I have sat through hundreds of technical interviews. The ones I remember are not the ones with the cleanest whiteboard code.
Here is the pattern I keep seeing. You run a rigorous loop. Four rounds. System design, a coding exercise, a take-home, and a values chat bolted on at the end because someone in HR asked for one. The candidate clears every stage. You send the offer. Seven months later they are gone. Or worse, they are still on the payroll and quietly checked out.
Nothing in your loop was broken. Your loop measured skill. Skill was never the thing about to fail.
Your Interview Loop Screens the Wrong 11 Percent
Leadership IQ ran a study across 5,247 hiring managers at 312 organizations, covering more than 20,000 hires. Their finding: 46% of new hires were considered failed hires within 18 months.
Now look at the reasons those hires failed:
- Coachability: 26%
- Emotional intelligence: 23%
- Motivation: 17%
- Temperament: 15%
- Technical competence: 11%
Eleven percent.
Read your interview scorecard again. Every rubric line, every rehearsed system design prompt, every take-home you asked someone to spend a weekend on. All of it aimed at the 11 percent slice. The other 89 percent gets a thirty minute "culture chat" with whoever had a free calendar slot.
We built an entire hiring industry around the smallest failure mode. Then we act surprised when good engineers wash out.
Purpose Fit Is Not Culture Fit
Before you tear up your process, understand the trap waiting for you.
Lauren Rivera studied hiring at elite law, banking, and consulting firms for her paper Hiring as Cultural Matching. More than half of the evaluators she studied rated cultural fit as the most important criterion at the interview stage, above analytical thinking and above communication.
Sounds like a win for my argument. It is the opposite.
Rivera found employers wanted candidates who were "not only competent but also culturally similar to themselves in terms of leisure pursuits, experiences, and self-presentation styles." Her conclusion was blunt: concerns about shared culture "often outweighed concerns about absolute productivity."
Leisure pursuits. They were hiring people to ski with.
Culture fit, as most companies practice it, means hiring your own reflection. Same schools. Same accent. Same weekend hobbies. It feels like rigor and it is bias with a scorecard. It narrows your team on every single hire until everyone in the room shares the same blind spots.
Purpose fit asks a different question: do the two of us believe this problem is worth solving?
Two people both answer yes and share nothing else. Different country, different background, different politics, different taste in editors. One drinks with the team on Friday, one goes home to the kids at 5pm sharp. Neither fact matters. They both want the thing built, and they agree on why.
Culture fit points inward at each other. Purpose fit points outward at the same horizon.
Ask Engineers What They Care About
The 2025 Stack Overflow Developer Survey found 24.5% of developers are happy at work. 47.1% are complacent. 28.4% are unhappy.
The top three drivers of job satisfaction, in order:
- Autonomy and trust to manage your own tasks
- Competitive pay
- Solving real-world problems
Solving real-world problems ranks third, above perks, above the tech stack, above title. And "you like your manager" comes in ninth.
Ninth deserves a moment of your attention. Engineers rank their boss below the problem they get to work on. My own research into bad bosses found 99.5% of respondents had experienced one or more types of bad boss, which is work I do with Step It Up HR. Most people have written off the idea of a good manager. They stopped expecting one. What keeps them in a chair is work worth doing.
So when you sell your open role on ping pong and a new laptop, you are answering a question nobody asked.
Five Questions I Ask Instead
None of this works as a vibe check at the end of the loop. Purpose fit needs the same rigor you give a system design round. Here are the questions I use.
1. What problem do you keep coming back to on your own time?
Not "what are your hobbies." What nags at them. Someone who rewrites their own budgeting tool every two years is telling you exactly where their attention goes for free.
2. Tell me about a piece of work you were proud of. Who did it help?
Watch what happens after "who." Some people light up describing the user. Some go quiet, then talk about the architecture again. Both answers are useful. Only one of them fits a team whose job is shipping outcomes.
3. What do you think we get wrong?
The best candidates arrive having formed an opinion about your product. A candidate with a sharp criticism of your onboarding flow has already started working. A candidate with nothing to say has applied to forty companies with the same cover letter.
4. Describe a time you pushed back on a spec. Why did you care enough to fight?
Coachability was the number one failure driver in the Leadership IQ data, and coachability is not obedience. I want someone who argues, then commits. The "why did you care" half of the question does the real work here.
5. Two years from now, what is different because you were here?
If the answer is entirely about their own growth, you have a passenger. If the answer involves the company, the users, or the team around them, you have a partner. Both are legitimate hires at different moments. Know which one you are getting.
None of these ask about weekend plans, favorite bars, or whether someone seems like "one of us." Purpose fit is testable. Culture fit is a feeling, and feelings are where bias lives.
"So You Want Me To Hire Bad Engineers?"
No. This is the objection I get every time, so let me kill it now.
Skill is a floor, not a ranking.
Set a real technical bar. Make it high enough for your codebase. Test it honestly, with work resembling the job rather than a puzzle from a 1990s interview book. Anyone below the bar is a no. No exceptions, no charm offensive, no "we will grow them into it."
Then stop ranking the people above the bar by skill.
Once three candidates clear your bar, the technical gap between them is noise next to the motivation gap. A bored senior engineer ships less in a quarter than a fired-up mid-level engineer ships in a month. I have watched it happen on my own teams more than once. The strongest resume in the stack has cost me more time than any weak one ever did.
Here is the compounding part. Skill grows on the job. A motivated engineer who cares about the problem picks up your domain, your stack, and your weird deployment quirks inside a quarter. Purpose does not grow on the job. Nobody starts caring about your mission in month eight because you sent out an all-hands deck.
What Getting This Backwards Costs You
Run the numbers on your own team. A failed hire burns the recruiting spend, the salary, six months of a senior engineer's time on onboarding and code review, the morale hit when they leave, and the hiring loop you have to run all over again. The Leadership IQ data puts close to half of your hires in the failed bucket.
No amount of extra algorithm rounds fixes it.
Go pull up the scorecard from your last engineering hire. Count the questions aimed at whether the candidate was able to do the work. Now count the questions aimed at whether they wanted this work, at this company, on this problem.
If the second number is zero, you have found your bottleneck. What is stopping you from adding one of those questions to next week's loop?