Your first growth hire, and which stage they should own

Why is the first growth hire so often the wrong one?

Because the job description gets written from the loudest symptom rather than from the funnel. Spend is up and demos are flat, so the company hires someone to run spend. The new person arrives, discovers that the landing page was written by a founder in 2023, that nobody agrees what a qualified lead is, and that two systems report different numbers for the same week. They spend their first quarter on that, which is the right work, but it is not the work they were hired for or judged on, and by month five both sides are disappointed.

The sequence below exists to prevent that. It takes about a day.

Step 1: find the stage with the worst ratio

Write five numbers for the last 90 days, each the ratio between two steps you already record.

• Acquisition: qualified leads per month, and what each one cost.

• Activation: of the people who arrived, what share reached first real use.

• Retention: of a month's cohort, what share is still active at month three.

• Referral: what share of new customers came from an existing one.

• Revenue: of the people who reached first value, what share paid, and how long until a customer paid back what it cost to win them.

Then compare each one against your own history, not an industry figure. Published benchmarks for these are wide enough to justify any conclusion you want: cost per lead alone runs from $26.84 to $131.63 by industry inside a single dataset (LocaliQ, updated June 1, 2026). Your own last four quarters are a tighter comparison and nobody can argue with them.

What goes wrong: three of the five numbers do not exist, so the stage that looks worst is just the stage you measure. Write down which ones you had to guess. That list matters in step 2.

Step 2: check whether you are allowed to believe the diagnosis

If two of the five numbers had to be guessed, or if marketing and sales would report different figures for the same week, stop. You cannot diagnose a funnel through a broken instrument, and hiring against a wrong diagnosis costs a year rather than a quarter.

In that case the first hire is the measurement person, whatever the symptom is. This feels like a detour to every founder I have said it to. It is the shortest route, because every other hire's first ninety days will otherwise be spent on the same problem with less authority. The groundwork is in attribution without the arguments.

What goes wrong: a founder accepts this, then hires a junior analyst to produce reports rather than someone who can change how events are collected and what the CRM records. Reporting on bad collection produces prettier arguments.

Step 3: match the stage to one of three kinds of person

There are really only three first hires, and the titles people give them vary more than the jobs do.

The channel operator. Owns acquisition. Lives in the ad platforms, the creative brief and the budget. Judged on cost per qualified lead and volume at that cost. What they need to exist before they start: a landing page that converts at a rate you know, a written claim the ads can make, and a definition of qualified with a field in the CRM. Hire them when all three exist, and they will be productive in three weeks. Hire them when none of them exists, and they will build all three badly while being measured on something else.

The conversion and lifecycle builder. Owns activation and retention. Writes and ships pages, onboarding, flows and emails, and can read a test result. Judged on the rate from arrival to first value, and on the month-three cohort. What they need before they start: enough traffic to read a result on. Below a few thousand relevant visits a month they will be guessing, and the work belongs with the founder and a researcher instead.

The measurement person. Owns the numbers across every stage. Fixes event collection, defines the fields, builds the one report, and ends the argument about whose number is right. Judged on whether the weekly review stops being about methodology. What they need before they start: access, and a founder who will enforce the definitions they write.

A fourth kind exists, the generalist who owns the joins between stages, and that is a second or third hire rather than a first. They are worth a lot once there are two specialists whose work has to meet in the middle, and they are expensive to misuse when they are the only person, because a generalist with no specialists becomes a project manager for freelancers.

What goes wrong: hiring one person to be all three. The job description with ads, lifecycle, analytics and content in it attracts people who have done a little of each and never owned a number, because the people who have owned a number apply for jobs that name it.

Step 4: decide between full time and brought in

Hire full time for work you will need for years and can judge. Bring people in for work you are still learning the shape of.

The practical test: can you write down, today, the number this person owns and what good looks like on it in ninety days. If you can, hire. If you cannot, you are buying time to find out, and that is a shorter engagement with a specialist, not a permanent role filled on hope.

What goes wrong: the role is filled full time before the company knows which channel works, so the team cannot change as fast as the learning does, and the first hire's specialty quietly becomes the company's strategy.

Step 5: write the scorecard before the job description

One page. One number they own. Three things that will exist in ninety days that do not exist today. The names of the people whose cooperation they will need. And the thing they are explicitly not responsible for, which is the line that saves the relationship in month four.

An acquisition hire's version reads something like: you own cost per qualified lead, on the definition in this document, which is at a level we both agree on today. In ninety days there is one channel producing qualified leads under the ceiling for two consecutive months, a creative test running two weeks ahead of spend, and a weekly report that sales reads without disputing it. You do not own the product's onboarding.

What goes wrong: no scorecard, so the review in month three is a conversation about effort and attitude. Or a scorecard with four numbers on it, which is the same as none.

Step 6: interview for a number they moved, and one they did not

Three questions do most of the work.

Ask for a number they moved, with the before, the after and the time it took. Then ask what else was happening in the company at the same time. The honest answer always includes something else that was happening, and the people who claim sole credit for a number are telling you how they will report to you.

Ask for a number they could not move, and what they concluded. This is the question that separates operators from presenters.

Then show them your own funnel, the five numbers from step 1, for twenty minutes, and ask what they would do first. You are not looking for the right answer. You are looking for whether they ask what qualified means before proposing anything, and whether they notice the two numbers you had to guess.

What goes wrong: hiring on channel brand. Someone who ran paid at a company with a strong product and a large budget may never have had to find out whether the channel worked.

When is the answer not a hire at all?

When the constraint is response time rather than volume. If qualified leads already sit for hours before anybody calls them, more of them makes the problem bigger. The strongest evidence on this is old: across more than 15,000 leads and 100,000 call attempts at six companies, the odds of qualifying a lead fell by 21 times between a five-minute response and a thirty-minute one (Lead Response Management Study, James Oldroyd, MIT Sloan, with InsideSales.com, 2007). A week of plumbing beats a quarter of recruiting, and lead routing without a person in the middle is where that week goes.

On what to pay: I am not going to publish a range, because the ones I have seen vary by more than a factor of three across markets, stages and whether the role carries budget authority, and a number without those conditions attached is worse than no number. Ask three people who have hired the same role in your city and stage this year.