Your customers are asking AI first. What is it telling them?

Here's something I get people to try in our first conversation. Open ChatGPT, Gemini or Perplexity and ask it a question your best customer would ask, and make it a specific one, like what something costs, whether it's in stock or whether you deliver to their town. Then read the answer as if you'd never heard of your business.

For most brands it's vague at best. I've seen prices from last season, products that stopped selling a year ago, and plenty of answers that recommend a competitor simply because their details were easier to read.

It's not an AI problem

It's tempting to blame the assistant, but it isn't being unfair. It's working with whatever it can find about you, so if your prices live in images, your policies are buried in PDFs and your product details change from one page to the next, that's exactly what it repeats back.

AI doesn't create the gaps. It just finds them faster than your customers do.

That's actually good news, because it means the fix is in your hands.

Where I'd start

Write down the 20 questions customers ask you most, which your sales and support people will probably know off by heart. Ask each one in three different assistants and save the answers, then go through and mark every one that's wrong, vague or missing. The useful bit is tracing each of those back to where it came from, whether that's a page, a product feed or a fact that isn't written down anywhere.

What you end up with is a backlog, and most of it is content, data and journey work your team can ship without buying anything new.

Then check again next month, because the answers change and so do your competitors.