Why do new hires quit in the first three months?

Almost every early exit can be traced back to a decision made before the person started. The job ad was written from a list of duties instead of from what the role has to achieve. The interview tested how well someone talks about work rather than how they do it. Nobody wrote down what a good first year looks like, so nobody could check whether this candidate was likely to deliver it.

That produces a specific kind of hire: someone who is genuinely impressive in a room and genuinely wrong for the job. They said yes to a role they pictured from a vague description. You said yes to a person you liked. Three months in, both of you discover the picture and the person do not overlap, and the honest name for that is a hiring failure, not a loyalty failure.

The second common source is a job ad that oversells. A five-person company hiring its sixth writes an ad that describes a company twice its size, because that is the company it hopes to be by December. The candidate arrives expecting the structure that was implied - a manager with time, a tool budget, a team to lean on - and finds a desk, a laptop and a lot of unanswered questions. Nothing was technically a lie. The impression was still false.

• A duties-based job ad: candidates apply to the words, not the work
• An interview that rewards charm, confidence and good English over evidence
• No written definition of what this hire must have achieved in a year
• An ad that describes the company you are becoming rather than the one they are joining

Some early exits have nothing to do with the work at all. Commute is the biggest and the most predictable. A role in Victoria Island filled by someone living in Ikorodu is a role with a hidden four-hour daily tax on it, and enthusiasm covers that tax for about six weeks. If the person is expected in the office five days a week, ask where they live and what the journey actually looks like at 6am, before you make the offer. Not to screen anyone out, but so both of you have priced it honestly.

Pay is the second. Salary paid late, or paid differently from what was described, ends more probations in Nigeria than performance ever does. If the number quoted was gross and the person was hearing net, if allowances were mentioned that turn out to be discretionary, if the first payment lands two weeks after the date you promised, you have taught your new hire that your word is approximate. Someone will make them a cleaner offer, and they will take it.

Third is what probation is used for. A probation period is meant to be a structured onboarding with an honest checkpoint at the end of it. Used as a cheap trial - lower pay, no clarity, no feedback, a decision that arrives without warning - it selects for people who keep interviewing elsewhere while they work for you. Whatever you agree, write it into the offer letter: length, the pay during it, what is being assessed, when the review happens.

Ask a manager what their new hire will do on day one and you often get an answer about laptops and email accounts. Ask what the person should have done by the end of week two and there is usually silence. That silence is the onboarding gap, and the new hire fills it with guesses.

A capable person with no defined priorities will pick their own, work hard at them, and present the results to a manager who wanted something else. Do that twice and the hire concludes they are failing. Managers read the same evidence and conclude the hire is not who they interviewed. Both conclusions are wrong, and both are produced by the same missing document.

The manager problem sits underneath this. Most early attrition in small companies is not a bad hire, it is an unavailable manager. A founder hiring their first operations person is usually the busiest person in the building, and the new hire is the item that gets rescheduled. Ninety days of being the lowest priority on your manager's calendar teaches someone exactly how much the role matters.

• One hour with the manager in week one is worth more than a fifty-page handbook
• Name the first small win and make it achievable inside two weeks
• Say who they should ask when the manager is unreachable, and tell that person too
• Give feedback at week two, not at the probation review, when it is already a verdict

The work that prevents an early exit happens before the offer. It is not more interviews. It is being specific about the job earlier, and letting candidates see the real thing before they commit to it.

Start by writing what must be true in a year for this hire to have worked. Three to five outcomes, each one a result rather than an activity: not 'manage customer support' but 'first-response time under four hours with a support team of three by month nine'. Those outcomes become your scorecard, your job ad, your interview questions and your 30/60/90 plan. One document, written once, doing four jobs.

Then show candidates the actual work. A short task, a scenario drawn from something that happened last quarter, or a one-minute video answer to a real situation the role will face. Candidates self-select honestly when they can see the job, and you learn more from ten minutes of someone doing the work than from an hour of them describing it. When you ask about previous roles, ask why each one ended, and listen across the whole history instead of judging any single answer. One difficult exit is life. The same story four times is a pattern, and the pattern is the signal.

1. Write three to five year-one outcomes for the role before you write the ad
2. Turn those outcomes into a scorecard, and score every candidate against it
3. Add a demonstration stage: a task, a scenario, or a one-minute video answer to a real situation
4. Ask about every previous exit and listen for the repeated pattern, not the single bad chapter
5. Be explicit about pay, payday, probation terms and days in office before the offer is signed
6. Write the 30/60/90 plan before day one and give it to the hire on day one

Not every three-month departure is a failure to be prevented. If the evidence in month two says this person cannot do this job, ending it quickly is kinder and cheaper than carrying them to month ten and calling it patience. The same goes in the other direction: a hire who realises early that the role is wrong for them and says so is doing you a favour, not betraying you.

The measure of a hiring process is not zero early exits. It is whether the exits that happen are informed rather than surprising. If you defined outcomes, showed the real job, agreed the terms in writing and reviewed at week two and week six, then a month-three parting is a decision both sides could see coming. That is a functioning process. What you are trying to eliminate is the exit nobody saw coming, because that one was always avoidable at the job-ad stage.

Tools can carry part of this. Tezera builds the scorecard from the outcomes you define and runs a demonstration stage so you see candidates doing something before you hire them, with the AI analysing the evidence and a human making the call. Its limitation is the same as the method's: it cannot score anyone against outcomes you have not defined, and it does nothing about a manager who has no time in week one. Define the outcomes yourself, or the software has nothing to measure against.

The same question, phrased the other ways people ask it. Every one of them lands on this page.

• Why do employees leave shortly after being hired?
• Why do new staff resign during probation in Nigeria?
• How do I stop new hires from leaving early?
• What causes early attrition in small companies?

• →What is a hiring scorecard and how do I make one?
• →How do I verify a candidate's CV is genuine in Nigeria?
• →How much does it cost to hire an employee in Nigeria?
• →Interview questions for a customer success manager

Everything else we have written like this sits on the answers index.