State of Personality Testing 2027

Executive Summary

This report analyzes the state of personality and career assessment adoption, usage patterns, and validity across the United States and United Kingdom in 2026–2027. Drawing on JobCannon's own platform data (approximately 171,000 test completions in the 30 days to 22 August 2026), cross-referencing published surveys from SHRM and CIPD, the Pew Research Center, and the peer-reviewed literature, we map how professionals and organizations use personality tests for hiring, self-discovery, and career navigation.

The findings point to three trends. First, employers use assessments and AI widely but far from universally: in SHRM's 2024 Talent Trends survey, 54% of organizations reported using pre-employment assessments (14% of those users named personality tests), and SHRM's 2025 Talent Trends survey put AI adoption in HR tasks at 43%, up from 26% in 2024 (SHRM 2025). Second, the gap between the best-known tests and the best-validated tests is wide. The MBTI remains the culturally dominant personality framework even though its own publisher says it should not be used for selection, while the Big Five—the most researched framework in occupational psychology—is mostly met in research and professional assessment. Third, regulatory pressure from the EU AI Act, NYC Local Law 144, and emerging state-level transparency laws is reshaping how vendors build and document hiring tools.

Three findings anchor this section: (1) no public source gives a total for personality assessments taken each year in the US and UK, so this report does not state one; (2) the best survey evidence we found puts personality-questionnaire use at 20% of UK organizations that recruited in the past year (CIPD, 2026) and shows 54% of surveyed organizations using some pre-employment assessment (SHRM, 2024); and (3) the most widely recognized personality framework, the MBTI, is one its publisher advises against using for selection.

Methodology

Data Sources

JobCannon proprietary data: Aggregate test completions from our platform, anonymized at the group level. Our first stored result is dated 22 February 2026, so the window is February to August 2026 and no longer. In the 30 days to 22 August 2026 the platform recorded approximately 171,000 completed tests, and approximately 284,000 since launch, drawn from a catalogue of 205 assessments. The frameworks named in this report — Big Five (50-item), MBTI-style typology (16-type), Career Match (proprietary algorithm), RIASEC (6-theme), Enneagram (9-type) — sit inside that catalogue alongside the rest of our personality, skills, and relationship tests. Note: JobCannon serves primarily self-directed users and mid-market B2B clients. Our user base is global and majority non-US; it is not nationally representative of either country this report covers and should not be generalized as such.

Public data sources:

• SHRM Talent Trends (2024, 2025) and SHRM Benchmarking (2025): Surveys of HR professionals and SHRM members covering assessment use, AI use in HR, cost-per-hire, and time-to-fill.
• CIPD Resourcing and Talent Planning reports (2024, 2026): UK employer surveys, including which selection methods organizations use.
• Pew Research Center, AI in Hiring and Evaluating Workers: What Americans Think (2023): Survey of US adults on attitudes toward AI in hiring. Pew does not measure personality-assessment use.
• BLS Job Openings and Labor Turnover Survey (JOLTS): A monthly survey of about 21,000 establishments measuring job openings, hires, and separations. It does not measure applications or screening, so it supplies labor-market context only.
• EEOC charge statistics: The EEOC does not publish a separate count of AI-related charges.
• NBER Working Papers and peer-reviewed journals: Field experiments on algorithmic hiring outcomes, resume screening bias, and assessment validity.
• EU Commission and NYC Department of Consumer Affairs: Regulatory text and compliance data on hiring AI audits and bias testing mandates.

What We Measure

This section reviews: (1) what published surveys show about assessment use in hiring (SHRM for the US, CIPD for the UK); (2) the assessment volumes that vendors report about themselves; (3) the vendor landscape; (4) the regulatory and validity landscape. We do not claim causal relationships from observational data.

What We Do Not Claim

This report is limited to the US and UK markets. Data on other English-speaking countries (Canada, Australia) is sparse and excluded. We do not infer predictive validity of any single framework—that requires longitudinal workplace outcome data beyond our scope. We do not estimate a total annual volume of assessments or a market share by framework, because no published source supports one. We do not extrapolate from JobCannon usage to global test-taker populations; our platform skews toward self-directed users and may not reflect enterprise hiring patterns. We do not make policy recommendations; we present regulatory facts and industry trends for context only.

We also do not report demographics for our own users, because we do not collect them. JobCannon stores no education field at all; age bracket and career stage are optional profile fields that fewer than one in two hundred accounts have filled, which is far too few to describe a user base with. Self-declared country is present on roughly two in five accounts and is a property of registered profiles rather than of test completions, most of which are taken without an account — so we do not publish a country split for completions either. Where an earlier version of this report gave figures for age, education, career-change intent or geographic mix, those figures were not measured and have been removed rather than restated.

How Many People Are Taking Personality Tests in 2026–2027?

Test-Taking Volume: What Is Published

No public source gives a total for personality or career assessments taken each year in the US and UK, and this report does not state one. An earlier version of this section gave an estimate of 65–75 million a year; it was built from assumptions we could not source (the share of employers screening with assessments, the number of assessments per hire, and the volume of coaching platforms), so it has been removed. What exists instead are volumes that individual vendors report about themselves. They are self-reported, cover different kinds of assessment and different geographies, and cannot be added together:

• SHL: "45 million SHL assessments are taken every year, around the world" (shl.com). This covers all SHL assessment types, not only personality.
• Hogan: "One million personality assessments are processed every year" (Hogan FAQ, hoganassessments.com).
• Myers & Briggs Foundation: "Over 2 million people take the MBTI tool every year" (myersbriggs.org).
• 16Personalities: its free test "has been taken more than 1.5 billion times" in total, as of June 11, 2026 (16personalities.com, page for AI assistants). That is a cumulative worldwide count, not an annual one.

We do not state JobCannon's share of any of these, and the omission is deliberate. Our platform has six months of history rather than a year, so any annual figure for it would be an extrapolation from a steeply rising run rate rather than a measurement. More importantly, none of the figures above is a US-and-UK total while our completions are majority neither, so dividing one by the other would not be a market share — it would be two different populations in a single fraction.

Survey Evidence: Assessment Use in Hiring

The most direct published evidence on how many employers use assessments comes from employer and HR-professional surveys. Key findings:

• Assessment use (US): In SHRM's 2024 Talent Trends survey of 2,366 HR professionals (fielded 10–19 January 2024), "54% of organizations use pre-employment assessments." Among those that do, 14% named personality tests (respondents could choose more than one type).
• Why some employers do not use them: In the same survey, organizations without assessments cited doubts that assessments measure the needed skills well (38%), return on investment (36%), bias or discrimination concerns (36%), and lack of resources (36%); 19% cited potential negative applicant reactions. Separately, 36% of HR professionals who use pre-employment assessments said their time-to-fill had worsened because of them.
• AI in HR (US): SHRM's 2025 Talent Trends survey of 2,040 HR professionals reported AI adoption in HR tasks at 43%, up from 26% in 2024. SHRM's AI in HR 2026 report (1,908 HR professionals, published 8 April 2026) says 39% of organizations are currently using AI in their HR functions. The two are different surveys with different bases, so they should not be read as a downward trend.
• Cost-per-hire and quality of hire (US): SHRM's 2025 Benchmarking reports (2,371 members responded between 9 January and 3 March 2025) give an average cost-per-hire of $5,475 for nonexecutive roles and $35,879 for executive roles, and note that only 20% of organizations track the quality of hire.
• Time-to-fill (US): SHRM's 2026 Recruiting Executives Benchmarking data brief reports a median of 39 calendar days for nonexecutive positions. SHRM measures time-to-fill (requisition opened to offer accepted), not time-to-hire.

Read together, these figures show that assessments are common but far from universal, and that personality tests are a minority of what employers use. We found no published source for the share of applications that are screened by humans rather than by software, or for the number of candidates screened per hire, so this report does not state either.

UK Evidence: CIPD

The CIPD's Resourcing and Talent Planning reports are the main UK benchmark. The 2024 edition (September 2024; over 1,000 HR and people professionals, n=1,016) found that around three-fifths of organizations (61%) use some form of test or assessment, and 18% used personality or psychological questionnaires (17% in 2022 and 22% in 2021, from the same chart). The 2026 edition (September 2026; n=791 organizations that had recruited in the previous 12 months) found that over three-fifths (63%) use some form of test or assessment, and 20% used personality or psychological questionnaires. "Some form of test or assessment" includes work samples, assessment centres, group exercises and reasoning tests, not only personality measures. The two editions use different bases (all respondents in 2024, only organizations that recruited in 2026), so the change from 18% to 20% is not a like-for-like trend.

On the candidate side, Criteria's 2025 Candidate Experience Report (more than 2,000 candidates worldwide) found that 86% of job seekers either like or feel neutral about taking assessments when applying for jobs.

We found no published breakdown of personality-test use by sector, company size, UK region or remote-work policy that we are willing to cite. An earlier version of this section gave adoption percentages for each of those splits; none could be traced to a source and they have been removed.

Which Types of Tests Are in Use?

Five Frameworks Compared

We found no published source that quantifies market share by personality framework, so this report does not rank frameworks by share. An earlier version of this section gave a percentage range for each framework; none could be traced to a source and they have been removed. The table below compares the frameworks on use and evidence instead. It is not drawn from JobCannon usage: what a single platform's visitors choose measures that platform's catalogue and traffic mix, not the world's.

The MBTI Paradox

The MBTI is among the best-known personality frameworks: the Myers & Briggs Foundation says over 2 million people take it every year, while the Myers-Briggs Company says as many as 1.5 million assessments are administered annually (both self-reported). Yet its publisher advises against using it for hiring. The Myers-Briggs Company states that the assessment "should never be used in recruitment or selection because it does not measure a person's skills or abilities," and the Myers & Briggs Foundation says it "is not designed to be used for hiring purposes." Pittenger (2005, Consulting Psychology Journal: Practice and Research 57(3), 210–221) concluded that the MBTI "may not yet be able to support the claims its promoters make." We found no peer-reviewed meta-analysis estimating the correlation between MBTI type and job performance; the meta-analyses we could verify address reliability (Capraro & Capraro, 2002; Randall et al., 2017). An earlier version of this report cited a meta-analysis reporting near-zero validity; we could not locate it, so the claim has been removed.

This gap between recognition and validity is not unique to the MBTI, but it is most pronounced there. The Big Five, by contrast, has the largest research literature in occupational psychology: Barrick and Mount (1991) found conscientiousness the most consistent Big Five predictor of job performance (mean corrected validity .22), and later syntheses put it near .2 (Hurtz & Donovan, 2000; Zell & Lesick, 2022; see Section 3). We have no data on the relative consumer reach of the two frameworks.

Vendor Consolidation and AI Integration

Assessment vendors have been changing hands. Harver announced the completion of its acquisition of pymetrics on August 11, 2022; Phenom announced its acquisition of Plum on April 28, 2026; and Crosschq announced its acquisition of Traitify on March 31, 2026. None of these announcements disclosed terms. SHL has been through several owners: CEB bought it in 2012, Gartner acquired CEB in 2017, and Gartner sold the business to Exponent Private Equity for $400 million in a sale completed on April 3, 2018 (Gartner SEC filing). Gartner also publishes a Magic Quadrant for Talent Acquisition (Recruiting) Suites, issued in April 2025, which positions vendors of recruiting software; we cite no figures from it.

AI-based analysis of text, voice and video is also changing how candidates are screened. Regulators are paying attention: NYC Local Law 144 and the EU AI Act both regulate automated tools used in hiring (see Section 2). HireVue announced in January 2021 that it was dropping facial analysis from its assessments (Fortune, 19 January 2021).

Who's Taking Them and Why

Test-Taker Motivations: Three Channels

Personality test engagement falls into three overlapping channels. We have no published data on how much of the total each channel accounts for, or on typical completion times and dropout rates in each, so none is given; an earlier version of this section stated percentages for all three, and they have been removed.

1. Employer-administered screening: Job applicants encounter assessments as a step in hiring funnels. The one candidate-side figure we could verify is attitudinal: in Criteria's 2025 Candidate Experience Report (more than 2,000 candidates worldwide), 86% of job seekers either like or feel neutral about taking assessments when applying for jobs.

2. Self-directed career exploration: Professionals pay or volunteer their time to take assessments, typically driven by career uncertainty, transition planning, or personal curiosity. JobCannon's platform sits in this segment: the overwhelming majority of our completions are self-directed rather than employer-administered. We do not describe those users by age, education or intent, for the reason given under What We Do Not Claim.

3. Corporate wellness and team-building: HR departments administer assessments as team workshops, leadership development, or employee engagement initiatives. Participation is typically voluntary, and results are often shared in group settings. We found no published outcome data on whether such workshops change retention or collaboration.

Demographics and Patterns

We do not publish a demographic profile of self-directed test-takers. We have no third-party dataset that breaks that population down by sex, age, education or occupation, and we will not assemble one from our own traffic for the reasons given under What We Do Not Claim. An earlier version of this section gave a split by all four; it was not measured, and it is gone rather than restated.

What we can say about our own mix is the channel, because it is recorded on the result row instead of being self-declared: fewer than one in five hundred of our stored results carries an employer marker — a partner-API run, a hosted invite, or a cohort batch — so our completions are almost entirely self-directed, as of 22 August 2026. That is a fact about this platform's traffic mix, not about who self-directed test-takers are in general, and the two should not be read as the same claim.

The same applies to the two breakdowns this section used to carry underneath: assessment exposure by job category, and by US and UK region. Both were stated as percentage ranges to the point, neither was traceable to any source in the list above, and an occupation split in particular would have contradicted the paragraph directly before it. Vendor positioning does appear to concentrate assessment use in people-facing hiring — that is visible in what the vendors themselves market — but we have no exposure rate by occupation or by metro that we are willing to print next to a citation block, so none is printed.

Non-English-speaking markets are under-represented in every major platform, including this one. That gap is acknowledged throughout this report and is a limit on its scope rather than a finding within it.

Conclusion: Framing the Landscape

No public source tells us how many people take personality assessments each year, which is itself a finding. What the published evidence does show is a split between brand recognition (the MBTI's cultural reach) and scientific support (the Big Five's research backing). That gap is not new, and regulation of algorithmic hiring tools is putting pressure on vendors to document validity and bias testing.

The best-known tests are not the best-validated ones. The MBTI, whose publisher says it should not be used for selection, has no meta-analytic evidence of predicting job performance that we could verify. The Big Five is the most researched framework in occupational psychology. Career-matching frameworks, including JobCannon's own algorithm, depend on the quality of the data behind them, and we found no peer-reviewed meta-analysis of them.

The following sections of this report cover the regulatory and vendor landscape (Section 2), which assessment methods predict job performance and what we can and cannot say about return on investment (Section 3), the evidence on bias and what audits can detect (Section 4), what employers and candidates say stands in the way of wider adoption (Section 5), and where the rules and the research are heading (Section 6).