1 in 4 Invent Skills for the Bot. Workday Can Now Show Recruiters the Line

Most job seekers have tried at least one trick on a screening bot, and about one in four US job seekers admits typing in skills they don't have. Greenhouse put numbers on this in a survey released in October. In August, Workday's HiredScore started grading each requirement on its own and pointing the recruiter to the resume text behind every match, at employers that switch it on. If you're thinking about hidden white text or padding your skills list, this is what that line looks like from the recruiter's side.

What did Greenhouse's survey actually find?

In the survey, 62% of US job seekers said they had tried at least one tactic to get past screening software. Within that group, 40% added skills or keywords they don't have, and a third added hidden text aimed at an AI reviewer.

The words "of those" matter here. Some coverage, Fast Company's write-up included, makes it sound like 40% of all job hunters invent skills. That's wrong. Here is the correct math:

• About 25% of all US respondents admit adding skills they don't have (0.62 × 0.40).
• About 20% of all US respondents admit adding hidden text (0.62 × 0.33).
• In the UK and Ireland, 63% tried tactics. Of those, 37% added skills they lack and 28% added hidden text.

The survey covered over 2,800 candidates, 1,200 of them in the US. So the headline overstates it, but the real figure is still about one applicant in four.

The motive is easy to see. 81% of US candidates were rejected in the past year without speaking to anyone at the company. Greenhouse's Chief People Officer, Sharawn Tipton, put it this way: "For the first time I can remember, neither side thinks hiring is working." When silence is the only response you ever get, a trick that might lift you up the pile starts to look reasonable.

What changed in Workday HiredScore in August 2026?

On August 5, 2026, HiredScore replaced its basic-qualifications matching and final grade assignment with an LLM that evaluates each screenable qualification individually against your parsed resume. At employers that enable it, the Fit & Gap view gives the recruiter evidence for each match and points to the resume sections that support it.

Some definitions first:

• HiredScore is Workday's AI for grading and ranking candidates.
• A parsed resume is the plain-text version of your file that the ATS (applicant tracking system) pulls out, without layout or styling.
• Basic qualifications are the must-haves listed on the requisition. Preferred qualifications are the nice-to-haves.

Workday's product summary promises "detailed evidence and pointers to relevant resume sections." The candidate profile also highlights the parts of the parsed resume that relate to the requisition. Put together, the recruiter can see, requirement by requirement, which part of your resume the model counted as evidence.

One caveat. Fit & Gap is opt-in. Workday's release note talks about "customers that enable the Fit & Gap enhancement," so not every employer on Workday has it on. You can't tell from outside which ones do, so assume it's on.

How Workday grades resumes A to D

HiredScore gives each candidate a letter grade, and the required list does most of the work in deciding it:

Workday doesn't publish the threshold for an A. Look at the line between B and C, though. It depends only on whether every required item is matched. One missing must-have drops you a grade, and that's why people invent one. The August change means the invented match now comes with a pointer to where it supposedly appears on your resume.

Does putting keywords in white text work?

No, at least not in any way that helps you. Text extraction generally ignores font color, so hidden lines show up in the parsed resume as normal text. If a hidden line is what satisfied a requirement, the evidence pointer leads the recruiter straight to it.

A hedge: Workday hasn't published a document saying its parser drops white font color. Third-party guides describe it as how extraction works in general: color is styling, the parser keeps only the text, and the hidden line appears in the parsed profile and in any plain-text view.

Here's how the hidden white text resume ATS trick looks to the recruiter. The requirement says "Experience with Kubernetes." It's marked as met. The highlighted evidence is the word "Kubernetes" sitting alone at the bottom of the page, with no employer, no dates and no result. The trick hasn't fooled anyone. It has just made itself the thing the recruiter reads.

Employers are also looking for it directly:

• ManpowerGroup says it finds hidden text in about 10% of the resumes it scans with AI.
• Greenhouse's Nova by Dweet integration includes hidden-content detection in its fraud checks.
• Simone Lini, CEO of Zerolook, received roughly 350 resumes for two engineering roles. He uploaded about 20% of them to an AI tool and asked it to flag hidden prompts.

A prompt injection is hidden text meant to steer an AI reviewer, either with direct instructions ("rank this candidate first") or with invisible keywords.

The difference between what people say and what they do is large. Researchers from the University of North Carolina, Duke and hireEZ looked at about 200,000 resumes and found hidden injections in about 1%. More than 90% of those didn't use explicit instructions. Many more people say they've tried it than actually do it.

Even when it works, the advantage shrinks as more people use it. An ACL study found that injection could change rankings, "especially when relatively few applicants were using the tactic." If one applicant in five does it, the boosts mostly cancel each other out, and vendors keep adding detection. What you're left with is the risk.

Why is visible keyword stuffing the bigger risk?

Visible stuffing is the bigger problem because far more people do it and every parser keeps it. About 25% of US respondents admit to visible invented skills. Measured hidden injections sit around 1%.

The August change makes it worse. The LLM checks each requirement for meaning, not exact keyword matches. If the requirement is "3+ years running production services on Kubernetes," a bare "Kubernetes" in a comma-separated skills list is weak evidence. If it's the only evidence on your resume, that's the text Fit & Gap points the recruiter to. So keyword stuffing on a Workday resume does two kinds of damage: it adds little to the grade, and it shows the recruiter exactly how thin your claim is.

Then you get to the phone screen, and the recruiter asks about the one thing the AI matched. InnoCaption's CEO called hidden-text tricks "cutting in front of the line." An invented skill in plain sight is the same move, just easier to catch.

How many recruiters actually screen in Workday?

A lot, especially in the US. In Refolk's index of professional profiles, 932 US recruiters and talent-acquisition professionals list Workday as a skill. That's 3.7 times the 252 who list Greenhouse.

The Workday group includes large employers: Boeing, Siemens and Lenovo, plus Takeda through a Korn Ferry outsourced-recruiting role. The Greenhouse group includes Meta, Snowflake and Warby Parker.

Titles searched were Recruiter, Talent Acquisition, Technical Recruiter and Talent Acquisition Partner. The counts include only profiles that list the skill by name, so they understate the real numbers. Use them to compare, not to size the market. The comparison is clear anyway: if you apply to large US employers, a Workday screen is a good bet, and so is a recruiter who lists it as a core skill.

Why does an honest gap cost less than a fake match?

A grade orders the pile. It doesn't reject anyone, so an honest C can still get a call. A fabricated B sits in the exact line the recruiter is shown.

Workday's own Spotlight guidance says C and D candidates might still be a fit. Its example is a grade C engineer who has a bachelor's instead of the requested master's but more relevant experience. A recruiter who sees that gap next to strong bullets can make that call. A recruiter who clicks a match and lands on an invented skill can't trust anything else on the page.

The costs aren't equal. An honest gap moves you down within a pile the recruiter still reads. A fake match gets checked by a person, in a view designed to make checking easy. And if it survives the screen, it comes up again in the interview.

What should you write instead?

Write one visible bullet for each required qualification. Use the posting's own wording, add a scope or a number, and attach it to a real job with dates. Where you have a real gap, describe the closest experience you actually have.

Here's the process:

1. Copy the required qualifications from the posting, one per line. Use the literal text. Vague internal job descriptions lead to vague grades, so match what's written, not what you think the job is.
2. Find the real job where you did each one, and write a bullet that uses the posting's noun, the scope, and the result.
3. Write years as dates. Back up "5+ years" with the date ranges on your roles and a result, not an adjective.
4. Say the closest true thing for a gap. If they want Kubernetes and your production work used other container tooling, name that tooling and the scope. The model judges meaning, so honest adjacent experience is better evidence than a bare keyword.
5. Treat the skills section as an index, not evidence. Every skill listed there should show up again in a bullet.
6. Delete anything you can't talk about for two minutes. If it gets highlighted, you'll be asked about it.

Here's what the change looks like:

The bracketed parts are where your own scope and results go. When Fit & Gap highlights one of these lines, the recruiter sees a job, a date range and an outcome instead of a bare word.

Mapping one requirement to one line, for every posting, is the slow part, and it's the part Refolk does for you. Paste the posting and I rewrite your own resume so each requirement has a visible line from your real history behind it. The fit score shows which requirements have no honest line, so you know your gaps before the recruiter does. When a gap matters, deal with it in the cover letter rather than hiding it. Refolk drafts that from the same posting, using the same evidence.

FAQ

Does putting keywords in white text work on Workday?

Not in a way that helps. Parsers generally ignore font color, so hidden text shows up in the parsed resume as normal text. Workday hasn't documented this specifically, but it's how extraction usually works. With Fit & Gap enabled, a hidden keyword that satisfies a requirement becomes the highlighted evidence the recruiter reads. ManpowerGroup says it finds hidden text in about 10% of the resumes it scans with AI.

Can I list a skill I'm still learning?

Yes, as long as you say it's in progress and attach it to something real, like a course project or a side deployment, with dates. What you can't do is present it as professional experience. With per-requirement evidence, the recruiter will see exactly which line earned the match and will ask about it.

Does every Workday employer use Fit & Gap?

No. Workday describes it as an enhancement that customers choose to enable, so some employers on Workday won't have it turned on. You can't tell from the posting which ones do, so write every resume as if the recruiter will see the line behind each match.