RehearseaBlog
4 min read

What a 45-minute interview is actually costing you

The research behind this product contains exactly one measurement of what an interview process costs the candidate, and it is a single reply in someone else's thread. It is still the only one there is.

  • interview process
  • job search
  • time cost
  • hiring

Nobody counts their applications and everybody counts their hours. That is backwards, because the applications are nearly free and the hours are not. A candidate can send two hundred applications without spending more than a week on them. The process on the other side of that sends you somewhere for a day, sometimes more than once, and there is no way to opt out of the parts you cannot see.

The one measurement in twenty-one threads

In a thread about a graduate struggling to find entry-level work, a reply describes an interview of three to four hours, eight people, and a salary of around 65k.

The author applied for IT Helpdesk Support at a small company, and "literally asked for the same salary" they had asked for before. They interviewed with seven different people in person, plus a phone screening, eight in total, and one of those people did a further phone call after the phone screening. Three rounds. The in-person interview included the other two helpdesk people, the IT director, the DevOps manager, the infrastructure manager, the HR manager, and the COO. It was broken into four different groups, some questions repeated across different people.

They walked out "feeling defeated". They also got the job, and it turned out to be the best company they had worked for.

Read those two facts next to each other. The same day, the same interview, produced the feeling that it had gone badly and the outcome that it had gone well. Whatever an interview feels like from the inside and whatever it decides are not the same variable, and the gap between them is most of what makes a search hard to reason about.

What it actually cost

Three to four hours in a room, plus travel each way, plus a phone screen, plus a second call. For a role advertised at about 65k, and for one of the smaller companies in its market rather than a large employer with a formal process.

Multiply that by the size of a typical search. A candidate with thirty applications might have six interviews in them. At four hours each, that is a day and a half in rooms, before travel, before the phone calls, before the waiting. A candidate with a hundred applications and a ten per cent interview rate has ten, which is three or four days of their life spent sitting in other people's offices being asked about a GPA.

None of that is visible in a tracker that counts applications. An application is a row. An interview is a row. The number that is actually large, the one that competes with a job, with a course, with a life, does not appear anywhere.

What to do about it

Not much, and the honest answer is that the hours are largely fixed. What can change is how many of them you spend.

Count the hours once, properly, and then stop counting applications. The arithmetic above is simple and it is the only version of the question worth asking: how many interviews have you actually had, how long was each, and what did they add up to. A candidate who has never done this has no idea, and the number is usually larger than expected.

Use it as a limit rather than as a score. The person who posted it came away feeling defeated from a process that went well for them. An interview that exhausts you is not evidence that you did badly, and treating exhaustion as a verdict is how a long process produces a short search.

Set the threshold in advance. The three-interview limit that appears elsewhere in this research is the same idea as a rule about information. Beyond a certain depth, additional rounds are costing you a day and telling you something you can act on. Deciding that in advance is cheaper than deciding it at round five, having already paid for rounds one to five.

Notice the shape of the process, not just its length. Four groups, some questions repeated, a COO in the room for a helpdesk role. A process that repeats itself is a process that has been built around a template rather than around the role, and that tells you something about what the round is measuring.

What this product does with it

The tracker records which stages a candidate reported going through, per application, and the search summary adds them up: sent, replied, interviewing, offers, rejected, and how long the silences ran.

That is a count of stages rather than a count of hours, and the difference is deliberate. Converting stages into hours would mean inventing a duration for a stage the product has never observed, and an invented duration is exactly the kind of number this product refuses to produce. The honest version is the one above: count your own hours once, because only you know them.

And the same reason applies to the comparison the product will not make. It is not going to tell you that your search is going well, because there is no baseline to say so against.

The limit of this, stated plainly

This is one reply in someone else's thread. It is not a study, not a sample, and not something that can be checked. It is the only measurement of what an interview process costs a candidate in the whole corpus this product was built from, and one data point is not a rate.

What survives that limit is narrow and still worth having: the hours are larger than the application count suggests, they are invisible in the tools candidates use, and the same interview can feel like a failure and turn out to have worked. Those are arithmetic and experience rather than statistics, and they are the reasons the tracker exists.

Related