It’s Because of the FAQ’s about our MYOB and Xero courses
We’ve been rolling out AI-assisted customer service across a few of our own websites lately, and it’s given us a pretty clear, real-world lesson: the AI isn’t the hard part. The knowledge behind it is.
Take two of our own sites as an example. On computertrainingonline.com.au, our AI chat agent handles pre-sales questions about our Xero, MYOB, QuickBooks and Microsoft Office courses with genuine confidence, it gives specific, accurate answers about course access, certificates, payment plans, support channels, and more. Over on natbooks.com.au, our National Bookkeeping site, the same kind of AI-assisted chat doesn’t have nearly as much to work with, and it shows.
Same underlying AI technology. Very different quality of answer. Here’s why.
It All Comes Down to Our FAQ Page
The reason our computertrainingonline.com.au chat agent performs so well is our FAQ’s page, and we don’t say that lightly. It’s not a token “top 5 questions” page. It’s a genuinely thorough, 30-question resource covering everything from course duration and certificates, to payment plans, refund policy, support channels, CPD points, and even how insurers and rehab consultants can enrol clients.
That page didn’t happen by accident. It’s been built up and refined since 2011 by actually paying attention to the questions our students and prospective students ask, then writing clear, specific, honest answers to them.
When you feed an AI system that kind of detailed, well-organised, first-hand knowledge, it has something real to work with. It’s not guessing or generalising, it’s answering the way one of our own team would, because it’s essentially drawing on the same accumulated knowledge our support team has built over more than a decade.
Why This Matters for Any Business Thinking About AI Customer Service
This is a pattern we expect to see everywhere as more businesses experiment with AI chat agents: the AI is only ever as good as what you give it to work with.
A generic AI chatbot with no grounding in your actual business will happily make things up, or fall back on vague, unhelpful answers. But an AI system trained on a genuinely detailed FAQ, your policies, your course structures, your actual support history, starts to sound like it actually knows your business. Because, in a very real sense, it does.
This is exactly the kind of thing we’re exploring further on our AI in Marketing, Sales and Customer Service page, where we go into more detail on how tools like Claude can be set up with custom knowledge, instructions and integrations so they genuinely reflect how a business works, rather than giving generic, one-size-fits-all answers.
The Takeaway
If you’re a business owner thinking about adding AI to your customer service, sales or marketing, don’t start with the AI. Start with your knowledge base.
Sit down and write out, honestly and specifically, the questions your customers actually ask you, the ones your team answers over and over on email, phone and website chat, and answer them properly. That FAQ page, or knowledge base, or set of support documents, is what turns a generic AI chatbot into something that actually sounds like your business.
We did it the slow way over more than a decade on computertrainingonline.com.au, and it’s paying off now in a way we didn’t originally plan for.
Want to know more about how AI can be trained to work the way your business works? Read our AI in Marketing, Sales and Customer Service page, or get in touch, we’re happy to talk through what we’ve learned.
