Time-to-Hire Is the Wrong Clock

How long did it take each person you hired last year to do their job without help? Most talent teams can't answer that. Ask them about time-to-hire, though, and they'll give you a number to the day, probably with a trend line.

Nobody thinks time-to-hire is the more important of the two. It's just easier to measure. Time-to-hire starts and stops at clear events, a posting and an accepted offer, and both happen inside recruiting's own systems. Time-to-productive ends at a fuzzy point, weeks or months after the person starts, in some other department. So teams measure the first one, put it in the board deck, and feel good when it drops to 18 days.

But look at what that number can and can't tell you. It says how fast the process moved. It says nothing about whether the person who started last month is contributing on their own yet, or needed three months of hand-holding, or is slowly turning into a mis-hire.

There are two kinds of metric worth separating here. Process metrics, like time-to-hire, interviews completed, or offer acceptance rate, tell you how the machine is running. Outcome metrics, like ramp time, 90-day retention, or revenue per employee, tell you whether you hired the right people. Most recruiting dashboards are full of the first kind. That would be fine if the two moved together, but I don't think they do.

Why speed and quality drift apart

Think about what happens when a hiring manager needs someone in the seat by the end of the month. Nobody announces that the bar is being lowered. It just gets lower. Gut feel fills in where careful early assessment should have been, and since the team is judged on speed, nobody has much reason to stop and ask whether the process gathered enough information to predict how the person will actually do. The result is the hire who checks every box on paper and then struggles quietly for six months before both sides admit it isn't working.

Those mistakes are expensive. SHRM has put the cost of replacing an employee at anywhere from half to twice their annual salary, depending on the role. A process that saves a week on hiring and then produces one of those hasn't saved anything.

The answer isn't to slow down, though. The teams that get this right, as far as I can tell, improve the quality of what they learn early, so the speed they do achieve rests on better information. That's a different problem from the one most ATS optimization projects are trying to solve.

What a good first conversation tells you

This is where I think AI screening is usually sold on the wrong benefit. It does reduce time-to-fill, but the more lasting value is in what a hiring manager knows before spending an hour with a candidate.

A well-built AI recruiter can run a structured screening conversation with every applicant. It asks everyone about the same things, it doesn't get tired on the fortieth call, and it doesn't warm to people who sound familiar. It also hears how someone answers, and that part matters more than people admit. How a candidate explains a past project, handles an ambiguous question, or recovers after a moment of uncertainty tells you something about how fast they'll ramp, and none of it shows up on a resume.

When those answers are in the ATS before the first human interview, the hiring manager walks in with real context. They ask better questions and make better decisions, and the people who get through are more likely to fit. You won't see that in time-to-hire. You'll see it weeks or months later, in time-to-productive.

This is what we built Asendia to do. It holds a structured screening conversation with each applicant, including the ones who apply in the evening or at the weekend, which matters more than it sounds: people who already have jobs tend to apply then, and if screening only happens during office hours, some of them won't still be waiting when someone is back at a desk. What the recruiter finds in the ATS is a written summary of each of those conversations, and because every candidate was asked about the same things, the summaries can be read side by side and compared. So recruiters start from a shortlist with context instead of a pile of unread CVs.

The part I find most interesting is what it makes measurable. If every candidate had the same kind of first conversation, you have a baseline. You can start comparing what you learned at screening, such as how clearly someone communicated or how deep their role-specific knowledge went, with how they were doing at 90 days, and use that to adjust the screening criteria. Over time that should make hires both faster and better. We make a similar argument about the whole recruiting process in why every recruiting team needs an AI recruiter agent.

If you want to start without any new tools, take your last ten hires and ask each of their managers one question: from what date could this person do the job without help? The answers will be rough, and that's fine. It's still the first time you'll have measured whether your hiring worked, instead of how quickly it happened.

Ready to transform your hiring strategy? Schedule a Demo with our founders today!