A resume lists a person’s employment history and the job they claim to have done there. It cannot tell you if a candidate can handle the actual work. Years of experience turn out to be one of the weakest predictors of performance, so the candidates who look best on paper aren’t reliably the ones who perform best. Before you make an offer, test the two or three abilities that are necessary for the position.
Table of Contents
Most hiring managers have seen some version of this. The candidate had eight years in the field, two recognizable company names and a clean promotion history. The interviews went well. Everyone felt good about the offer.
Six weeks later, the new hire still needs help with work the team assumed they’d handle on day one.
When candidates with strong resumes can’t do the job, it’s easy to file it under “bad fit” and move on. But if it keeps happening, the problem probably isn’t the people you’re hiring. It’s what you’re using to judge them, and it’s the reason so many teams have started testing skills before they hire.
Key Takeaways
- A resume shows where someone has been. It can’t show what they’re able to do today.
- Years of experience and job titles are weak predictors of how well someone will perform in a new role.
- Casual, conversational interviews tend to reward confident storytelling more than job skill.
- Work samples, job knowledge tests and structured interviews give much stronger evidence of ability.
- Pick the few resume claims the role depends on and test them before making an offer.
Why do strong resumes fail to predict job performance?
A strong-looking resume does not always mean a candidate can do the job well. Three common resume factors can create the biggest problems, yet they tell you very little about the job-critical skills a candidate can actually use.
Experience is a measure of time
In 2019, a team led by Chad Van Iddekinge at Florida State University pulled together 81 studies on pre-hire work experience. The correlation with job performance came out at .06, on a scale where 1.0 would mean perfect prediction. With turnover, it was zero. Even experience in closely related roles only reached .07.
Ten years in a role tells you almost nothing about how someone will do in yours. Some people grow every year. Others coast. Either way, the resume reads the same.
Degrees and titles are shortcuts
Hiring teams lean on degrees and brand-name employers because they’re quick to scan. They’re also loose proxies for the actual work. A “Senior Analyst” at one company might be doing junior-level tasks at another.
Plenty of employers say they’ve moved past this. The data says otherwise. A 2024 study by Harvard Business School and the Burning Glass Institute found that dropping degree requirements changed the outcome for fewer than 1 in 700 hires in 2023. Where companies really did hire on skills, though, workers without degrees stayed longer. Their retention was 10 percentage points higher than their degree-holding colleagues.
Candidates control the story
A resume works like a personal sales pitch, and AI has made that pitch easier to polish. It can even help someone create a realistic-looking fake resume, one of several reasons resumes can mislead hiring teams. The subtler problem is that even a completely honest resume can’t show how well someone did the work it describes.
“Led a CRM migration” could imply that they participated in the weekly calls or that they created it. On paper, both read the same.
Want to know if candidates rely on AI during remote tests? Read our article on spotting AI use in remote assessments.
Why don’t interviews catch it?
You’d think the interview would correct an inflated resume. A lot of the time, it reinforces it.
Most interviews are informal. The interviewer skims the resume, asks the candidate to walk through their background and goes with a gut feel at the end. That format favors people who are good at talking about their work. Some of them are skilled at it as well. Many aren’t.
The research backs this up. In a large 2022 re-analysis of selection methods, Paul Sackett and colleagues put the validity of unstructured interviews at .19. Structured interviews, where every candidate gets the same job-related questions and the same scoring rubric, came in at .42. That gap is a big part of why AI interviewing tools are built around fixed questions and shared rubrics.
There’s a bias problem too. Once an interviewer has seen a well-known employer or an impressive title, they tend to hear every answer more kindly. That’s the halo effect, a textbook case of unconscious bias in hiring, and it means the resume ends up grading itself.
What actually predicts whether a candidate can handle the job?
Methods that make candidates show their skills instead of describing them. This is how common approaches compare in Sackett’s 2022 estimates (higher means a stronger link to later performance):
| Selection method | Mean validity | What it tells you |
| Structured interviews | .42 | How candidates handle job-specific situations, scored the same way for everyone |
| Job knowledge tests | .40 | Whether they know what the role requires |
| Work sample tests | .33 | Whether they can do real tasks from the job |
| Cognitive ability tests | .31 | How quickly they pick up and solve new problems |
| Unstructured interviews | .19 | Mostly how well they present themselves |
| Years of job experience | .07 | How long they’ve held similar roles |
The two things most hiring processes rely on, resume history and an informal chat, sit at the bottom. Most of what sits near the top asks candidates to show what they know or can do. Even structured interviews, which still rely on what candidates say, work better because they hold every answer to the same standard.
Resumes still have a job to do. They’re useful for eligibility, context and career history. They just aren’t proof.
How do you test for real ability before you hire?
Here’s what that looks like in practice. Where these checks sit in your hiring funnel matters too, and we cover that in why unqualified candidates still reach your interview stage.
Pick the claims that matter
Focus on the skills and experience that matter most for the job. For example, a candidate might mention building Excel models or leading a team of eight people. Check whether they can back up those claims with real examples. If you’re reviewing hundreds of resumes, resume parsing can collect the skills candidates list into a single format, helping you compare their qualifications without going through every resume manually.
Almasons, a company that builds SAP supply chain products, hit this wall with technical hiring. They had a big applicant pool, plenty of plausible profiles and no quick way to tell who had the depth. What helped was putting technical and psychometric tests in front of the interview, so the shortlist was built on how people performed rather than how they described themselves.
Turn each claim into a short task
The best way to check a skill is often to see it in action. Instead of simply taking an Excel claim at face value, give the candidate a cluttered spreadsheet and ask them to solve a practical business problem. For a developer, a small codebase with a bug can reveal how they actually troubleshoot and write code. If the job involves dealing with customers, use a realistic situation and ask the candidate to explain how they would respond.
Job simulations bring these real-world tasks into the hiring process. They give recruiters a much better idea of how someone could perform on the job before making a hiring decision.
If your engineering screen still leans on puzzle-style questions that don’t resemble the job, our guide to coding assessments is a good place to start.
Ask how the work got done
Use the same set of interview questions for every candidate. Ask them to explain their role, the results they achieved, any problems they faced, and what they would do differently next time. Then, rate their responses using a scoring guide that the team has agreed on beforehand.
Start with the same core questions for every candidate, then use follow-up questions to dig deeper into their answers. When someone makes a vague claim, ask them to explain what they mean and how they did it. If you’re deciding between interviews and other assessment approaches, our comparison to psychometric testing versus structured interviews shows how each method can support hiring.
Keep checking after day one
No screening process gets everything right. Look at new hires against the same skills at 30 and 90 days, and treat any gaps as something to coach. A regular skills gap analysis also shows which hiring signals actually predicted success on your team, so each round of hiring gets a little sharper.
See it on a role you’re hiring for now.
Pick one open role and see how its top applicants score on a real task. Xobin checks resume claims with role-specific tests, job simulations and structured AI interviews, then tracks the same skills after the hire. Try it with Xobin.
Which resume claims are worth checking first?
Not every line on a resume deserves a test. Start with the claims that show up on almost every application for the role and that would hurt most if they turned out to be hollow. These are some of the usual suspects, with a quick way to check each one:
| Common resume claim | A fast way to check it | Rough time |
| “Advanced Excel” | Clean a messy dataset and build a summary in a live spreadsheet | 20 minutes |
| “Writes production code” | Fix a failing test or a small bug in an unfamiliar codebase | 30 minutes |
| “Managed a team of eight” | Structured questions about one specific underperformer they handled, what they did and what happened next | 15 minutes |
| “Strong communicator” | Reply in writing to a frustrated client, then record a short spoken answer | 15 minutes |
| “Data-driven marketer” | Read a campaign dashboard and recommend where to move next month’s budget | 20 minutes |
| “Fluent in English” | A CEFR-level language test that measures speaking and writing | 20 minutes |
You don’t need all six for any one role. Two or three well-chosen checks will cover the claims that matter most. Skip anything you’re happy to teach after the hire, and focus on what the person needs to do well from week one. If you want help designing the tasks themselves, our guide to skills testing for recruiters covers the details.
Hire for what people can do, then help them grow
Xobin brings role-specific skills tests, job simulations, structured AI interviews and proctoring into one hiring flow. Xobin Grow then carries the same skills into development after the hire. Book your 30-minute personalized demo and we’ll walk through how to verify resume claims for one of your open roles.
Frequently asked questions
Does work experience predict job performance?
Only weakly. Experience shows how long someone has done similar work. It says little about how well they did it. Testing the skills that experience is supposed to build gives you far better evidence.
How can you check whether a candidate has overstated their experience on a resume?
Ask them to do the work. A short task tied to their main claims, followed by a few structured questions about their approach, usually separates real experience from inflated wording.
Should we still screen resumes?
Yes, for context and basic eligibility. Use the resume to decide what to test, and let the test results carry the hiring decision.
How long should a pre-hire skills test be?
Short enough that strong candidates will finish it. A focused task on the role’s core skills, usually under an hour in total, tends to tell you more than a long general test.
Are structured interviews better than skills tests?
They do different jobs. Skills tests show what a candidate can do, while structured interviews show how they think through real situations. Using both gives you a fuller picture than either one alone.
