Growth Lead

Every team has ideas about growth. Marketing wants a bigger ad budget, the product team wants the first screen rebuilt, engineering wants to fix how slowly the signup page loads. None of the three is wrong, so the meeting runs long and the loudest or most senior voice wins. The teams that lost still push their own version quietly, so by the end of the quarter all three are half-finished.

The meeting that follows a flat quarter always has the same shape. Asked why signups didn't grow, marketing says the budget was too small and the product team says the redesign it requested got pushed. Because all three things happened at once, nobody can separate what worked from what didn't. No one lied, and nothing was learned.

The growth lead role exists to put that decision in one person's hands. They fix, in numbers, where the product is leaking, narrow this cycle down to one place, and write down when the rest gets revisited. The job sits outside any single team because the judgment can't be made from inside one. Ask the person who owns advertising and the answer always comes back pointing at advertising, not because anyone is protecting territory but because that is the part of the product they can see.

The day usually opens with numbers. Signups, first use, return visits and payments through yesterday get compared against last week, and any step with a visible dip gets opened up. The frame for that reading is AARRR, which splits the path a user travels into acquisition, activation, retention, referral and revenue and counts how many people clear each step. Cut that way, problems appear that a single aggregate number hides completely. Ten thousand signups with flat revenue says nothing about what to do; ten thousand signups, 1,200 first documents and 900 people back a week later narrows the work to one step.

The rest of the time splits roughly four ways.

1. Ordering — candidates in the experiment backlog get ranked again. Likely size of the effect, confidence that it works at all, and days of build time all go on the table together. A six-week candidate loses to three two-day ones.
2. Saying no — a large share of the actual hours go here. Around thirty ideas surface each quarter and five or six can run, so twenty-five of them go back with a "not this time" and a reason attached.
3. Sharpening the hypothesis — each chosen candidate gets compressed into one sentence covering what changes, which number moves, and why. A candidate with an empty reason field is unreadable when the result lands, so it waits for the next cycle.
4. Calling results — the outcome of each growth experiment gets read as a win or a loss. Refusing to read an ambiguous result as a win is part of this job. Losers get deleted on the spot and winners stay.

Once a quarter a deliberately large swing goes on the list. It's a guard against a local maximum, the state a team reaches by repeating small improvements until it misses a bigger jump. Those candidates carry low confidence, so ranking always pushes them down; without a reserved slot they never come back around.

The closest working pair is the growth engineer, who builds each approved test and deletes the losing side once the result is in. The round-trip speed between these two decides directly how many cycles fit in a quarter. Put the pair together with a designer and an analyst and you have a growth team.

With the growth marketer the overlap is at the top of the funnel. Bringing people in belongs to the marketer; which step those people fall out of belongs to the growth lead. When 1,000 arrivals produce only 80 people who create a first document, the conclusion that the ads are fine and the next screen isn't comes from this seam.

With the product team the work needs coordinating wherever it collides. One side wants to add a feature to a screen while the other wants to strip fields off it, often enough that many companies settle at the start of the quarter who touches which screen.

Leadership gets two things each quarter: how many tests ran and how many won, and which numbers moved as a result. That report only holds up when growth analytics is in place. A change nobody instrumented has no way to prove it did anything, and a quarter of work then reads as an opinion about whether things feel better.

"We're only looking at activation this quarter. Acquisition reopens next quarter."

The target has been narrowed to one step, so other proposals aren't being taken right now.

"It isn't killed, it's ranked below the line. It's twelfth in the backlog."

Letting someone know this was a priority call rather than a rejection.

"I can't read that as a win. The gap is inside the noise for this sample."

The difference could still be chance, so it won't be rolled out on that basis.

Pagedeck created the role once the company passed forty-five people. Before that, Grace set the growth items herself while drafting the quarterly plan, and the problem was that whatever got decided stayed fixed for the whole quarter. Reading the numbers and reordering the list every two weeks turned out not to work as a second job.

Pagedeck's growth lead now spends an hour every two weeks with Ravi and Theo. They read the results of the two tests that ran, roll the winner out to everyone, delete the loser, and pick the next two off the backlog. If Mia is reworking the same screen, the sequencing gets settled in that room.

The most common assumption is that the job is about generating ideas. Someone who owns growth sounds like the person who brings new things to try. In practice half the work runs the other direction. Other people supply plenty of candidates; what's left is turning down twenty-five of them with a reason. A lead who can't do that ends up touching six things a little bit each quarter, and then which one worked never becomes clear — the same fog the role was created to clear.

The word "lead" in the title also gets read as people management. At many companies the role starts with nobody reporting into it. What gets led here is the direction of the testing, not a headcount. Whether management responsibility comes attached varies by company, and its absence doesn't mean the decision-making power is small.