“AI SEO” is everywhere right now. GEO, AEO, AI Search Optimisation, whatever name gets used, the promise is the same: get your business mentioned when someone asks ChatGPT, Gemini, or Perplexity for a recommendation. And with that promise comes exactly what you’d expect from any young, fast-moving corner of an industry. A lot of relabelling. A lot of agencies bolting “AI SEO” onto a website without changing what they actually do.
So here’s what we think a business should genuinely ask before handing anyone money for this. Not because we’re guessing, but because this is the same process we run on ourselves.
Five Takeaways
“AI SEO” is being widely relabelled. A lot of agencies are bolting the term onto existing services without changing what they actually deliver, so it’s worth asking pointed questions before paying for it.
Not appearing in an AI answer rarely means the content was bad. It’s usually a structural gap: missing reviews, a missing directory listing, or a technical indexing issue stopping content from being seen at all.
Different questions pull from different sources. Local questions can be driven entirely by Google Business Profile signals, comparison questions by B2B directories like Clutch, and informational questions by written content, so a single generic fix won’t cover everything.
There’s no “page one of ChatGPT” to promise. Genuine progress is measured by re-testing the same realistic questions over time, not by a fixed ranking position.
Pricing should match the business. A UK SME shouldn’t be quoted the same retainer as an enterprise AI search programme.
1. Where do you currently appear when a realistic customer asks?
Not a hypothetical. Not a demo. The actual questions a real customer would type into ChatGPT or Perplexity, asked in a fresh conversation with no memory of your business skewing the result.
We do this by writing out realistic buyer questions in three categories: direct category questions (“best SEO agency in Worcester”), comparison questions (“should I hire a local agency or a national one”), and need-based questions from someone who hasn’t named a service yet (“my website isn’t generating enquiries, what should I check first”). Then we run each one, honestly, and log exactly what comes back, including when the answer is that we don’t appear at all.
2. Which competitors are being recommended instead, and why?
This is where the useful work actually starts. When you don’t appear in an answer, the reason is rarely “the content wasn’t good enough.” Often it’s something structurally different: a local business listing gap, a missing directory profile, a technical indexing issue stopping content from ever being seen in the first place.
We’ve found all three in our own testing. A local SEO question pulled entirely from Google Business Profile signals, ratings and review counts, not written content at all. A web design question cited B2B agency directories like Clutch and GoodFirms by name. A pricing question pulled from written guides with real, specific numbers. Three different questions, three genuinely different mechanisms. Treating them as one problem with one fix is exactly how a lot of “AI SEO” packages fall short.
3. What specific work will actually improve it?
Once you know why you’re missing, the fix is usually one of a handful of things: cleaning up technical indexing issues so Google and AI crawlers can actually find and trust your content, building genuine reviews and directory presence where local or comparison questions are drawing from, or writing content with real, specific, provable claims rather than vague marketing copy, because that’s consistently what gets cited over generic pages.
None of that is exotic. It’s disciplined, honest, and specific to what your actual testing turns up, not a generic package applied the same way to every client.
4. How will you actually measure improvement?
Not a promise of “page one of ChatGPT,” because that’s not how AI answers work. There’s no fixed ranking position to chase. What can genuinely be measured is whether you start appearing in the same fresh-conversation tests over time, and whether the reasons you were missing before have actually been addressed.
We track this by re-running the same question set periodically, in memory-free conversations, and comparing honestly against the baseline. If nothing’s moved, we say so.
5. What’s a sensible cost for a business like mine?
If a UK small business is being quoted an enterprise-level retainer for this, that’s worth questioning. AI search optimisation done properly for an SME shouldn’t cost the same as a national brand’s ongoing programme. Be wary of anyone pricing it as though it does.
Why we approach it this way
We’ll be honest: AI search is still genuinely new, and nobody, including us, has this fully solved. What we won’t do is relabel standard SEO as “AI SEO” and charge more for the same work. The process above is exactly what we run for ourselves, testing our own visibility, finding our own gaps, and fixing them properly rather than assuming a generic package will do the job. Check out two case studies where we’ve put this into practice to great effect for two clients: BladeRoom and Comlux.
If you’re a Worcester or West Midlands business weighing up whether this is worth investing in, we’d rather show you exactly where you currently stand than promise you something we can’t measure.
Want to see where your business actually stands in AI search right now?
Fill in the form and tell us
FAQs
They’re largely different names for the same goal: getting a business mentioned, cited, or recommended when someone asks an AI platform like ChatGPT, Gemini, or Perplexity for a supplier.
Ask them to show where your business currently appears in realistic AI-generated answers, which competitors are recommended instead, and how they’ll measure improvement. A genuine provider can answer all three specifically.
Yes, particularly for location-specific questions. National specialists often have no presence in local answers at all, which is a genuine opening for a well-prepared local business.
There’s no fixed timeline, since it depends on what’s causing the gap. Technical indexing fixes can show movement within weeks, while building genuine reviews and directory authority is a slower, ongoing process.
Where you currently appear for realistic customer questions, which competitors are recommended instead, what specific work will be done to close the gap, how improvement will be measured, and whether the pricing is realistic for a business of your size.


