If your company is growing, your hiring process has probably gained stages and interviewers along the way, and it’s fair to ask whether each interview should get longer too.

Interview length decides how much interviewer time every candidate costs. Book a slot too long and you waste interviewer time on every candidate in the pipeline. Book it too short and the interview ends before it has covered what the stage is for.

Before you book longer slots, it helps to know what interview length by company size looks like. The figures below answer that from our Metaview Notetaker’s recordings of completed interviews at customer companies of five sizes, from startups to companies with thousands of employees.

The data is aggregated and anonymized, and each company’s size comes from Salesforce. Every figure is how long the recorded conversation ran, so it measures the interview itself rather than the slot booked on the calendar.

The figures show mid-size and large companies running slightly longer interviews than small ones, and the very largest companies not running the longest ones on average.

Each figure counts every interview stage together, from first screens to final rounds, so I’d leave company size out of the decision. Instead, size slots by stage: list what each stage has to learn about the candidate, and book the time its questions take.

Below are the figures for five company sizes, then a way to size your own slots and check them.

Interview length by company size.

The table below gives the median and the mean interview length for each of the five company sizes.¹ The median interview captured on Metaview ran from 34.0 minutes at customer companies with 200 or fewer employees to 37.8 minutes at those with 1,000 to 5,000.

The median is the length of the middle interview, so it shows what a typical interview looks like. The mean is the plain average, and it’s higher than the median at every size, so the median is the better guide to a typical slot.

What the figures say about bigger companies.

Yes: companies with more than 200 employees ran longer interviews than those with 200 or fewer, by a few minutes at most.

That holds at every size above 200 employees, on the median and the mean alike.

Length doesn’t keep rising with size, though. Companies with 1,000 to 5,000 employees ran the longest interviews on average, and companies with 5,000 or more ran shorter ones than companies with 201 to 5,000.

I wouldn’t plan your slots around gaps of a few minutes. Each figure mixes every stage a company recorded, from screens to final rounds, and you book slots one stage at a time, so the figure that helps you is the one for each of your own stages.

That’s why I’d size slots from the stages themselves, starting with the job each one does.

How to set the length of each interview.

These six steps size each stage’s slot and keep the whole process in check, ending with a test of each slot against the interviews it holds:

1. List what the stage needs to find out about the candidate: the skills it tests, and the questions or exercise that test them.
2. Add up the time those questions take, plus a few minutes at the start for introductions and at the end for the candidate’s own questions. Round up to the block your calendar uses.
3. When your team adds a stage, check first whether an existing stage already covers its questions. If none does, give the new stage its own list and its own slot rather than stretching an existing one to cover new ground.
4. Add up the interviewer time each candidate costs across the whole process: every slot multiplied by the number of interviewers in it. That total grows with every stage or panelist you add, so work it out again each time and ask whether the new stage is worth the extra hours.
5. After a few weeks, check how long stages ran against the slots you booked. Recordings or video-call logs give you that length, while the calendar only shows the booking. Adjust the stage that keeps running over or finishing early, and leave the stages that fit alone.
6. Read the transcripts or notes from a few interviews in each stage as well, because a stage can finish on time and still be too short. If a question on the stage’s list keeps going unasked, the slot may be too short for it.

When one stage keeps running past its slot, we have a separate guide on interviews that run over. It covers how to tell whether the slot or the conversation needs to change.

Where Metaview fits.

Steps five and six both need a record of each interview: how long it ran, and what was asked.

The Metaview Notetaker joins each interview as a visible participant and records the conversation, so it captures every spoken word. That one record gives you both checks: how long each call ran, on the “Calls” page, and what was asked, in the call’s transcript.

1

2

1. 1The “Duration” column shows how long each call ran.
2. 2The line under each name gives the stage, such as a recruiter screen or a system design interview.

The “Calls” page is the view I’d open before changing any slot, because it shows whether a stage runs long for every candidate or only now and then.

When your company is about to grow.

Company size won’t tell you how long to book. If I were planning interviews at a company about to double, I’d leave slot lengths alone until a stage’s own interviews kept running over or finishing early, and I’d watch interviewer time per candidate as the company adds stages and panelists.

Your calendar knows what you booked; only the recording knows what you ran.

Sources.

¹ Aggregated and anonymized Metaview interview data from Salesforce-linked customer organizations, with the median and mean length of interviews for each company size.