Skills-based hiring needs people who never applied

By SkavanderHR · 13 September 2026

Dropping the degree requirement and testing for skills instead is a real improvement, and it only works on the people who turned up. If the strongest people in your market never applied, a better way of assessing applicants does not reach them. Skills-based hiring is a measurement method. It is not a way of widening the pool.

Key takeaways

• Adoption is now mainstream: 70% of employers use skills-based hiring for entry level roles, up from 65%.
• What a written application proves has weakened. 74% of US job seekers use AI in the process, 32% have claimed AI skills they do not have, and 28% admit generating fake work samples.
• Between 74% and 87% of developers describe themselves as not actively looking, so they are in nobody’s applicant pool.
• Assessing better and reaching further are two separate jobs. Doing the first well does not do the second at all.
• Work evidence you found yourself, before the person knew about your role, is stronger than work evidence they prepared for you.

What does skills-based hiring do well?

It fixes a filter that was measuring the wrong thing. A degree, a former employer’s brand or a tidy career line tells you about access and circumstance more than capability, and dropping them widens who can be considered. Structured work samples and role-relevant tests are a genuine gain in accuracy. That is why adoption keeps climbing rather than fading as a slogan.

The method has one boundary, and it is worth naming plainly: it improves the judgement you make about a candidate. It does nothing about which candidates exist to judge.

What is actually in the applicant pool in 2026?

Two pressures are squeezing the pool from opposite sides. Volume is up and the evidence inside that volume is weaker than it used to be.

Skills-based hiring adoption, and what the applicant pool now contains

Figure	What it tells you	Source
70% use skills-based hiring for entry level roles, up from 65%	The method is mainstream, not an experiment	NACE, Job Outlook 2026
254 applicants per posting	How much material a skills assessment has to sort	Greenhouse platform data, 2025
74% of US job seekers use AI in their application	A written application is weaker evidence than it was	Greenhouse, 2025
32% have claimed AI skills they do not have	Claimed skills need verifying, not reading	Greenhouse, 2025
28% admit generating fake work samples	The work sample is now also a contested artefact	Greenhouse, 2025
74% to 87% of developers not actively looking	The size of the group that is in no pool at all	Stack Overflow Developer Survey

Sources: NACE’s Job Outlook 2026 survey for adoption; Greenhouse’s own platform data and job seeker survey for the volume and self-reported figures; Stack Overflow’s developer surveys across several years for the not-looking share. The Greenhouse figures are self-reported by job seekers, and the Stack Overflow share is a self-description, so both are directional rather than audited.

Why does the method get weaker as applications get easier to produce?

A skills-based process asks for evidence of skill. That was a sound trade when producing evidence cost the candidate real effort, because effort itself was a signal. Now a plausible written answer, a tailored summary of experience and even a work sample can be generated in minutes, and the people most willing to do that at volume are not the people you are trying to find. The industry has a word for the result already: skillfishing, claiming skills that do not survive contact with the actual job.

None of this makes assessment pointless. It means the burden shifts to evidence you can verify independently: a public contribution, a shipped product, a talk, a body of work with dates on it. That evidence is easiest to trust when you found it first and approached the person because of it, rather than receiving it as part of a submission.

Does a wider pool contradict skills-based hiring?

No, it completes it. The reason people support skills-based hiring is usually that the old filters excluded capable people. That argument applies with even more force to the largest excluded group of all: capable people who never applied, and were therefore never filtered at any stage.

Look at what “we widened the pool” usually means in practice. The requirements list was shortened, the degree line came out, and the posting was promoted a bit harder. Every one of those changes operates on people who read job postings. The share of a technical market that reads them at all is a minority.

What this means for a company of twenty to five hundred people

At this size the hiring volume rarely justifies an assessment platform per role, and there is usually a mix: some roles with plenty of strong applicants and one or two with none. The split worth making is this. Where applicants arrive and some are strong, the skills-based process is the right investment, because the open question is judgement and the method improves judgement. Where the applicants arrive and none are close, the process is not the bottleneck, and a better test applied to the same weak pool produces the same answer faster.

The second case is the one worth naming out loud in a hiring meeting, because it gets misdiagnosed as an assessment problem for months. It is a reach problem. The comparison worth making at that point is between an agency, RPO, an embedded recruiter and a standing monthly search, on the single axis of who each of them actually reaches.

Where we fit

We do the reaching, not the assessing. For one specific role, the system looks through professional networks and company websites for every person in your market whose real work matches your criteria, and each of them is told about the role about that role. You keep your own skills-based process exactly as it is, and you run it on a set of people who were found by their work rather than by their willingness to apply. Everyone relevant is usually reached within two weeks, and it costs about a tenth of a recruiting agency. For a family such as AI engineers, where public work is abundant and applications are rare, that is the difference between assessing a pool and assessing a market.

Skills-based hiring deserves the attention it is getting. Give it something worth measuring.

Numbers from our own searches. Written and checked by SkavanderHR.