What to Do After a Bad Hire: A Step-by-Step Plan

The moment you know

It's usually not day one. It's week three or four. You've given the same instruction twice, the client email that should have gone out on Tuesday is still a draft on Friday, and the rest of the team has started routing around this person instead of through them. You're not panicking yet. You're just quietly certain.

Most owners sit on that certainty for another six weeks, hoping it resolves itself. It rarely does. Here's what to actually do instead.

The two ways people get this wrong

One group waits too long — they keep hoping, avoid the awkward conversation, and let the problem cost them a quarter of lost output and a demoralised team before they act.

The other group panics and fires on gut feeling within a week, without ever finding out whether the problem was fixable. Sometimes that's the right call. Often it isn't — some of what looks like a bad hire is actually a bad first month: unclear expectations, no proper handover, a manager who's too busy to check in.

The fix for both is the same: slow down for two weeks, gather actual evidence, then move fast.

Step 1: Name the problem precisely

"It's not working out" isn't a diagnosis. Force yourself to write one sentence that says exactly what's wrong. There are basically three categories:

• Skill gap — they genuinely can't do the work at the level the job needs. Missed technical detail, can't structure a report, doesn't understand the tools.
• Effort or reliability — the skill is probably there, but deadlines slip, they're late, they disengage in meetings.
• Fit or conduct — they undermine colleagues, argue with feedback, or the way they operate clashes badly with how your team runs.

These need different responses. A skill gap might be coachable in thirty days. A conduct problem rarely is.

Step 2: Write down the evidence before you talk to anyone

Not how you feel about them — what actually happened, with dates. "Missed the March 14 deadline by two days," not "seems unmotivated." You need this for three reasons: it keeps the coming conversation honest and specific instead of vague and defensive, it protects you if the exit ever gets legally messy, and — this is the one people skip — it tells you whether the pattern is as bad as your gut says, or whether you're reacting to two bad weeks.

Step 3: Have one direct conversation

Not a hint. Not a passive-aggressive email. Sit down and say something like:

"I want to be straight with you. Over the last month I've noticed [the two or three specific things]. That's not the level we need for this role. I want to know if you think this is fixable, and if so, what would need to change."

Watch the response as closely as the words. Someone who owns it and asks how to improve is a different case from someone who blames the tools, the training, or everyone else on the team.

Step 4: Set a short, measurable improvement window — or decide it's over

If you're giving them a chance, make it concrete and short. Not "try harder for a while" — something like: "Over the next two weeks, I need to see X delivered on time, Y done without me checking it, and Z stopped happening." Put it in writing. Two weeks is usually enough to see if the trend actually changes; a month just delays the decision you already know you'll have to make.

If it's a conduct issue, or if you already know in your gut this isn't fixable, don't run the fake improvement plan. It wastes everyone's time and it's not kind to string someone along when you've already decided.

Step 5: If it's exit, do it cleanly

Check your local notice period and probation terms first — rules differ enough between Malaysia, the Philippines, UAE, South Africa and Singapore that this is worth ten minutes with your contract or a local advisor before you say anything. Then:

• Tell them in person or on a call, not by message.
• Be brief and factual. You don't need to relitigate every incident — one or two examples is enough to make it clear this isn't arbitrary.
• Have their final pay and logistics (equipment return, access removal) sorted before the conversation, not scrambled after.
• Don't promise a reference you're not prepared to give. "I can confirm dates of employment" is honest if that's all you can say.

Step 6: Tell the team something true

Your team already knows something was wrong — they've been covering for it. You don't owe them the full story, but a vague "there's been a change" makes people nervous about their own job security. Something like "[Name] and I agreed this role wasn't the right fit — we're starting the search again" is honest without being a post-mortem.

Now look at how they got hired in the first place

This is the part people skip because it's uncomfortable — it means looking at your own process, not just the person who didn't work out. Ask plainly: did the interview actually test the thing that turned out to be the problem? If someone interviewed brilliantly and then couldn't structure a report, the gap is usually that nothing in the process asked them to structure one before you hired them. A skills-based test or work sample closes that specific gap — it's the difference between someone describing how they'd do the job and actually doing a slice of it. AssessFit exists for exactly that piece: candidates take a proctored, adaptive test built from calibrated items, and the CV can never count for more than 30% of the score, so confident interviewing can't paper over a real skill gap the way it can in a conversation alone.

The one honest caveat

Not every bad hire is a process failure. Sometimes someone tests well, references check out, the work sample is solid, and they still don't work out — people change, or the fit was genuinely unpredictable until they were in the seat. Don't over-correct into six-week hiring processes for every junior role because one hire went badly. Fix what you can verify was actually missing, and accept that some of this is judgement under uncertainty, every time.