UK Tech Recruitment AI Visibility Index 2026: which recruiters ChatGPT, Gemini and Google AI name on their own

Last updated: 8 September 2026

There is plenty of opinion about which recruitment agencies matter in UK tech hiring, and almost none of it says which ones an AI engine will actually bring up when a founder or hiring manager asks for help. This index closes that gap. On 8 September 2026 we put 55 real hiring questions to ChatGPT, Gemini and Google AI Overview, covering everything from a first engineering hire to a VP of Engineering search, and recorded 622 answers. We tracked 13 UK tech recruitment brands against that single denominator. The finding worth remembering: the most-named agency in the whole set is still absent from roughly 96 percent of answers, and five of the 13 brands we tracked, more than a third of the field, never appeared once.

The numbers in brief

• Harnham is named most often, in 4.02 percent of the 622 answers. That is roughly one answer in twenty five.
• The median across the eight agencies that appeared at least once is 0.88 percent.
• Richard Wheeler Associates follows at 2.73 percent, then Talentful at 1.29 percent and Understanding Recruitment at 0.96 percent.
• Five of the 13 tracked brands, more than a third of the field, never appeared in a single answer.
• LinkedIn is retrieved in 9.0 percent of answers, more often than any recruiter's own website.

What we measured, and how

We track 55 questions phrased the way a founder or hiring manager actually types them, split across five themes: choosing a tech recruitment partner, hiring AI and data talent, building engineering and product teams, executive search and leadership hires, and contract, embedded and interim hiring. Each question is put to three engines. The figures in this report are drawn from 622 answers recorded on 8 September 2026 across ChatGPT, Gemini and Google AI Overview. We record what the engine returns, not what it was asked to return.

Three measures, in plain words:

• How often AI names them: the share of the 622 answers in which the agency is named.
• Share of the answers: the agency's portion of all 110 brand mentions across every answer.
• Average position: where the agency tends to land in a list when it is named. Lower is better.

Every agency was measured against the same 622-answer denominator. No agency paid to be included, and the names in this report identify research subjects, not customers and not endorsements.

Who the engines actually name

That table is not a truncated top slice. It is the entire scoring field. Of the 13 brands we tracked, only these eight were named in a single one of the 622 answers. The other five, more than a third of the tracked field, did not appear once.

The finding that matters more than the ranking

Leading this category means almost nothing yet. Harnham, the most-named agency in the set, is still missing from about 96 percent of the answers we recorded. The gap between first and second place is one and a quarter percentage points, and by the fourth name the whole table is bunched inside a single percentage point. In a field this thin, no agency has claimed the answer, which means the recommendation is still open rather than defended.

There is a second layer worth sitting with. Thirteen tracked brands and only eight ever named is not a crowded market failing to differentiate. It is a market where most of the named competitors that a founder would list off the top of their head are, from an engine's point of view, invisible. The eight that do appear are not necessarily the eight a London operator would name first. They are the eight whose evidence the engines could actually find.

Frequency, position and share are three different things

Two agencies named in a similar number of answers are not always doing the same job.

Harnham does not just lead, it leads on every axis we measured. It is named in the most answers, at 4.02 percent. It lands earliest when it does appear, with an average position of 1.8. And it holds close to half of every brand mention across the whole category: 45.45 percent of the 110 total mentions logged across all eight scoring agencies. No other agency in this set matches it on more than one of those three measures.

The exceptions worth noting sit further down the table. Talentful is named in only 1.29 percent of answers, well back in third place, yet its average position of 2.0 is the second best in the table, ahead of Richard Wheeler Associates at 2.6. When Talentful does get named, it tends to land near the top rather than buried in a long list. Understanding Recruitment and Signify Technology log the warmest sentiment scores in the category, 75 and 77 respectively, despite each being named in under 1 percent of answers. Being named rarely and being named warmly are not the same problem, and they do not have the same fix.

Leadership shifts depending on the question

The category splits into five distinct hiring topics, and no single agency leads more than two of them.

Read this table against the category-wide numbers above and the pattern is stark. Harnham's 4.02 percent category figure hides a much stronger position on one specific question set: it is named in 17.74 percent of the 124 answers about hiring AI and data talent, more than four times its overall rate. Talentful's category rank is third at 1.29 percent, yet it leads the contract and embedded hiring topic outright at 6.96 percent of that topic's 115 answers. La Fosse sits fifth category-wide at 0.80 percent but is the name most often given when someone asks how to choose a recruitment partner in the first place. Understanding Recruitment, fourth overall, tops the engineering and product team building topic even though its topic-leading share is still under 2 percent.

The practical read is that this category rewards a narrow, specific reputation more than a broad one. An agency's overall visibility number can understate how strong its position is on the one or two questions it is actually built to answer.

Where the answers come from

Two things stand out. First, LinkedIn alone is retrieved in 9.0 percent of answers, more than any single recruiter's own website, including Harnham's. Harnham leads the category on being named, yet harnham.com is retrieved in only 2.57 percent of answers, less often than LinkedIn, Reddit or gov.uk. Being named does not require your own site to be the source the engine actually opens.

Second, community platforms carry real weight here. LinkedIn, Reddit and YouTube together account for a larger share of retrieval than any single vendor domain in the table. For a category built on professional reputation, that tracks: LinkedIn posts, Reddit threads about hiring experiences and gov.uk pages on contractor rules are exactly the kind of evidence an engine would reach for when a founder asks who to trust with a search.

What this report can and cannot tell you

These results are directional, not definitive. They describe one question set, on three engines, on one day.

Read this way, the data can answer a bounded set of questions:

• Which UK tech recruitment agencies the engines name without being prompted with a brand name.
• How concentrated, or in this case how thin, the field is across a meaningful set of hiring questions.
• Which source types the engines reach for when they build a recruitment shortlist.
• Where leadership genuinely sits by topic, rather than by a single blended score.

It cannot do several other things, and it is worth being blunt about them. It cannot prove that appearing more often produces more briefs or more placements. It cannot establish a causal link between anything a given agency did and its position here. It cannot tell any individual agency that it ought to appear across all five topics, because most agencies have no real claim on all five. A boutique AI and data specialist has no business chasing a graduate-hire question.

The operating instruction is therefore not to chase a single blended score. Split the topics by the ones you could genuinely win, read the exact answers the engines give for those, and repair the weakest layer of evidence underneath them.

What this means if you run a UK tech recruitment agency

If you are a boutique or founder-led firm, the thinness of this field is the opportunity, not a wall. A category where the leader sits at 4.02 percent has no incumbent worth deferring to. Harnham's position looks strong mainly because it is one of the few agencies whose evidence, on LinkedIn, on Reddit, on its own domain, is actually there for an engine to find. That evidence base is buildable by anyone with a genuine specialism and the patience to document it.

If you specialise in one hiring motion rather than trying to cover all five, this dataset says that is the right call. The topic table shows leadership going to the sharpest reputation, not the broadest one. An agency that owns AI and data hiring, or contract and embedded placements, has more to gain from strengthening that one position than from trying to appear everywhere at a shallow depth.

If you are comparing this category to others we have measured, the contrast is useful. Our contact centre QA software index covers a category where the median brand appears in over 14 percent of answers, more than fifteen times the median here. Our Texas dining index sits somewhere in between, with a 1.17 percent median across restaurants. Recruitment, at least in this UK tech corner of it, is one of the least settled categories we have measured yet. The same open-field pattern shows up in adjacent professional-services work too, including our Middle East on-demand strategy consulting index, and for a sense of how differently this plays out in a consumer category, see our crypto wallet, London burger and London cookie bakery indexes.

Sources

• Schmitdy category tracking, UK tech recruitment question set, 8 September 2026. The 622 answers, 55 tracked questions and 13 tracked brands behind every figure in this report.
• Harnham, harnham.com. Official site of the most-named brand in the index.
• Richard Wheeler Associates, rwaconsultants.com. Official site, second in the published table.
• Talentful, talentful.com. Official site, third in the published table.
• Understanding Recruitment, understandingrecruitment.co.uk. Official site, fourth in the published table.
• La Fosse, lafosse.com. Official site, fifth in the published table.
• Xcede, xcede.com. Official site, sixth in the published table.
• Signify Technology, signifytechnology.com. Official site, seventh in the published table.
• Trust in SODA, trustinsoda.com. Official site, eighth in the published table.
• LinkedIn, linkedin.com. The most-retrieved source across the tracked answers.
• Reddit, reddit.com. Second most-retrieved community source.

Every brand named in this report is a research subject. Inclusion is not an endorsement, and no agency paid to appear.

If you want to know which of these questions your own agency already shows up in, we can run the same measurement for your market.