New Hire Not Working Out? Ramp Problem vs. Fit Problem


When a new hire is not working out, the useful first question is whether you are watching a ramp problem or a fit problem. They look identical at day 45 and they call for opposite responses. A ramp problem needs support and time. A fit problem gets worse with both.
The test that separates them is whether last month's feedback shows up in this month's work. What follows covers the signals to read, the conversation to hold, and what to fix in the process before you restart a search. The 30, 60 and 90 day markers are a management review cadence. They carry no legal weight as a probationary period.
A structured review is ordinary management, and bad hires are common enough to plan for. In CareerBuilder's survey of 2,257 US hiring managers, 74% said they had hired the wrong person for a role, at an average reported cost of $14,900 per bad hire. The more useful number in that 2017 survey is the top reason employers gave. At 35%, it was hiring someone without all the needed skills because they thought the person could learn quickly. That is a bet on ramp. The carry-forward test below is how you find out whether the bet is paying off.
Rules on ending employment vary by state and by worker classification, so take any exit decision to HR or employment counsel before you act on it.
Bad hire recovery starts with that distinction. A ramp problem is a person moving in the right direction more slowly than you expected. Their questions get sharper each week, and what they are missing is knowledge: how the account works, who owns what, where the data lives.
A fit problem shows no directional movement. The same questions come back, feedback does not change behavior, and what is missing is judgment: which problem to work on first, when to escalate, what to stop.
Ramp also takes longer than most managers budget for. Gallup puts the time for a new employee to reach full performance potential at around 12 months, and warns against comparing new hires with veterans too early. So day 90 is too soon to judge output and early enough to judge direction.
The carry-forward test is a monthly check of whether last month's feedback shows up in this month's work, and it helps hiring managers separate a ramp problem from a fit problem before the day-90 decision. Knowledge arrives with time. Judgment mostly does not.
The test only works if the feedback was specific enough to carry. In a Gallup study of nearly 15,000 employees, only 16% said their last conversation with their manager was extremely meaningful. Write down each piece of feedback and the deliverable where you expect to see it. If you can't point to either, the test is measuring your feedback.
Before you read anything as a fit problem, rule out the causes that sit on your side. Some hires look like they're failing when the setup failed them.
Start with four checks. Did they get the access they needed, or spend two weeks waiting on permissions? Has the scope changed since the offer? Are instructions arriving from a founder and a manager who disagree? Did anyone write down what working looks like at 90 days? If you ran the brand rows of the scorecard in our 30-60-90 day plan for onboarding an ecommerce hire, you already have the answers.
If access was late, the clock has not started yet, so move the checkpoint back by however long they waited. Third-party access is the usual culprit, since agency-owned ad accounts, 3PL portals and retention tools each sit with a different owner. Scope drift shows up on the hire's side too. In the same CareerBuilder survey, 37% of workers who realized a job was a bad fit said it didn't match what was described in the listing and interviews, and 33% pointed to a lack of clear expectations.
At 60 days, read the five signals below together. One owned cycle of work should have closed by now, and how it closed tells you more than whether it did.
One weak signal is normal. When several point the same way at day 60, act, and at this stage acting means an explicit conversation about the gap. The decision comes at 90.
By 90 days you should be able to say whether this is working. If you cannot, the usual reason is that nobody wrote down what working meant, and that is an onboarding gap before it is a performance one.
Two inputs decide it: the carry-forward result, and whether you ruled out the causes on your side.
Hold the conversation directly and early. Name the specific gaps, state what would need to change and by when, and ask what they need from you. Some hires turn around here. They are usually the ones whose problem was scope or access, with capability never in question.
Don't wait for the hire to resign. BLS describes the quits rate as a measure of workers' willingness or ability to leave jobs, and it sat at 1.9% in July 2026. When few people are quitting anywhere, a struggling hire is unlikely to resolve it by leaving.
If the decision is to end the arrangement, bring in HR or employment counsel now. Every state except Montana allows at-will employment, but the reason for an exit still can't be illegal, such as discrimination or retaliation.
Plenty of failed hires were mis-specified. The brand picked the right person for the role it wrote down, and the role it wrote down was wrong. Speed makes it worse: 30% of employers in the CareerBuilder survey said pressure to fill the role quickly led to their bad hire. This is the least comfortable part of the review. A hiring mistake ecommerce teams make over and over is treating the person as the only variable.
Four patterns come up. The role was written for one seniority and filled at another. The scope changed between interview and start. The person was hired to fix something the founder would not stop doing. Or the interview tested for experience the job did not need and never tested the one thing it did.
Two recent searches show how early this starts. An apparel brand came back to us to backfill its creative lead. The previous hire leaned toward brand storytelling, and the role needed a direct response operator running strategy behind $1M+ a month in Meta spend. A DTC health brand's search opened as a senior performance marketer role, was discussed internally as head of growth, and closed as performance marketing director. Both were scoping questions before they were hiring questions.
Ask what you would have needed to see in the process to catch this. If no stage would have caught it, add one. On head of growth searches, a common starting point is a job description that reads like a senior media buyer, a CMO and an analyst rolled into one title.
Rewrite the description before reposting it. If the last hire failed on scope, the scope was wrong on paper first.
Fix the reporting line if instructions were arriving from two directions. Fix the band if the level you need costs more than you offered. Hiring a cheaper version of the same role is how a replacement hire process ends up repeating the original error.
Add the interview stage that would have surfaced the gap. If judgment was the problem, add a work sample. If ownership was the problem, add a conversation with the person who would report to them. Whatever you add, structure it with the same questions, in the same order, scored against the same rubric. A 2022 re-analysis of personnel-selection research in the Journal of Applied Psychology ranked structured interviews as the strongest predictor of job performance it studied.
Contingency recruiters commonly offer a replacement guarantee, and the terms decide what it is actually worth.
A replacement guarantee usually means that if a placed candidate leaves or is exited inside a defined window, the recruiter runs a further search without a new fee. Terms vary between firms. Before you sign, check whether it is a replacement search or a refund, what voids it, whether it survives a change to the role's scope, and when the window starts counting.
Get the terms in writing, and tell your recruiter at the 60-day conversation. Waiting until the exit is too late. At Constant Hire, our searches run on a contingent model, so no fee is due until you hire, and the guarantee window is set in each client agreement. Every placement sits in our ATS under an in-guarantee status until that window closes, so an early flag starts the replacement conversation while the window is still open.
Say less than you think, and say it once. Teams notice when something is not working long before a manager raises it, and silence feeds the speculation. A short, calm statement that the role is under review, followed by a clear account of who covers what in the interim, settles most of the anxiety.
Keep the conversation on the work and the coverage plan, and leave the individual out of it. The team wants to know who they go to now and whether their own job is safe. Answer both directly.
A failed hire usually lands in one of three places, and each has a different fix. If feedback carried forward but results lagged, it was a ramp problem, and the answer is time. If brand-side causes were present, it was an onboarding problem, and the hire's scores were measuring your setup. If the carry-forward test failed with those causes ruled out, it was a fit problem, and some part of the process let it through.
The expensive mistake is treating the first two as the third. That's how brands exit someone who was two months from working, then rehire into the same scope. When one of our placements doesn't work out, the replacement search reopens the intake before it reopens sourcing, because the same scope tends to produce the same hire.
Check whether last month's feedback appears in this month's work. A ramp problem shows steady directional movement and knowledge gaps that close over time. A fit problem shows judgment gaps and repeated feedback that never changes behavior. Knowing when to let a new hire go depends on that distinction more than on any single missed target.
No. The 30, 60 and 90 day checkpoints here are a management review cadence with no legal status. Employment law varies by state and by classification. Montana, for example, is the one state that doesn't allow at-will employment, so take any exit decision to HR or qualified employment counsel first.
Terms vary by firm. Generally it means a further search at no new fee if a placed candidate leaves inside a defined window. Ask in writing whether it is a replacement or a refund, what voids it, and whether it survives a change to the role's scope.
Judge direction at 60 and 90 days and output later. Gallup puts full performance at around 12 months, so a hire who is improving but not yet producing is on track. If feedback isn't carrying forward at 60 days and you've ruled out access, scope and conflicting instructions, a decision at 90 is reasonable. If any of those causes were present, move the checkpoint by the time lost.
Yes, and early. Guarantee windows are time-bound, so a flag at the 60-day conversation keeps your options open, while a flag at the exit may not. The recruiter's intake notes also record what the role was scoped as, which is usually the fastest way to see whether it drifted.
Check your side first. Late access, a scope change since the offer, conflicting instructions from a founder and a manager, or no written definition of success at 90 days will all make a capable hire look like a poor one. If any of those failed, fix it and rerun the check before you judge the person. Our onboarding plan includes a scorecard for the brand's side.
First interviews in 5 business days,
or 10 for executive searches.