one isn’t the end of traditional technical writing. Instead we should see it as a challenge to our technical writing traditions If we’re writing to enable users and reduce friction between the audience and their goals then incorporating a chatbot front door is one step towards engineering documentation that truly achieves that goal It’s beyond time for digital publication to step away from the traditions of print media and embrace all of the superpowers of publishing in n-dimensional digital space.
In the show notes, we’ll provide links to Michael’s post and to Ardis’s post so you can read them in full.
Let’s start to reflect on this by asking why the change?
For years users’ behaviour with computers with information has been conditioned by their interactions with search engines like Google. And when people use search engines asking questions is one of the most common ways in which they seek information.
And the statistics provide evidence of that.
There’s an article on SEOCopilot about some research by Surefire Search. And according to that research around twenty seven percent of search engine queries are posed as questions.
And among these question formats, how is the most commonly used one at thirty eight percent of questions followed by y which is twenty four percent.
In addition, the study found that half of all queries are now four words or more in length indicating that users are becoming more specific and detailed in their searches.
In addition to this, voice searching tends to be more conversational. And when people use a service like Siri Their interactions are more likely to be phrased. as questions. And the surveys suggest around sixty five percent of voice searches are in a question format.
And now we have chatbots where the promise is that users can get quick direct answers that the chat bots can understand the context of the question far better than a static document which means they’re able to give more tailored personalised responses and that chat bots can reduce the cognitive load that sifting through documentation is work where getting an answer from a chatbot directly just feels effortless
A question-and-answer environment like this doesn’t mean that we’re talking about static traditional FAQs that you have seen for many years on websites. As the government digital service along with others have pointed out. traditional FAQs rarely work.
Often, they’re a dumping ground for content that doesn’t fit anywhere A miscellaneous section on a website or a knowledge base. They tend to reflect organizational. assumptions about what users ask rather than actually reflecting what users need. They’re often a source of content duplication and a headache to maintain They’re often a poor user experience because they force users to scan a long list of questions that might or might not be relevant to their specific situation and problem. And the government digital service and others argue that if information is important enough to be in an FAQ it should be findable and well-integrated into the main content where users naturally look.
So, a Q and A world shouldn’t be a world of static FAQs
Instead of directing users to a fixed list every part of the knowledge base becomes a source of answers The FAQ is everywhere and nowhere specific. So, it doesn’t mean users will be navigating to a particular list Instead the idea is they ask other questions. via a chatbot or an intelligent search bar.
The big question, what does this mean for technical communicators? We don’t think this shift is a death knell for technical authors or technical writers as the profession is called in the USA.
Because chatbots and other AI systems rely on a body of knowledge and a body of knowledge with good quality information
It’s more likely to provide good quality answers than one where the chatbot is relying on woolly incomplete contradictory information.
So, this is more about a refocusing of core skills a subtle change.
Long form guides and help systems are likely to continue to be needed They become the foundational content for these question-and-answer experiences
But here are some ways in which the role of a technical writer may need to adapt.
- A Q&A approach focuses on the direct path from question to isolated answer. This doesn’t mean abandoning structure, but rather ensuring each part can be effectively decontextualised and re-contextualised by an AI.
- Content must be granular, capable of standing alone as a coherent answer to a specific question.
- Clear headings, structured sections, and rich metadata (schemas, taxonomies, and consistent terminology) help AI understand, index, and retrieve information accurately.
- Ambiguity is the enemy of AI comprehension. Our writing must be precise and unambiguous.
- We must become adept at predicting and understanding what users will ask, not just what tasks they perform. This means analysing search queries, support tickets, and chatbot logs to continually refine content.
One comment about these changes from Craig Wright on LinkedIn was, with the exception of metadata isn’t that what the job has been for a long time anyway?
And he’s right. If you’ve been on a training course like Cherryleaf’s foundation course. Then a lot of these. Principles are ones that we teach.
Our city was a subtle change. So many respects. May simply be that you write your content as you have. Before, but you check it you check do the topics that you have written. To the questions that people are likely to ask, are there questions that people might ask where there are gaps?
Although he said there is unlikely to be a series of long static FAQs in documentation, it might be that your content is repackaged and repurpose for AI systems to use where it has a list of questions and a direct mapping. To the answers within your knowledge base,
So, the future isn’t about crafting a list of 20 commonly asked questions. It’s about building the robust, intelligent, large system that is able to handle any question that a user might ask and deliver a meaningful answer to them.
We think the skill set of the technical author will evolve the core competencies will remain.
More likely to need to learn tools publishing content to the AI ready formats or learn the new features that are added to existing platforms. The tools that we know and love today. We’re likely to need to design content specifically for AI consumption and for question-based discovery.
Only two you need to continuously maintain, and we find this content to make sure that the AI systems are learned from the right sources and that they improve over time.
You know these changes will shift does bring challenges.
AI systems can hallucinate.
They can provide incorrect answers on this important nuance.
A badly designed chatbot interface can very easily feel like a modern day Clippy.
And there is a risk that users receive shallow answers when what they need, what they want is deeper context.
And this is where technical communicators shine. They can structure content really well.
And they can oversee the quality of the AI driven experience.
The promise, the opportunity is that we end up with faster open resolution and happy users.
That users find it easier to find the content that they need. That there’s less strain on support teams by being able to deflect more calls.
And potentially a more strategic role for technical communication professionals. Where they are, the architects and curators of knowledge systems.
I mentioned that they have completed a survey into the use of AI technical communication. It’s the third survey that we’ve done. We’ve done one each year for the past three years and.
The survey didn’t show that people are doing this, yet there was no mention of a question-and-answer approach in the answers that we were given. I won’t go through all of the results in the survey.
You’ll find that on our blog, and I’ll provide a link to that in the show notes.
The survey indicated that at the moment or within the last 12 months, AI is being used by technical communicators to do things like generate and draught first draughts of topics generate outlines. And to what’s called scaffolding of content, using it to edit and rephrase complex sentences or UI text make things clearer. To summarise. The texts so you can get the gist of them.
Things like meeting notes and the like. To do some linking and checking of things like is the content following the style guide. Although there are some limitations in that in that. Often chatbot file to do that correctly, they’re using it to research new topics and gather information.
Either using it to create code samples and build regex or regex expressions help with some of the more technical sides of HTML and websites, CSS, JavaScript, automating some formatting, things like markdown tables.
And a little bit of process and workflow improvements to start automating some tasks of getting content published. Things like tagging articles.
We also noticed from the survey that. For the past three years that we have won this, there is a percentage of users that have said that their company has no plans to develop a chat bot. In 2024, that was 32% of respondents.
This year it’s 28. And it may well be that around the third of all organisations, for all of different reasons, we won’t have or won’t offer a chat bot to end users the numbers of. Respondents that said they now have a chatbot went from 9% in 2024 to 23% in 2025, and for those that have a chatbot behind firewall that moved from 8 to 11%.
So, I would expect at least half of all organisations to offer a chatbot that provides that draws on a knowledge base.
If you’d like to know more information on how familiar the respondents said they were with AI and the technology and the tools. How much they’re using Agentic AI today, then have a look at the blog post. Which is called “Results from our using AI Intercom 2025 survey”.
What do you think? Do you think we’re moving towards the question and answer the world?
How is your organisation navigating this shift? Are you already designing for question-based access?
You can share your thoughts on the blog post that we did on our website. You can e-mail us info@cherryleaf.com.
And we also posted it to LinkedIn so you can add your comments on the thread there.
And if you’re interested in trading the services that we provide the technical routing services and the training courses, you can contact us by e-mail or have a look at our website, which is www.cherryleaf.com. Again, thank you for listening.
Leave a Reply