Why isn't your recruitment agency showing up in AI search?

What the diagnostic result means

Why recruitment agencies don't appear in AI search

1. AI doesn't clearly understand what your staffing agency specialises in

AI models build a mental map of who does what before they ever answer a question. If your agency's specialism isn't stated plainly, consistently, in enough places, you're not on that map at all. Not blocked. Not penalised. Just absent.

This is different from a ranking problem. An AI platform needs to associate your recruitment agency with the sectors you recruit for, roles you fill, locations you cover and evidence supporting your expertise before it can confidently include you in relevant recommendations.

2. Your recruitment firm ranks in Google but isn't recommended by AI

This is the split that catches most agencies off guard.

For example, in our recruitment study Harvey Nash has strong, established Google visibility. Ask ChatGPT the equivalent buyer question and the picture changes: 5% AI recommendation share versus 82% on Google, a 77-point gap, and ChatGPT mentioned them zero times across all 11 queries. Matchtech tells a smaller version of the same story: 6/11 Google queries, one ChatGPT mention, zero Perplexity or Google AIO mentions.

Neither firm has a content problem in the traditional sense. Both rank. Neither gets recommended when the question moves from a search box to a chat window.

The mechanics behind this are measurable, too. Across AI Overviews generally, the overlap between what ranks organically and what gets cited in the AI answer has collapsed, from roughly 75% down to somewhere between 17% and 38%. Two different games. Winning one says less than it used to about the other.

3. Other websites aren't corroborating your positioning or specialism

AI systems lean hard on third-party confirmation. Not your homepage's claim about itself, someone else's page saying the same thing.

Adria Solutions was a genuinely strange case in the study. Their own roundup pages are one of the sources Google's AI Overview actually pulls from when answering recruitment questions, actively feeding citations to Harnham, Xcede and Understanding Recruitment. Ask ChatGPT about Adria Solutions directly, though, and it's never named them. Not once, across any of the eleven questions tracked. They're doing the citing. They're not being cited.

4. You're mentioned, but your website isn't the source

That's four agencies with genuine name recognition inside AI answers, and zero citations pointing back to anything they control. Every mention runs through somebody else's page.

Worth sitting with the difference between three things people tend to flatten into one: being mentioned somewhere, being cited as the source, and being actively recommended. A roundup article that happens to list your name gets you mentioned. It doesn't make an AI model treat your own site as evidence. And it definitely doesn't mean the model would put you forward first if asked directly.

The "fragile" label isn't decoration. A third party updates their roundup, changes their opinion, or takes the page down, and that mention disappears with it. Nothing about it belongs to the agency it's about.

I've watched this exact pattern show up in every vertical I've studied so far, not just recruitment. The names that keep getting reached for aren't always the ones doing the best work in the category. They're the ones that have been mentioned consistently enough, in enough places an AI model trusts, that reaching for them stopped requiring a second thought. Consistency of description travels further than quality of work, which is an uncomfortable truth for a writer, but the data says it every time I look.

5. Your recruitment pages are too generic

This is where "why recruitment agencies don't appear in AI search" tends to actually get decided, one page at a time. Imagine how a hiring manager might create a query for AI and try to include all relevant options when publishing site content. Include details such as employer, specialism, sector, role type, permanent or contract, and location/geography

Weak

Specialist technology recruitment solutions for ambitious organisations.

Strong

We recruit permanent and contract cybersecurity professionals for UK SaaS companies, including SOC analysts, security engineers, GRC consultants and CISOs.

The first sentence is true of roughly four hundred agencies. The second one is true of a handful. AI systems match questions to specifics. A page that could describe anyone gets treated like it describes no one in particular.

6. You don't answer buyer questions

Most agency content is written for candidates: job specs, career advice, "how to ace your interview." Reasonable, given who actually fills out application forms. But the questions driving AI recommendations come from the other side of the desk entirely:

• Which agency specialises in X?
• Who should I use for X?
• Which agency has experience recruiting X?
• Which recruitment agency would you trust for X?

If none of your content answers those questions in that language, there's nothing for an AI system to retrieve when someone asks them, so it instead retrieves the answers from your competitors or third-party sites.

7. ChatGPT visibility isn't the same as overall AI visibility

XEC Recruitment is the sharpest illustration of this in the whole study. Ask Google's AI Overview which agencies recruit for LLM and generative AI roles in the UK, and XEC comes back first, ahead of Harnham. Ask Perplexity the identical question, and XEC has never been named. Not on that question, not on any of the other ten tracked.

Same firm. Same specialism. Same question, word for word. Completely different answer depending on the AI system chosen.

ChatGPT optimisation is hugely important, their users grew from 400 million weekly users to 900 million in just twelve months. All companies want to get recommended by ChatGPT. But Perplexity and Google's AI Overviews are growing fast too. Each platform is pulling from its own shortlist, built its own way.

Worth remembering, too, that "which AI platform" is rapidly becoming the wrong question. As of September 2026 Microsoft Copilot alone already has over 30 million paid seats and sits inside Outlook, Teams and Word. Google Gemini is built into every Google Workspace app (Gmail, Docs, Sheets, Meet, Chat). Apple Intelligence / Siri already has around 410 million daily active users. Meta AI is embedded across WhatsApp, Instagram, Facebook and Messenger with 1.2 billion monthly active users.

Research powered by AI is happening quietly inside tools people already have open, across multiple platforms. Being visible in ChatGPT doesn't mean you're visible in AI search. ChatGPT visibility is only part of the picture. A recruitment agency invisible to AI risks being left off the shortlist when employers and hiring managers are deciding who to use.

Frequently asked questions