Why Your Blog Posts Are Not Showing Up in AI Answers
Look, I've been doing this for three years now, properly focused on AI search, and the question I get asked most often, hands down, is "why aren't my blog posts showing up when people ask ChatGPT or Perplexity about [whatever their business does]?"
And the answer is never what they expect.
You're still writing for Google in 2019
Here's what I see constantly. Business owner in Hitchin writes a blog post about, I don't know, boiler servicing. They've done the keyword research. They've got their H2s sorted. Meta description is spot on. Internal links? Yep. It's all there.
Then they go to ChatGPT, type in "best way to service a boiler in Hertfordshire" and... nothing. Their business doesn't come up. Their blog post isn't referenced. It's like they don't exist.
So they assume AI just doesn't know about them yet. Give it time, right?
Wrong.
The problem is they wrote that post for a search engine that ranks based on links and keywords and dwell time. ChatGPT doesn't care about any of that. Neither does Perplexity or Gemini or any of the other answer engines people are actually using now in 2026.
You wrote for the wrong reader. Simple as that.
AI needs different food
When ChatGPT or Perplexity pulls information to answer a query, it's looking for something specific. Not keywords. Not backlinks. It wants structured, definitive information that directly answers questions.
Your blog post probably does answer questions. But does it do it in a way an AI can easily extract and trust?
Most don't.
I was working with a solicitor in Letchworth a few months back. Family law stuff. She had this massive blog post about divorce proceedings, really detailed, probably took her hours to write. But it was written like... well, like a blog post. Long intro about how difficult divorce is. Personal story. Then eventually, buried four paragraphs down, the actual useful information about timeframes and costs and process.
ChatGPT couldn't use it. Too much fluff. The actual answers were camouflaged in narrative.
We rewrote it. Same information, completely different structure. Clear questions as subheadings. Direct answers in the first sentence of each section. Lists where they made sense. Specific numbers and timeframes.
Within two weeks, her firm started showing up in AI answers for local family law queries. Not because the AI suddenly discovered her. Because we gave it something it could actually work with.
Your content probably lacks structure that machines can parse
Right, so this is where it gets a bit technical but bear with me because it matters.
AI models are really good at understanding context now. Like, scary good. But they still prefer content that's explicitly structured. That means:
- Clear question-answer pairs. If someone asks "how long does X take", they want to see "X takes 3-5 weeks depending on Y" right at the start of your answer.
- Lists and steps. Not paragraphs that happen to mention steps. Actual formatted lists.
- Definitions at the top. If you're explaining something, define it first. Don't make the AI (or the reader) wade through three paragraphs to figure out what you're even talking about.
- Specificity over style. "Most homeowners in Hertfordshire pay between £800-1200" beats "costs can vary significantly depending on various factors" every single time.
And here's the thing. I've seen blog posts that are really well written, proper engaging, the kind of thing you'd actually want to read. But they're terrible for AEO because they're written to be enjoyed, not extracted.
You can do both. But you have to be intentional about it.
You're not being cited because you're not citeable
This one bugs people when I bring it up but it's true.
ChatGPT and Perplexity and the rest, they're not just summarising random web pages. They're looking for sources that seem authoritative and trustworthy. And one of the ways they judge that is whether your content looks like a source that should be cited.
What does that mean in practice?
Your "About Us" page matters now. Like, actually matters. If it's three sentences of marketing waffle, that's a problem. AI wants to know who you are, what your credentials are, why you're qualified to answer questions about your field.
I had a client in Baldock, runs a recruitment agency. Smart guy, knows his stuff. But his website had basically no information about his background or expertise. Just "we help businesses find great people" type stuff.
We added a proper bio. His 15 years in the industry. Specific companies he'd worked with (permission given, obviously). Some actual credentials.
His content started getting picked up way more often. Same blog posts. Just now the AI had context about why it should trust him.
The format you're publishing in might be invisible
OK so this is something I didn't realise until about six months ago.
Some CMS setups, some themes, some ways of structuring your content... they're basically invisible to AI crawlers. Not technically invisible. Like, the content is there. But the way it's marked up or nested or whatever means the AI models don't weight it properly.
I see this a lot with businesses using page builders. They've got these beautiful layouts with text in columns and fancy boxes and tabbed content. Looks great on screen. Completely confusing for an AI trying to parse it.
Or the content is there but it's not marked up semantically. There's no schema. The headings aren't proper heading tags. The lists aren't marked as lists.
This sounds really technical but the fix is usually straightforward. You just need someone to look at it who knows what they're looking for. Most web developers don't, to be honest. Not yet. This is still pretty new territory.
You're competing with sites that are optimised for AI
And here's the thing that'll really annoy you.
Your competitors, or at least some of them, have figured this out. They're writing content specifically for answer engines. Their blog posts are showing up in AI responses. Yours aren't.
Which means when someone in Stevenage asks ChatGPT for recommendations in your industry, they're getting your competitor's name. Not yours.
And the gap is only getting wider.
Because once a business starts showing up in AI answers, they get more visibility, more traffic, more authority. The AI models notice that and weight them higher. It's a bit like the old Google backlink cycle, except faster and more concentrated.
What actually works
Right, so what do you do about this?
First, stop thinking about your blog as something you write for humans that Google happens to index. Start thinking about it as a knowledge base that needs to work for both human readers and AI systems.
That means clarity over cleverness. Structure over style. Specifics over generalities.
Second, go through your existing content and ask yourself: if an AI model was scanning this page looking for an answer to extract, could it easily find one? If the answer is no, rewrite it. You don't need to delete it and start over. Just restructure it.
Third, get your schema sorted. Make sure your site is properly marked up. This isn't optional anymore.
And fourth, actually test it. Go to ChatGPT or Perplexity and ask questions your potential customers would ask. See what comes up. If it's not you, work backwards and figure out why.
I've been doing this in Hertfordshire for three years now and it's mad how quickly things have shifted. In 2023, businesses were still figuring out whether AI search was even a thing they needed to worry about. In 2026, if you're not showing up in AI answers, you're basically invisible to a huge chunk of your potential customers.
Anyway. If you're reading this and thinking "bloody hell, I need to sort this out but I don't have time to figure it all myself", that's literally what we do. We help local businesses in Hitchin, Royston, Stevenage, all over North Herts really, get their content working properly for AI search.
Book a call if you want to talk specifics about your situation. Or have a look at what we do with AEO in North Hertfordshire if you want more detail on how this stuff actually works. Either way, don't leave it too long. The gap between businesses that show up in AI answers and businesses that don't is getting wider every month.