The Skills-First Promise Still Needs Proof
Employers agree that skills should count more than degrees. Three new studies show the hard part now: checking that a person can actually do the work.
Skills-first won the argument and lost the execution
The Testlify and SHRM Labs "Skills-First Talent Report 2026" surveyed 131 HR and talent leaders in July 2026. Its headline is blunt. Skills-first hiring won the debate but is losing in practice. Read the report
44.5% of organizations admit that nobody formally validates their assessments. Decisions run on gut feel. Only 8.4% are confident that their process picks the best candidate rather than the best interviewer. And 45.8% say the outcome depends on which hiring manager happens to run the process.
That last number matters for veterans. A service member's skills are real, but they rarely arrive in civilian wording. If the result depends on who is reading the resume, translation risk lands on the candidate. A validated, repeatable assessment moves that risk off the person and onto the process, where it belongs.
Games can be checked against the gold standard
If assessments need validation, what does validation look like? A peer-reviewed study in JMIR Serious Games offers a model. Researchers Boutrouille, Fichet and Dinet tested a serious game called Yuzu, built to assess soft skills in HR settings. Read the study
Yuzu has three modules. Active listening was compared with the AELS scale. Decision-making under uncertainty was compared with the Iowa Gambling Task. Team interaction style was compared with the SYMLOG model. Thirty-nine adults in France took part.
The listening module correlated strongly with AELS, with high agreement on a Bland-Altman check. The decision module was statistically equivalent to the Iowa Gambling Task and reproduced its exploration versus exploitation pattern.
The sample is small, so treat this as early validation. Still, the method is the point. A game-based measure was held against established psychometric tools, and it held up. That is the bar any simulation-based assessment should be asked to clear, including ours.
AI made real skills harder to see
The second annual WGU "Workforce Decoded" report, released September 30, surveyed 3,128 US hiring professionals. 60% say AI has made it harder to assess a candidate's real skills. The share who have not yet worked out how to evaluate AI skills doubled, from 16% to 32%, in a year. Almost one in five names the check of "is this a person, not an AI" as a challenge. Read the release
The cost shows up at the entry level. Among employers who say AI complicated assessment, 54% cut entry-level hiring. Among the rest, it was 20%. When employers cannot verify skills, they hire fewer people at the start of a career, and a veteran changing fields is effectively starting one.
WGU's president, Scott Pulsipher, called for assessments that confirm what a person can really do. Meanwhile, the value employers place on degrees rose from 68% to 73%, and on certificates from 86% to 91%. Credentials are not losing ground. The gap is in proving the skill behind them.
What this means for PowerTechs
For employers, the lesson is that a skills-first label is not a process. A process has a measure, a check against something trusted, and a record of how the person acted. Without those, hiring drifts back to the best interviewer.
For veterans, the lesson is hopeful. When a task is simulated and every action is recorded, a service member can show a skill directly. No job-title translation is needed, and no one has to guess from a resume line.
For PowerTechs, it confirms our direction. We build simulations that measure what people do, step by step, and we want that evidence to stand up to the same scrutiny the Yuzu team applied to theirs.
Takeaway
Everyone now agrees skills matter. The next advantage goes to whoever can prove them: validated, explainable and tied to real tasks.
For veterans: see roles that match your military experience For employers: book a 30-minute call