Their work is excellent. You can't argue with the output — the deck is polished, the analysis is right, the document holds up. The problem is it took two weeks. You pinged them three times. They missed the meeting it was supposed to inform. By the time it landed, the moment had passed and you'd already winged it without them. The quality isn't the issue. The pace is.
Most managers diagnose this as a quality problem and back off. "Their work is great, I can't really complain." That framing is the trap. Slow-but-correct isn't a quality problem — it's a calibration problem. The person doesn't know what "done" looks like, so they keep going until something else (a deadline, a manager pinging them, sheer exhaustion) forces them to stop.
Research summarized in HBR found that perfectionism has essentially no relationship with job performance. Perfectionists don't produce better work. They just spend more time producing the same work. What you're seeing isn't a person who cares more about quality than their peers. It's a person who can't tell the difference between work that's done and work that could still be done.
That's a coaching gap, not a personality trait. And it's yours to close.
Three things to do this week.
Define "done" explicitly, in writing, before they start. Most slow people aren't shown what good enough looks like — they're given a vague brief and told to make it good. So they keep polishing because there's no signal that says stop. Before assigning work, write out what the deliverable needs to include, what level of polish it requires, and what it explicitly does not need. "Two-page summary, bullets are fine, no need to format charts. Audience is me, not the exec team." Then point at an example of past work that hit that bar so they have something concrete to calibrate against.
Give time budgets, not just deadlines. A deadline says when. A budget says how much. "This is a four-hour task" is a different instruction than "this is due Friday." Friday lets them spend twelve hours on it. Four hours forces a tradeoff. When you assign work, attach an estimate of how long it should take. If they routinely blow through your estimates, that's useful diagnostic information — either you're under-scoping, or they're over-engineering, or there's a skill gap making the work take longer than it should. All three are addressable, but only if you're tracking the gap.
Make them ship something at 70% on purpose. The single most useful intervention for a chronic over-polisher is forcing them to deliver something visibly unfinished and watching nothing terrible happen. Pick a low-stakes deliverable, tell them you want it at draft quality with caveats and gaps, and have them send it. Afterward, point out that nothing broke. The mental model you're trying to install is that shipping at 70% twice is usually more valuable than shipping at 100% once — because you get feedback in between.
The conversation itself is short: "Your work is good. The problem is the pace, not the quality. I need to show you what the bar actually is, because I think you've been aiming higher than the work requires."
This doesn't apply to work that genuinely needs perfectionism — board materials, regulatory filings, security-critical code. The job isn't to make everyone fast. It's to teach calibration: which work deserves three drafts and which deserves twenty minutes. Slow-but-correct people are usually excellent at the first kind. The fix is helping them see the second kind exists, and that most of their work is the second kind.
They're not slow because they care more. They're slow because nobody ever told them when to stop. That's your job.
