Gem, HireEZ and Manatal tied for the top of this category: each was named in 4 of the 6 buyer answers we collected. On 2026-08-07 we put six candidate-sourcing questions to ChatGPT and Gemini, let each assistant run one live web search, and recorded every product named and every source cited. No single product ran the table. Even the three co-leaders missed a third of the questions, topping out at 4 of 6 (67%).
Behind the leaders, five products landed in exactly 3 of 6 answers (50%): Betterleap, Indeed, JazzHR, Pin and SignalHire. Three more were named in 2 of 6 (33%): Humanly, LinkedIn Recruiter and Zoho Recruit. The tail is the real story of the field: 14 of the 25 named products showed up in just 1 of 6 answers, so more than half the category was a single mention that a slightly different question would have missed.
The shortlist rewrote itself for every question
The bare query "best candidate sourcing software" returned 8 products, and 2 of them appeared nowhere else in the study: Deel Hire and Joveo. Ask the most generic question and you already collect names that vanish the moment a buyer adds one qualifier.
"Best free candidate sourcing software" produced its own list of 8, in which 4 names were unique to that single question: SeekOut, Juicebox (PeopleGPT), Fetcher and Wellfound. Half of the "free" answer was built from products no other phrasing surfaced.
"Most affordable candidate sourcing software" returned a different 8, with 5 names unique to it: Recruit CRM, GoHire, Cognism, Crustdata and ZoomInfo. Three of those, Cognism, Crustdata and ZoomInfo, are contact-data tools rather than applicant systems, and they appeared only under "affordable," in 1 of 6 answers each.
Here is the fan-out in one number. The "free" answer and the "affordable" answer, two questions most buyers would treat as synonyms, shared exactly 1 product between them: Pin. Eight names on each side, one in common. Whichever word your customer types decides almost the entire shortlist they see.
The neutral phrasings held together far better. "Best," "what should I use" and "recommendations" shared 4 products across all three answers: HireEZ, Betterleap, Indeed and SignalHire. The general questions converge on a stable core while the qualified ones scatter. The other three questions each added their own one-time names too: Team Engine (only in "small teams"), PowerToFly (only in "what should I use") and Recruiterflow (only in "recommendations").
One more sign of how much phrasing matters: Gem, a co-leader at 4 of 6, was absent from the bare "best" answer and from "most affordable." It earned its ranking entirely through "small teams," "free," "what should I use" and "recommendations."
The source map: who fed the answers
One site fed every answer. peoplemanagingpeople.com, an editorial roundup, was cited in 6 of 6 questions, more reach than any other source by a clear margin. If you sell in this category and you are not in that roundup, you are missing the one page both assistants leaned on 100% of the time.
Community discussion showed up as Reddit, cited in 4 of 6 answers. It fed the two budget questions ("free" and "affordable") and the two broad ones ("best" and "small teams"), but not the two direct-advice questions. Quora did not appear at all. Neither did any review directory: G2 and Capterra are absent from all 6 answers, which is worth sitting with if your team spends its energy maintaining directory profiles.
The key finding is a vendor's own domain doing the work of a review site. gem.com, Gem's own website, was the second most-cited source at 5 of 6 answers, behind only the top roundup, and Gem was named in 4 of 6. A vendor publishing its own comparison content, earning citations for it, and getting named on the back of it: that is the entire playbook on one line.
Gem is not alone in it. pin.com fed 2 of 6 answers and Pin was named in 3 of 6; zoho.com fed 2 of 6 and Zoho Recruit in 2 of 6. The counter-example is just as useful: juicebox.ai was cited in 4 of 6 answers, tying Reddit, yet Juicebox (PeopleGPT) itself was named in only 1 of 6. Its content is read as a neutral roundup, so it hands citations to competitors more than to itself. The remaining influential sources were reviews and roundups: selectsoftwarereviews.com, metaview.ai and recruiterflow.com were each cited in 3 of 6 answers.
What a vendor here would do to get named
The pattern across all 6 answers is blunt: the source list, not market share, decides who gets named. Neither ChatGPT nor Gemini holds a prior opinion about candidate sourcing software. Each ran one web search and named whatever the top pages named, so getting named means getting onto those pages.
Concretely, that is a short list of specific sources: the peoplemanagingpeople.com roundup (in 6 of 6 answers), the Reddit threads (4 of 6), and the editorial reviews at selectsoftwarereviews.com and metaview.ai (3 of 6 each). Gem's result points to the fourth move: publish your own comparison content and earn citations for it, because gem.com alone fed 5 of 6 answers.
And because a single qualifier reshuffles the whole list, you measure this per question, not once. Pin was named in 3 of 6 answers overall but in 0 of the three neutral questions; it lives entirely inside "small teams," "free" and "affordable." A vendor watching only the word "best" would conclude Pin is invisible, when in fact it owns the budget queries.
Where this leaves Glotier
Glotier does not sell candidate sourcing software, so we are correctly absent from all 6 of these answers, and we would not have it any other way. Naming ourselves here would be exactly the fabricated placement this study exists to catch. We ran the measurement because it is the same one we run for a customer's own category.
Everything above came from six questions and two assistants on a single afternoon. The useful version is the one pointed at your product: which of the 6 answers name you, which of these sources you are missing, and how your shortlist changes from "best" to "free" to "affordable." The check is free, needs no account, and takes about a minute.
The full list, counted
| Product | Named in | Share |
|---|---|---|
| Gem | 4 of 6 | 67% |
| HireEZ | 4 of 6 | 67% |
| Manatal | 4 of 6 | 67% |
| Betterleap | 3 of 6 | 50% |
| Indeed | 3 of 6 | 50% |
| JazzHR | 3 of 6 | 50% |
| Pin | 3 of 6 | 50% |
| SignalHire | 3 of 6 | 50% |
| Humanly | 2 of 6 | 33% |
| LinkedIn Recruiter | 2 of 6 | 33% |
| Zoho Recruit | 2 of 6 | 33% |
The other 14 products were each named in 1 of 6 answers (17%): Cognism, Crustdata, Deel Hire, Fetcher, GoHire, Joveo, Juicebox (PeopleGPT), PowerToFly, Recruit CRM, Recruiterflow, SeekOut, Team Engine, Wellfound and ZoomInfo. Each brand appeared under a single label, so no row in this table folds two names for one company into a number the measurement did not record.