Skills Test Score Benchmarks: What Job Applicants Actually Score
Based on 56,900 test administrations by EmployTest, August 2021 – August 2026
Most hiring managers set a passing score without knowing what a normal score looks like. A candidate gets 68% on an Excel test and there’s no reference point — is that strong, average, or a problem?
This page answers that. Every figure below comes from EmployTest’s own assessment platform: real job applicants, tested by real employers, during actual hiring processes. Not a survey, not self-reported.
How many job applicants can’t actually do what their resume says?
Across 24,278 skills and knowledge test administrations, 1 in 4 applicants scored below 70%, 1 in 7 scored below 60%, and 1 in 14 scored below 50%.
Based on 46 assessment titles covering Microsoft Office, computer literacy, accounting, attention to detail, communication, and general workplace skills, 2021–2026.
A note on these thresholds. 70%, 60%, and 50% are reference points, not passing scores. We publish them because they’re familiar — most people carry a rough sense from school of what those numbers mean — not because we consider any of them a standard.
On software and Microsoft Office tests specifically, the rate rises to 1 in 3. Across 12,896 application skills tests, 33% of applicants scored below 70%, 19% below 60%, and 10% below 50%. These are the skills candidates state most plainly on a resume — “proficient in Excel,” “advanced Microsoft Office” — and the ones most often taken at face value in an interview.
We don’t set cut scores, and no assessment vendor should. A cut score is a judgment about what a specific job requires, made by the employer who knows the role: how much of the work depends on the tested skill, how much training a new hire will get, how deep the applicant pool is, and what the cost of getting it wrong looks like. The right threshold for a financial analyst who lives in spreadsheets is not the right threshold for a coordinator who opens one twice a week. Both may be entirely defensible.
What benchmark data can do is tell you what a threshold will cost you. If you require 70%, roughly a quarter of your applicants won’t clear it — a third on software tests. That’s information for a hiring decision, not the decision itself.
Where the gaps concentrate. Rates vary widely by test. About 1 in 5 applicants scores below 60% on a combined Word-Excel-Outlook test, and roughly 1 in 3 falls below 60% on advanced Excel. Focused single-skill tests run much lower: 11% on intermediate Word, 8% on attention to detail, 6% on basic employment skills.
Breadth is where candidates overstate. Someone can genuinely know Word well, be described as “proficient in Microsoft Office,” and still be unable to do the Excel half of the job. A combined test surfaces that; a conversation usually doesn’t.
This is the argument for testing at all. A resume line reading “proficient in Excel” or “strong attention to detail” carries no information about where a candidate falls in these distributions. Neither does an interview conversation about it. A 20-minute assessment does — and then you decide what the number has to be.
What is a good score on an Excel test for a job applicant?
The average score on an intermediate Excel test is 74%, and the median is 78%. Half of all applicants score between 64% and 86%.
The gap between the average and the median matters. A minority of very low scorers pulls the average down, so 74% is not the typical applicant — 78% is closer. If you’re setting a cutoff, anchor it to the median, not the mean.
Difficulty level changes the picture substantially:
| Excel test level | Average | Median | Middle 50% of applicants |
|---|---|---|---|
| Beginner | 78% | 81% | 72–91% |
| Intermediate | 74% | 78% | 64–86% |
| Advanced | 69% | 73% | 56–85% |
What do applicants score on a combined Microsoft Office test?
Multi-application Office tests average 71–75%, slightly below what the same candidates score on single-application tests. Covering Word, Excel, and Outlook in one assessment produces a wider spread of scores than testing any one program alone.
| Test | Average | Median | Middle 50% | Share scoring 70%+ |
|---|---|---|---|---|
| Word-Excel Core Skills | 75% | 80% | 67–87% | 70% |
| Microsoft Office Test 365 | 73% | 75% | 64–86% | 62% |
| Word-Excel-Outlook combined | 72% | 77% | 60–87% | 65% |
The wider middle band on the combined test — 60% to 87%, versus 64% to 86% on Excel alone — reflects that few candidates are equally strong across all three programs. A combined score of 72% can mean solid competence throughout, or strong Word paired with weak Excel. If the role depends on one program in particular, a single-application test gives you a clearer answer.
What score should I require on a pre-employment test?
It depends far more on the type of test than most hiring managers expect. A 70% cutoff is reasonable on a software skills test and nearly disqualifying on an aptitude test.
Here’s what actually happens to your applicant pool at a 70% bar:
| Test | Average score | Share of applicants scoring 70% or higher |
|---|---|---|
| Accounting — Accountant Knowledge | 81% | 85% |
| Basic Employment Skills | 81% | 84% |
| Word — Intermediate | 78% | 81% |
| Attention to Detail | 82% | 80% |
| Computer Literacy | 78% | 79% |
| Excel — Beginner | 78% | 76% |
| Word-Excel Core Skills | 75% | 70% |
| Outlook — Intermediate | 73% | 68% |
| Excel — Intermediate | 74% | 66% |
| Accounting — Bookkeeper | 73% | 65% |
| Word-Excel-Outlook combined | 72% | 65% |
| Microsoft Office Test 365 | 73% | 62% |
| Grammar | 71% | 61% |
| Excel — Advanced | 69% | 56% |
| Accounts Payable | 68% | 54% |
| Cognitive ability | 51% | 16% |
| Workplace aptitude | 41% | 8% |
The practical takeaway: applying one universal cutoff across a mixed assessment battery will reject almost every candidate on the aptitude portion. Set thresholds per test, not per candidate.
One process note worth more than it sounds: decide your cutoff before you review candidates. Setting the bar after you’ve met someone is how personal preference re-enters a process meant to reduce it.
What is the average score on the Elite Cognitive ability test?
Applicants answer about half the questions correctly on average — 51%. Half score between 40% and 62%, and only about 1 in 10 answers three-quarters or more correctly.
| Share of questions answered correctly | Share of applicants reaching it |
|---|---|
| Half or more | 52% |
| 60% or more | 27% |
| 75% or more | 10% |
| 95% or more | 0.2% |
The Workplace aptitude test runs harder still: 41% average, with only 8% of applicants answering 70% or more correctly (552 administrations).
If those numbers look alarming, they shouldn’t. Cognitive and aptitude tests are built to spread candidates out across the full range, which means the average lands near the middle by design. A skills test asks whether you can do something; an aptitude test asks how quickly you work through unfamiliar problems. They are not scored to the same standard and should never be compared to each other.
Why do candidates score so much lower on aptitude tests than on skills tests?
Because the two test types are designed for different jobs. Skills and knowledge tests measure whether someone has learned a specific thing — Excel formulas, accounts payable procedures, grammar rules. Most candidates who apply for those roles have learned most of it, so scores cluster high: averages of 73–82%, with medians higher still.
Aptitude and cognitive tests measure reasoning speed on unfamiliar material. They’re built so scores spread across the entire range, which puts the average near the middle: 51% and 41% in our data.
A candidate who scores 78% on Excel and 55% on cognitive ability hasn’t failed the second test. They’ve scored above average on both.
How long does skills testing add to my hiring process?
Most tests take 20 to 30 minutes. Median completion times from actual applicants:
| Test | Median time | Middle 50% |
|---|---|---|
| Attention to Detail | 6 min | 5–7 min |
| Grammar | 10 min | 7–12 min |
| Excel — Beginner | 18 min | 14–25 min |
| Basic Employment Skills | 18 min | 15–21 min |
| Computer Literacy | 20 min | 16–27 min |
| Accounting — Accountant Knowledge | 22 min | 16–29 min |
| Word — Intermediate | 23 min | 18–31 min |
| Word-Excel-Outlook combined | 25 min | 19–34 min |
| Outlook — Intermediate | 26 min | 20–34 min |
| Excel — Intermediate | 29 min | 22–40 min |
| Excel — Advanced | 32 min | 24–42 min |
| Accounting — Bookkeeper | 37 min | 28–49 min |
Since candidates test on their own time before an interview, testing typically removes time from the hiring process rather than adding it — a 30-minute assessment replaces the portion of an interview spent trying to gauge skills through conversation.
Do timed tests lower candidate scores?
Yes — and the effect is far larger on harder tests. Comparing the same Excel content with and without a 30-minute limit:
| Test | Untimed average | Timed average | Difference |
|---|---|---|---|
| Excel — Beginner | 78% | 72% | −6 points |
| Excel — Intermediate | 73% | 67% | −5 points |
| Excel — Advanced | 68% | 56% | −12 points |
Beginners lose about six points to the clock. Advanced candidates lose twice that. The harder the material, the more time pressure costs.
Two practical consequences. If your role genuinely involves working under deadline, a timed test measures something real and you should use one. If it doesn’t, an untimed test gives you a cleaner read on capability. And when comparing a candidate’s score against any published benchmark, confirm you’re comparing the same format — a timed score judged against untimed norms will make a competent candidate look weak.
Are workers’ computer skills getting worse?
Not in our data. Average scores on intermediate Excel tests have been essentially flat for five years:
| Year | Average | Median |
|---|---|---|
| 2022 | 73% | 78% |
| 2023 | 74% | 78% |
| 2024 | 73% | 78% |
| 2025 | 75% | 81% |
| 2026 | 73% | 81% |
Year-to-year movement is within normal variation. Applicant Excel ability has not measurably declined — or improved.
What percentage of applicants pass a skills test?
There’s no universal pass rate, because there’s no universal passing score. But at a 70% threshold, pass rates run from 85% on accounting knowledge down to 54% on accounts payable among skills and knowledge tests — and drop to 8–16% on cognitive and aptitude tests.
A more useful way to think about it: a cutoff that passes 80% of applicants is a screen against clear gaps. A cutoff that passes 50% is a ranking tool. Decide which one you’re building before you pick a number.
About this data
These figures come from EmployTest’s assessment platform: 56,900 completed test administrations across 396 assessment titles between August 2021 and August 2026.
They represent job applicants who reached the assessment stage of hiring at organizations that chose to use skills testing. They are not a representative sample of the U.S. workforce or of applicants generally.
We publish no figure based on fewer than 100 test administrations. Some assessments report percentage of items answered correctly; others report a percentile rank against a norm group — we label which throughout, and scores on different scales are not directly comparable.
EmployTest sells the assessments this data comes from. We publish it because we think it’s useful, and we document our methods so you can judge it accordingly.
Media and research inquiries: [email protected] — we’re happy to provide additional cuts of this data on request.
Last updated: August 10, 2026