Candidate identity checks are moving to the front of the hiring funnel, and the vendors selling them describe the benefit in terms of trust. Any HR team that buys one should write down what it expects to change before the first candidate is verified, because afterwards nobody will remember the baseline.
The product category is arriving in the stack
On 19 October, Match2 announced enhancements to what it calls Connected Talent Solutions, including optional biometric identity verification, credential-first discovery and pre-apply qualification. The company says the tools are available through SAP Store and Phenom Marketplace and integrate with more than 100 hiring platforms. It frames them as an addition to the applicant tracking system, which, in the release’s words, “remains the System of Record”.
I am using Match2 as an example of a category: tools that confirm who a candidate is and what they hold before the application reaches a recruiter. The company describes the aim as giving employers “stronger signals to mitigate impersonation, misrepresented qualifications, and bot-driven noise”. The release also says plainly that “Verification strengthens trust; it does not eliminate all fraud”. That sentence is the most useful one in the document, because it tells a buyer that the product reduces a problem and leaves a remainder.
My argument: write the baseline first
HR teams should set three numbers before any verification step goes live: how many applications a recruiter screens per hire, how many of those are discarded as fake or misrepresented, and how long that screening takes. Without them, a purchase approved on the promise of trust cannot be judged a year later. Every team will be able to say that candidates seem more trustworthy. Few will be able to say by how much, or what it cost them in applicants who left at the verification screen.
The reason is that identity checks change two things at once. They filter out some bad applications, which is the intended effect. They also add a step, and any added step changes who finishes the application. A team that only measures the first effect will credit the tool with a cleaner pool and miss a smaller one. The release itself puts the aim as “more useful applications, not just higher volume”, which is a goal about volume and quality together. Only a before-and-after count shows whether both moved in the direction the vendor describes.
The same discipline applies to the AI systems that sit downstream. Match2’s chief executive, Darren Simons, says in the release that “As AI matching becomes ubiquitous, trusted, real-time candidate intelligence becomes the advantage.” If that is right, the signals feeding the matching layer matter as much as the matching itself, and HR owns the question of which signals it accepts. Our earlier analysis found that hiring reports split on whether trust followed faster sourcing. A verification layer is one way to try to close that gap, and it deserves the same measurement as any other.
The strongest objection
The best counter-argument is that a baseline is a luxury. Recruiters already face impersonation and bot-written applications, the argument runs, and a tool that confirms identity earlier will save time whether or not anyone builds a spreadsheet around it. Waiting to measure means living with the problem for another quarter. Candidates also gain something real: the release says they keep control of what they share and carry their information between participating sites, which spares them retyping the same details.
I accept that identity checks can help and that candidate-side benefits are real. The baseline takes a few days of work, though, and it protects against the failure that tends to follow a quiet rollout: a tool that stays switched on for years because turning it off feels risky, while nobody can show it paid for itself. A team that records the three numbers can keep the tool, tune it or drop it on evidence. A team that skips them has only the vendor’s account to go on.
What I would do this quarter
- Pull the last three months of screening data from your ATS and fix the three baseline numbers.
- Ask any verification vendor which of those three numbers it expects to move, and by how much, in writing.
- Decide in advance what result would end the pilot, and who makes that call.
- Record how many candidates abandon the process at the verification step, since that cost never appears in a recruiter’s queue.
- Add the tool to the hiring-AI inventory that Q4 is the last window to build.
Verification may well become a standard part of recruiting platforms. The teams that benefit will be the ones that can say what it did to their own funnel.
