Channel sheet · CH-09 · gain 5 min · logged November 14, 2026

AI MarketingDirect input

72% of Workers Say Nobody Has Ever Tested Their AI Skills

72% of workers say their AI skills were never formally tested, yet 69% of managers weigh AI ability in hiring — a gap PPC managers can close with their own account logs.

By Nathan Brooks5 min read905 words

Signal notes

  1. 72% of surveyed workers say their AI skills have never been formally evaluated, while 69% of managers treat AI ability as a hiring factor.
  2. Only 35% of workers say employer training built the AI skills they actually use; 46% received training they cannot fully apply.
  3. AI-required roles rose from 9.4% to 16.7% — a relative 78% jump, not percentage points.
  4. CodeSignal surveyed 1,000 US employees and 1,000 US managers via Pollfish in September 2026.
  5. 82% of employees say their manager understands their skills — a finding CodeSignal itself volunteered against its own business case.
72% Of Worker’s Say Their AI Skills Are Untested – Here’s How PPC Managers Prove Theirs
Input monitor72% Of Worker’s Say Their AI Skills Are Untested – Here’s How PPC Managers Prove Theirs — 6689062 / pixabay

72% of workers say their AI skills have never been formally evaluated, yet 69% of managers call AI ability a major or moderate hiring factor. Employers, in short, are hiring for a skill they mostly do not measure.

The numbers come from new research by CodeSignal, which surveyed 1,000 U.S. employees and 1,000 U.S. people managers through Pollfish in September 2026. Both samples were U.S.-only. The survey defined knowledge workers by the work they do, not job title, and did not screen out AI abstainers: 12% said they never use AI at work.

One caveat belongs up front. CodeSignal sells assessments, and its survey concludes that skills are not being assessed. The company acknowledged its interest openly and volunteered a counter-finding: 82% of employees said their manager understands their skills very or somewhat well. CodeSignal called that knowledge real but informal, uneven across managers, non-transferable between teams, and impossible to compare company-wide. The survey method was not described in the launch announcement for its Agentic Skills Intelligence product; the details came only in answers to follow-up questions.

How fast is AI use actually growing?

CodeSignal reports that roles where AI use is required or expected "jumped 78%." That is a relative increase, not percentage points. When respondents started their current jobs, 9.4% said AI use was required or expected. Today, 16.7% say so.

That is nearly double — and still fewer than one in five. Paid search sits well past that curve already, since automated bidding has been Google's default recommendation for years. The useful question is whether anyone can tell who uses AI well.

Does employer training actually stick?

Mostly no. 62% of employees said they received employer AI training, but only 35% said that training built the skills they actually use. CodeSignal confirmed the training was specifically about AI and reads the gap as a sign it was too thin to apply on the job.

The follow-up number is worse. Only about one in four trained workers can apply most or all of that training; the other 74% apply some, very little, or none of it. Across the full sample, roughly 46% of workers got AI training they cannot fully use. Employees named lack of time first, then training that does not fit their role, then lack of access.

So where do people learn? Multiple answers were allowed:

  • Free resources found on their own: 46%
  • Trial and error on the job: 39%
  • Employer training: 35%
  • Colleagues: 22%
  • Self-paid courses: 12%

"Companies are buying AI courses or building their own, but if most workers are going elsewhere to learn, you have to wonder if that training is actually helping," said Tigran Sloyan, CodeSignal's co-founder and CEO.

In paid search this pattern is old news. What changed is that the thing managers now learn by trial and error is the ad platform's own AI, and each error is paid for out of a client's budget.

Is AI the real measurement problem?

No. 66% of workers say their skills in general have never been formally evaluated — a six-point gap with the AI figure. Asked whether companies simply do not assess much of anything, CodeSignal answered, "Yes, exactly." More than six in ten managers admitted making people decisions on assumptions rather than reliable data.

The hiring side already shows the strain. When fluency is unmeasured, people get judged on visibility and confidence. The person who says the automation will figure it out can sound sharper than the person who spent the afternoon reading search terms and adding negatives. Only one of them is protecting the budget.

"A resume is a story someone tells about their skills," Sloyan said. "The work is the proof. Right now, most companies are making their biggest decisions about people based on the story."

CodeSignal separates the concepts this way: literacy means understanding what AI is, prompt-writing is a tactic, and fluency means knowing what to delegate, checking and correcting output, and recognizing when not to use AI at all. Mapped onto Google Ads, that is choosing which campaigns get Smart Bidding, reading the search terms and auto-created assets the system produces, and holding back a campaign with too little conversion data.

What can a certificate show?

CodeSignal admitted an assessment is only as useful as what it was designed and validated to measure. The same applies to Google's free AI-Powered Performance Ads Certification on Skillshop, a timed 50-question exam. It teaches Google's playbook. It cannot show the afternoon you decided not to follow it.

CodeSignal's advice for PPC managers is blunt: saying you use AI for everything proves nothing. Fluency is explaining exactly how you use it, and when you don't.

  • Keep an automation log. Change history records what changed, not why. Log every override, rejection, and result.
  • Bring the overrides to your review. Specific stories beat general claims.
  • Teach someone. Explaining your thinking makes the skill visible.

For team leads: start with managers, write down what good AI use looks like per role, check real accounts before and after training, and find your self-taught experts. "Training only works when it's built around the job someone actually does," Sloyan said.

PPC work already leaves a record. The managers who turn that record into evidence will get promoted while everyone else tells a story.

via searchenginejournal.com (Source)

Filed under

  • ai-skills-assessment
  • ai-training
  • ai-fluency
  • paid-search
  • talent-assessment
Share this article:

More from Nathan Brooks

Nathan Brooks

Show full bio

Correspondent covering regulation, policy and their impact on marketing professionals running AI in the stack at Mart Signal.

154 articles

Bus out

‹ Previous articleNext article ›