What screening discrepancies reveal about the workforce risk most companies can't see

Screening data is usually reported as a pass rate. A percentage of candidates cleared. A volume of checks completed. These are useful numbers for procurement and reporting, but they obscure the thing that matters most: what happens when a check comes back with a finding, and what it tells you about the risk sitting inside the hiring process.

Across the Veremark Screening Benchmark 2026 dataset, roughly 1 in 20 completed checks flagged a discrepancy. At a glance, that sounds manageable. The picture changes when you look at where those discrepancies concentrate. Education checks flagged roughly 1 in 5 candidates. Employment checks flagged around 1 in 7. CV gap checks flagged more than 1 in 2. Database checks (sanctions, criminal record, ID, financial history, right to work) flagged fewer than 1 in 100.

The overall average smooths out the extremes. The reality, for the checks that verify what candidates say about their own careers, is that discrepancies are common.

1 in 5 candidates have an academic credential that doesn't verify as stated

For every five candidates who go through academic verification, one has a finding. Across thousands of completed education checks, the same categories appear repeatedly.

• A candidate lists a degree from a recognised university. The university confirms enrolment but no award conferred. The candidate's CV says "BSc Computer Science." What's verifiable is two years of attendance and an incomplete programme.
• A candidate lists a qualification from an institution that, when contacted, has no record of them. The institution exists. The programme exists. The candidate was never enrolled.
• A candidate provides a graduation year two years earlier than the institution confirms. It could be a genuine error. It could be an attempt to conceal a gap between leaving education and starting work.
• A candidate lists a professional certification. The awarding body confirms it was issued but has since lapsed. The candidate didn't disclose that.

For a hiring manager, a 1-in-5 rate means that in a shortlist of ten candidates, statistically two of them have an academic credential that won't fully verify. For roles where qualifications are a regulatory requirement or a condition of practice (healthcare, financial services, engineering), that's a direct compliance exposure. For roles where they're a proxy for capability and credibility, it's a judgement being made on information that may be wrong.

1 in 7 candidates have an employment history their previous employer doesn't confirm

Employment checks contact previous employers and verify dates, title, and sometimes reason for leaving. At 1 in 7, findings are routine rather than exceptional.

• A candidate lists a role from 2019 to 2023. The employer confirms 2020 to 2022. A two-year tenure becomes four on the CV, one year added to each end.
• A candidate lists their title as "Head of Operations." The employer confirms "Operations Coordinator." The responsibilities described on the CV map more closely to the inflated title than the verified one.
• A candidate lists continuous employment across three consecutive roles. When the verified dates are mapped sequentially, a ten-month gap appears between roles two and three that didn't exist on the CV. The candidate adjusted the end date of one role and the start date of the next to close the gap on paper.
• A candidate lists a company that, when contacted, has no record of them. The company exists. The person was never employed there.

Each of these findings has a different severity. Adjusted dates may have a benign explanation. A fabricated employer doesn't. What they share is that none of them would have surfaced without direct verification. The CV presented a clean picture. The verified record told a different story.

For organisations making hiring decisions based on claimed experience, the risk is that seniority, competence, and track record are being assessed against a version of events that doesn't hold up when checked. A candidate hired into a leadership role based on four years of claimed experience may have two. A candidate trusted with client relationships based on a specific previous employer may never have worked there.

More than half of career timelines contain gaps the candidate never disclosed

CV gap analysis maps the full employment timeline against verified dates and flags where unexplained breaks of six months or more appear. At more than 1 in 2, it has the highest discrepancy rate of any check type in the dataset.

Some gaps have straightforward explanations. Parental leave, travel, study, redundancy. These become clear when the candidate is asked, and they rarely affect a hiring decision.

Others are the fingerprints of the adjustments described in the sections above. When a role is removed from a CV, it leaves a gap. When dates are adjusted to cover a period of unemployment, the verified timeline and the claimed timeline diverge. When a candidate has assembled a career narrative from selective truths, the seams show when the full picture is put together.

The reason this check flags more often than any other is that it examines the narrative layer of a career rather than individual claims within it. A candidate can have each listed role verified successfully and still have a timeline that doesn't hold together, because the issue isn't in the roles they included. It's in the ones they left out.

For senior hires and regulated roles, undisclosed gaps carry particular weight. A gap might conceal a termination, a period under investigation, a role at a competitor the candidate doesn't want to disclose, or a stretch of time that contradicts the progression narrative they've presented. The check doesn't determine which. It surfaces the inconsistency so that the hiring team can ask the question rather than accept the narrative at face value.

The checks that would catch this sit outside most standard programmes

Database checks account for 58% of all screening volume in the Veremark dataset. In Financial Services, 80%. These checks flag findings in fewer than 1 in 100 candidates. They serve a clear compliance function, confirming identity, checking registers, verifying right to work. When they do produce a finding, it's typically serious.

The checks that surface the discrepancies described above (education verification, employment verification, CV gap analysis) occupy the remaining share, and in many standard packages they're add-ons rather than defaults. They're triggered for senior roles or regulated positions, but the majority of hires may pass through without them.

The practical consequence is a screening process that confirms a candidate's identity and register status while leaving the most commonly inaccurate part of their application, their account of their own career, either lightly verified or unverified entirely.

What organisations are exposed to when verification isn't part of the process

The discrepancies described in this article aren't rare events surfaced by unusually thorough screening. They're what shows up at consistent rates when standard verification checks are actually conducted. 1 in 5 on education. 1 in 7 on employment. 1 in 2 on career timelines.

For organisations that include these checks, the findings trigger a review and a conversation. A discrepancy gets examined, context is gathered, and a decision is made with complete information. The hire might still proceed, but it proceeds with the organisation's eyes open.

For organisations that don't include these checks, the discrepancies still exist. They just go undetected. The candidate with the incomplete degree gets hired. The candidate who added two years to their tenure gets promoted based on experience they don't have. The candidate with the undisclosed gap fills a regulated role where the omission would matter if anyone knew about it.

The risk isn't theoretical. When 1 in 5 education checks and 1 in 7 employment checks produce findings across a dataset of this size, the pattern is structural. It reflects a consistent gap between what candidates present and what can be verified. Organisations that verify it can manage it. Organisations that don't are carrying the risk without knowing its size.

The data behind this article comes from the Veremark Screening Benchmark 2026. The full report covers discrepancy rates across all major check types, industry screening profiles, and a framework for assessing where a programme's detection gaps sit.

{{the-2026-veremark-screening-benchmark="/components"}}

‍