For founders

How long does it take to hire an engineer?

Five to seven weeks is the number most teams plan around. Here is where those weeks actually go, which of them you control, and why time to hire is the wrong thing to be measuring.

Zelcast

Filling a senior engineering role takes about five to seven weeks on average, measured from opening the search to a signed offer, and another two to twelve weeks on top if the person has notice to work. Most of that time is not interviewing. It goes into finding people who are any good, and then into the gaps between steps, where you wait for replies, for calendars, and for a decision from someone who is travelling.

Where the weeks actually go

Writing the role takes longer than founders expect, because the first draft describes a job the company does not have yet. A week disappears here, often more, and it disappears again later if the draft was wrong.

Sourcing is the long pole. If you are starting from an empty list, you are finding names, reading their work, and writing to strangers who are employed and content. Most working developers are employed and not actively looking, which the Stack Overflow developer survey finds year after year, so reply rates on cold outreach are low enough that the arithmetic needs volume, and volume takes weeks. We wrote about why those messages go unanswered separately.

Screening and the loop are quick in theory. In practice each step waits on a calendar. Four interviewers, one candidate with a full time job, one week per round if you are organised, and the loop alone runs two to three weeks.

The offer stage adds days that nobody plans for. References, a compensation conversation, a counteroffer from their current employer, a weekend to think.

The part you do not control

Notice periods are not negotiable and they are not your problem to solve. Two weeks is normal in the US, a month or three is normal elsewhere, and none of it counts as time to hire while it looks exactly like a delay to the person waiting for the work to start.

For a roadmap, the date that counts is the start date. A search that closes in four weeks and starts in twelve has taken twelve weeks.

Why time to hire is the wrong clock

Time to hire measures how fast you filled a seat. It says nothing about whether you filled it with the right person, and it rewards exactly the wrong behaviour: lowering the bar in week six because the number is getting embarrassing.

The number to watch is time to the right person. That number is dominated by one thing, which is whether the people you want are reachable when you need them. If they are, the rest is scheduling. If they are not, no amount of process discipline fixes it.

What actually compresses it

Knowing the people already. Every week saved comes out of sourcing, and sourcing only collapses when someone already knows who is good and can call them. It is the whole reason introductions beat searches on time.

A short loop with a decision maker in it. Three conversations that end in a decision beat six that end in a debrief. If a founder is not in the loop at an early stage hire, the loop will take longer and decide less.

Deciding before you start. Agreeing what the role is, what you will pay, and who says yes removes most of the delay that shows up later as thinking time.

What fast looks like

When the people are already known, an introduction reaches an offer in days. Our own average from introduction to offer is seven days, and the fastest is one. That comes from already knowing the people. No amount of process discipline gets a cold search there.

The industry average of five to seven weeks is what it costs to find a stranger. Five to seven weeks is what finding a stranger costs. Hiring costs less when the person is known in advance.