How to Reduce Time-to-Hire Without Lowering the Bar

When a search drags, the usual reaction is to add recruiters or post on more job boards. That is almost always the wrong reaction, because the lost time is rarely in sourcing candidates.

It is in the dead days between stages. And those days do not show up in the total.

Two metrics almost everyone blends

Time-to-fill counts from requisition approval to candidate acceptance. It includes however long it took you to decide you needed the person.

Time-to-hire counts from the candidate entering the process to acceptance. It measures your process, not your planning.

Blending them makes the number meaningless. A long time-to-fill with a short time-to-hire is an internal approvals problem. The reverse is a process problem. The two causes call for opposite actions.

Why the benchmark makes a bad target

The next question is always what the normal number is. SHRM’s own data moved from roughly 54 days on average in 2022 to around 44 in a later measurement, and the figures shift a lot by sector and role type.

A range that wide does not give you a target. It gives you an excuse in either direction. The two comparisons that hold up are against your own number from last quarter, with the same definition, and across searches inside your own company in the same period.

Where the time actually goes

Before touching anything, split the process and measure the lag at each stage. The five stretches that matter.

Application to first review. Review to first interview. Between interviews. Last interview to decision. Decision to offer sent.

In most processes, four of those five stretches are pure waiting. Nobody is working, the file is sitting on a desk. That is the same failure that kills most hiring kanbans, developed in the hiring pipeline.

The five levers that move the number

1. Define the profile before posting. A search that starts with no written criterion turns into three weeks of calibration disguised as interviews. The team interviews in order to discover what it is looking for. How to write the profile is in job profile with data.

2. Move the assessment forward. The usual order puts it after the third interview, once six hours of expensive people have already been spent. Moving it to the start sorts the shortlist before calendars get involved. Same information, two weeks earlier.

3. Compress the calendar, not the process. Four interviews in one day are not worse than four across three weeks. They are better, because the candidate compares while it is fresh and the team decides with everything in mind. What stretches processes is the calendar, not the number of steps.

4. One decider, with written criteria. A committee deciding by consensus takes twice as long and produces the average of the opinions. One person decides, everyone else brings evidence against a criterion agreed beforehand.

5. Close the “let’s see one more” loop. It is the most expensive leak. Past a certain point, each additional candidate adds little information and a lot of calendar, and raises the risk of losing the finalist you already had. Agreeing in advance how many finalists is enough cuts that spiral.

What not to cut

Speeding up is easy if you give up the criterion. It is also expensive, and the cost shows up months later. It is laid out in the real cost of a bad hire.

Three things hold even under pressure.

Interview structure. It is what makes two candidates comparable. Improvising it to go faster turns the process into a series of chats. The guide is in competency-based interviews.

The written criterion. Without it, speed only accelerates the decision you were always going to make, skew included.

The conversation with the manager about the role’s real context. Half an hour that saves weeks.

Candidate experience is part of the number

A slow process loses finalists, and losing a finalist restarts the clock. That restart never appears in the report, even though it is the most common way a thirty-day search ends up at seventy.

Two cheap practices cut the leak sharply. Always respond, rejections included, within a stated window. And tell the candidate on the first call how many stages there are and how long it will take.

What it looks like when it works

A tidy process usually reads like this. Profile written before posting. Assessment in the first 48 hours in the pipeline. One fit conversation with the manager. One block of technical and competency interviews on the same day. Decision within the next 72 hours. Offer the day after.

The steps are the usual ones. What changed is that none of them waits for someone to remember.

Where the instrument fits

Moving the assessment forward only works if it is fast for the candidate and produces something the manager can read without translation. A forty-minute assessment with high abandonment makes the problem worse, because it adds a stage that filters for patience.

At Talen.to the initial diagnostic takes about ten minutes and produces the profile against the role and against the company’s values. That moves the shortlist to the top of the funnel, where time is still cheap. The full logic is in the talent lifecycle.

Before changing anything, measure the five stretches of your last closed search. There is almost always one stage eating half the calendar, and it is almost never the one the team expects.