Should You Use Personality Tests to Hire? (DISC, Myers-Briggs, and What Actually Works)

A few years back, an owner I was coaching called me up, thrilled. He'd given his top two candidates a Myers-Briggs test. One came back "ENTJ," which a website told him meant "natural leader." He'd already decided that was his hire.

I asked him one question. "What does that person do when a customer screams at them over the phone?" He had no idea. The test never asked. He'd spent his decision on a label that told him nothing about the actual job.

So let's settle this. Should you use personality tests to hire? Yes, sometimes, in a specific way. As a verdict on who walks out the door with the job? Never. In my 25 years training small-business owners, the test has never once been the thing that made a hire work or fail. The process around it did.

The trap owners fall into

You're hiring on your own, with no team of analysts. So when a tidy report shows up with colors and percentages and a confident four-letter code, it feels like real data. It feels like the thing you've been missing.

Here's the problem. Most of the popular tests were never built to predict whether someone will be good at a job. They were built to sort people into types so they'd understand themselves better. That's a different goal, and the gap between those two goals is where owners get burned.

DISC and Myers-Briggs are the two you'll run into most. They're useful for what they were made for. They're a bad foundation for a hiring decision. Let me show you why, and then show you what to use instead.

What DISC and MBTI actually tell you

DISC sorts people on four traits: Dominance, Influence, Steadiness, and Conscientiousness. Myers-Briggs sorts you into one of sixteen types based on four pairs, like introvert versus extrovert. People love them because they're easy to read and they feel true. You take one, you read your result, you nod.

Here's what they do well. They give you a shared language for talking about how people work and communicate. A team can use DISC to understand why one person wants the bottom line in three bullet points and another wants to talk it through for twenty minutes. That's real value, after you've hired someone.

Here's what they don't do. They don't tell you whether a person can do the job. They don't measure honesty. They don't measure whether someone will show up on time, follow through, or stay calm when things go sideways. Myers-Briggs in particular has a known reliability problem. Take it twice a few weeks apart and a big share of people get a different type. A measuring tape that gives you a different number every time isn't one you should buy a house with.

There's another problem you can't see from the report. These tests are self-reported. The candidate is telling you who they think they are, or worse, who they think you want them to be. Smart candidates figure out the "right" answers fast. The test doesn't catch the gap between what someone says and what they do. Your process has to.

The tests that were actually built to predict performance

This is the part most owners never hear. There's a whole category of assessment built for one job: predicting how well someone will perform. The research backing it goes back decades.

The landmark study is Schmidt and Hunter's 1998 meta-analysis, which pulled together 85 years of research on what selection methods actually predict job performance. Their finding has held up: the strongest predictors are general mental ability (basically, how quickly someone can learn and solve problems), work-sample tests (having someone do a piece of the actual job), and structured interviews. Type-sorting tools like MBTI don't make that list of strong predictors. They weren't designed to.

So if you're going to spend money on a test, here's the order that matters.

A work-sample test is the closest thing to a crystal ball you'll find. Want to hire a bookkeeper? Hand them a messy set of books and ask them to reconcile a month. Hiring a salesperson? Have them sell you the product cold, right there. You're watching the actual work, not reading a description of it.

Cognitive ability tests measure how fast someone learns. For any role where the person has to think on their feet or pick up new systems, this matters more than personality.

Structured personality assessments are the legitimate cousin of DISC. The trait that predicts job performance across almost every role is conscientiousness, which is the tendency to be organized, dependable, and follow through. A validated assessment that measures it is worth real money. The pop-psychology version that sorts you into a color is not the same thing.

The difference between these two families of tests is the whole ballgame. One sorts people into types. The other was built and tested to predict performance. Know which one you're holding.

How I actually use testing

Here's the rule I give every owner I work with. A test is one input. It is never the verdict.

In my system, testing sits at Phase 7. By the time I get there, I've already written real interview questions and watched the person answer them. I've done a working interview or a work sample. I've often done a second interview, which catches more than people expect (I wrote about why you should always do a second interview). The test is one more reading on a candidate I already know well. It carries nothing on its own.

I use a test two ways.

First, as a tie-breaker. When I've got two strong finalists and I genuinely can't separate them on the work, a validated assessment can tip the scale. Not a four-letter code. A conscientiousness score, or a cognitive result for a role that demands fast learning.

Second, as a conversation starter. I'll take a test result into the interview and probe it. "Your assessment suggests you prefer working alone. This role has a lot of team handoffs. Walk me through how you'd handle that." Now the test is earning its keep. It's pointing me toward better questions, not making the call for me. If you want a feel for what good questions sound like, I broke down the kind that predict performance.

What I never do is use a test as a gate. I don't screen people out because they came back the "wrong" type. I've watched owners reject excellent candidates over a DISC profile, and I've watched those same candidates go thrive somewhere smarter. A test that filters people before a human ever talks to them is a lawsuit and a missed hire waiting to happen.

The legal and fairness part you can't skip

This isn't optional, and most small-business owners don't know the rules. In the United States, an employment test that affects hiring decisions can be challenged if it screens out protected groups at different rates, and "but the test said so" is not a defense. The agency that handles this is the EEOC, and their guidance is clear: a test used in hiring has to be job-related and validated for the job you're using it for.

Translation for you. A validated, job-relevant assessment used the right way is defensible. A personality quiz you found online, used to reject people, is not. If a test ever becomes the reason someone didn't get hired, you'd better be able to show it actually measures something the job requires. DISC and MBTI were never built to clear that bar.

What to do, step by step

Here's how to put this to work without overthinking it.

1. Decide if you even need a test. For most small-business hires, a strong work sample plus a structured interview already gives you more than any personality report will. Add a test only when it answers a question you can't answer otherwise.

2. Pick the right kind. If you want to predict performance, reach for a work sample, a cognitive ability test, or a validated conscientiousness assessment. Skip DISC and MBTI for the hiring decision itself. Keep them in your pocket for developing the team after they're hired.

3. Check that it's validated. Ask the vendor one question: "Has this been validated to predict job performance, and can you show me the evidence?" A real assessment company will have an answer. A personality-quiz site will get quiet.

4. Use it late, never early. Run the test after your interviews and work sample, not before. It confirms or complicates a picture you already have. It doesn't create the picture.

5. Make it an input, not a gate. Let the test inform a question or break a tie. Never let it auto-reject a candidate a human hasn't met.

6. Pair it with everything else. The whole point of reference checks, work samples, and second interviews is that each one catches what the others miss. A test is one more layer, not a replacement for the rest.

The bottom line

A personality test can help you hire. It cannot hire for you. The popular ones, DISC and Myers-Briggs, are useful for understanding people you already employ and weak for predicting whether a stranger will be good at the job. The tools that predict performance are work samples, cognitive ability, and validated measures of traits like conscientiousness, and they only work as one piece of a process you run with your own eyes and ears.

Testing is Phase 7 in my 10-phase hiring system, and it sits where it does on purpose. By the time you're testing, you should already know who you're leaning toward. The test just helps you check your work.

If you want the rest of the system, the interviewing playbook I've built over 25 years lives in The Naked Interview. And if you're curious how your own instincts shape your hiring before you ever pick a test, the hiring style quiz is a good place to start.