The pitch is familiar by now. Stop screening on degrees, start screening on skills. Open the funnel to candidates traditional hiring missed. Retain better, perform better, look better in your DEI report.
Then you read the data. The Burning Glass Institute and Harvard Business School, in their joint 2024 study "The Emerging Degree Reset", looked at what actually happened after companies committed publicly to skills-based hiring. Job postings without degree requirements rose from 49% in 2017 to about 56% by 2023, a real shift. But when researchers tracked actual hiring outcomes at the same companies, the share of new hires without four-year degrees changed by less than 4 percentage points across the same period. Pronouncements moved one way. Practice moved much less.
This article is about the gap. What skills-based hiring is in 2026, what it actually changes at the level of one recruiter filling one role, what the proof says about who does it well, and the three working pieces that separate the companies running it from the companies signaling it.
What skills-based hiring actually is
Skills-based hiring is the practice of evaluating candidates against the specific skills a role requires, rather than against degrees, job titles, or years of experience as proxies for those skills. The shift sounds small in theory and is large in practice, because almost every step of the standard hiring loop is built on the proxies. Job descriptions list degree requirements. ATS filters screen by years of experience. Interview rubrics ask about past job titles. Sourcing tools index by school. Removing the proxies and replacing them with skill evidence touches every one of those steps.
The terminology often gets tangled with "skills matching", which is the tooling category. Skills matching is what platforms like Lightcast, CareerOneStop, BigTime, BuildWithin, and to some extent Glozo do when they ingest a job description and a candidate profile and produce a skill alignment highlight. Skills-based hiring is the editorial decision that the company will make hiring decisions on the basis of that signal rather than on degrees. The first one is software. The second one is policy. Companies need both to do this for real.
Why it keeps coming up in 2026
Three drivers keep skills-based hiring on the agenda. None of them are about generosity.
The first is labor supply. Across multiple skilled-trade and tech-adjacent roles, the unemployment rate sits below 3% and the share of open roles per available candidate keeps climbing. LinkedIn's 2023 Future of Work analysis estimated that removing the degree requirement on a typical technical role expands the candidate pool by roughly an order of magnitude, depending on the role. When you cannot find the candidate inside the degree-filtered pool, the calculus on dropping the filter changes.
The second is AI. The same technologies that let recruiters draft a job description in three minutes also let them parse a candidate's evidence (work samples, code, project narratives) at the sourcing stage without reading every profile page by page. The cost of evaluating skill evidence before deciding whom to reach out to has fallen sharply. McKinsey's 2023 work on skills-based hiring economies has documented meaningful reductions in time-to-shortlist in pilot programs at large employers when AI-assisted evidence parsing supplements manual sourcing review. The exact figure is contested, but the direction is clear: the manual cost of evaluating on skills at the top of the funnel is no longer the blocker it was.
The third is the federal example. Multiple US federal hiring directives over the past five years have explicitly removed degree requirements from large categories of federal positions. State governments have followed. Pennsylvania, Maryland, and Utah have all issued executive orders dropping degree requirements on most state jobs. When the largest employer in the country starts hiring this way at scale, private-sector benchmarks shift.
What the data says about who does it well
This is where the honest article diverges from the press-release article. A small number of companies do skills-based hiring well. A larger number signal it.
IBM is the case study most commonly cited. The company began removing degree requirements on technical roles around 2017 and reports that by 2023 roughly 50% of US job openings had no degree requirement, against a baseline of nearly all roles requiring one a decade earlier. Walmart, in its 2022 announcement of its own degree-requirement reduction, reported that 75% of US openings had been adjusted. General Motors, Bank of America, and Accenture have all made and partially executed similar pledges.
What separated the actual practitioners from the announcements, according to the Burning Glass Institute follow-up work, was three things. First, the change had to be operationalized: ATS filters had to be reconfigured, job description templates had to be edited, hiring manager training had to happen. Companies that announced the policy without rebuilding the pipeline mostly kept hiring the way they had been. Second, leadership had to absorb the early friction. Skills-based hiring slows the first few cycles before it speeds them up, because internal stakeholders are evaluating new candidate categories. Third, the company had to have meaningful skill evidence to evaluate, which usually meant work samples, structured assessments, or a documented evaluation framework with consistent rubrics applied by trained reviewers.
Companies that announced skills-based hiring without one or more of those three pieces saw the press cycle move and the hiring data not.
The three working pieces
For a recruiter or hiring team actually running skills-based hiring in 2026, three pieces have to work together. These are the same three pieces that make modern proactive sourcing different from 2015-era Boolean LinkedIn work, which is not an accident: sourcing and skills-based hiring share most of their infrastructure.
A note on stage. The three pieces below operate at the sourcing layer, where the recruiter is deciding which candidates to identify, evaluate at a high level, and reach out to. They are not automated decision systems for ranking applicants who have already applied. The distinction matters legally. Applicant ranking by AI falls under NYC Local Law 144 (the Automated Employment Decision Tools rule), the Illinois Artificial Intelligence Video Interview Act, the EU AI Act's employment provisions, and EEOC guidance on AI in selection procedures. Skills-based hiring done well separates the sourcing layer, where AI widens and accelerates the search, from the selection layer, where human judgment is central and any automated assist is auditable for adverse impact. Conflating the two is the most common compliance mistake in skills-based hiring rollouts.
A skill graph that reads evidence. The keyword-match approach (search the resume for "Python", flag the matches) misses too much. A senior engineer with five years of unspecified backend work and a public GitHub full of Go code never appears for a Go role under keyword filtering. A skill graph reads the work, projects, conference talks, and code, and builds a weighted picture of what the candidate can actually do. The match becomes about evidence rather than about whether the right vocabulary was in the resume.
A compensation filter applied at the start of the pipeline. Skills-based hiring widens the candidate pool by removing the degree filter. That width matters only if the recruiter does not then spend the saved time talking to candidates whose compensation expectations the role cannot meet. A live, role-specific salary estimate attached to each surfaced candidate is the highest-precision filter most recruiting teams are not running. Most ATS platforms do not have the data infrastructure to do this, which is why companies that take skills-based hiring seriously usually augment their stack.
A predictive engagement signal. "Open to Work" on LinkedIn is self-reported, lagging, and noisy. A predictive signal that surfaces candidates likely to engage with a new opportunity now, even if they are not publicly searching, is what separates a 4% response rate on outreach from something closer to 15%. With a degree filter removed, the candidate pool widens by an order of magnitude. Without an engagement signal, that wider pool just means more cold outreach with the same low response rate.
These three pieces are what skills-based hiring looks like in practice, on a Tuesday, at the level of one recruiter filling one role. Skills matching as a category is the tooling layer underneath them.
Where companies trip up
Five failure modes show up consistently in the postmortem literature on skills-based hiring rollouts.
The first is internal mobility. A company that removes degree requirements on external hiring while still requiring degrees for internal promotions creates a two-class system that is visible to employees and corrosive to morale. The Burning Glass Institute research found this to be one of the most common reasons skills-based hiring announcements were quietly walked back within 18 months.
The second is manager pushback. Hiring managers who have spent careers evaluating candidates by school and title resist evaluating by skill evidence, particularly when they have to learn what counts as evidence. Without training and without leadership air cover, managers route around the new policy by adding "preferred" requirements that effectively restore the old filters.
The third is fake skills evidence. Once the system rewards skill claims, candidates inflate them. Skills-based hiring without an evaluation mechanism (assessments, work samples, references on specific skills) becomes worse than degree-based hiring, because at least the degree was verified by the institution.
The fourth is performance review misalignment. If the new hire is brought in on skills evidence but evaluated on the old proxies after they start, the company is signaling that the policy is a hiring filter, not a culture. Retention collapses.
The fifth is comp band misalignment. A candidate hired into a role traditionally requiring a degree, paid the same as their degreed peers, may be undervalued relative to the labor market for their actual skill stack, or overvalued if the role itself was overpriced. Companies that do not revisit comp bands when they revisit degree requirements end up with strange wage compression patterns. Glozo's compensation intelligence is built specifically for this problem: pricing roles to the skill stack rather than to the credential.
When NOT to do skills-based hiring
Three conditions make skills-based hiring a poor fit, and recruiters who run it indiscriminately get the worst of both worlds.
The role is heavily regulated. Some roles legally require credentials. A nurse, a CPA, a licensed clinical psychologist needs the credential because the state requires it. Skills-based hiring on these roles is illegal, not unwise.
The hiring volume is too low to justify the system cost. Skills-based hiring requires real infrastructure: an evaluation layer, a skill graph, manager training. A company hiring 3 senior engineers a year does not have the throughput to amortize that infrastructure. The right answer is good old-fashioned referral and high-touch sourcing.
The employer brand is not strong enough to attract non-traditional candidates. Skills-based hiring expands the candidate pool, but the candidates in that expansion still need a reason to take the call. A Series B startup that nobody outside the tech industry has heard of will struggle to convert non-degreed candidates from outside-the-funnel sources, because those candidates have less information about the company and less incentive to bet on it.
The honest takeaway
Skills-based hiring is real, it is not hype, and most companies that announce it are not doing it. The gap between the announcement and the practice is the most important fact in the space. For recruiters and hiring teams running it for real, the work is technical: rebuild the pipeline, train the managers, install the evaluation layer, watch the comp bands. For companies signaling it, the gap will close on its own eventually, because the data is now public.
If you want to see what skills-based hiring infrastructure looks like when it is built on a proper skill graph, live compensation data, and a real engagement signal, go to glozo.com and look at the product. For the sourcing techniques that pair with this approach, the candidate sourcing vs recruiting piece is the starting point, and the free resume search tools roundup covers the supporting toolkit.
The press cycle on skills-based hiring will continue. The question worth holding onto: does your company's hiring data look different than it did three years ago, or only your hiring policy?