Key Person Risk in Marketing: What Happens the Day Your Head of Growth Quits

It's 1:52 a.m. and she's still looking at her phone, brightness turned down so it doesn't wake the room. The MER is fine. 3.4, comfortably above target, the number the board wants on Thursday. Instead of putting the phone down, she's running a calculation that never shows up on any dashboard: what happens to that 3.4 if she isn't the one who has to explain it anymore.

She's VP of Growth at a DTC brand doing eight figures, and she knows exactly why September's creative fatigued, which channel the team abandoned and why, and which retention segment actually drives the number everyone applauds on Monday. None of it is written down. If she quit tomorrow, whoever sat in her chair by Q2 would inherit the login and two years of historical data. They would not inherit the reasoning.

That's the real single point of failure in a scaled DTC growth function. Not the ad account, not the platform. It's the context sitting in one person's head, and it can't be re-hired. It can only be re-learned, at the cost of a quarter's momentum.

Why Growth Ends Up Running Through One Person

Search "head of growth quits dtc" and most of what comes back is backfill advice: post the role, run the search, get someone senior in the seat inside sixty days. That advice assumes the job is a job. For most $30M+ brands, it stopped being one a while ago.

Nobody built it this way on purpose. The brand hit $8M with three people and a shared Slack channel, and the growth function grew the way DTC growth functions always grow: fast, lean, concentrated in whoever was fastest at reading the numbers. At $8M that concentration is a feature, because the whole funnel still fits in one head. By $30M she's running paid social, email, and the agency relationship, still in her head, because nobody stopped to ask when the concentration turned from an efficiency into a liability. The org chart says Head of Growth. What it describes is a single point of failure, wearing a job title as a disguise.

Security teams have a name for this. They call it key person dependency, and they build redundancy plans to survive it. Marketing teams carry the identical exposure and mostly lack the vocabulary for it. This is key person risk, practiced by accident, because the number that would expose it stays healthy right up until the week it doesn't.

What Actually Walks Out the Door

The mistake is thinking the loss is the seat. It isn't. What leaves with a departing growth lead is the substrate under every decision that looked, from the outside, like instinct: which segment converts on urgency and which on social proof, which agency got fired and why, which creative angle is burned, which channel got tried two years ago and quietly dropped.

The next hire inherits the login, not the reasoning. She'll see the same ad accounts, the same Klaviyo flows, the same reporting cadence. She won't see the eleven months of trial and error that produced them, because that rarely gets written down while it's happening. It becomes instinct, and instinct doesn't export.

The people who've run growth at scale say a version of this out loud, usually when the talk turns to agencies. "So many of our brands just gave the keys, and I don't recommend that... the brand has to be more responsible," says Ari Murray, Head of Growth at Salt and Stone. She's talking about an outside firm, but the instinct applies just as well to a single internal hire: a serious brand doesn't let the whole of its growth function live somewhere it can't see, whether that's an agency's account team or one person's unwritten judgment. Keep the keys. That rule doesn't stop being true because the person holding them is on payroll instead of retainer.

The Scramble

When the departure happens, and at that concentration it eventually does, the timeline is predictable enough to put on a calendar. Reopen the req. Run the search, six to ten weeks if the market cooperates. Close the hire, another two to four. Onboard them into a business they've never seen, with no document that explains the decisions that got the funnel here.

Call it a quarter, conservatively. The cost of that quarter is mostly invisible while it's happening, which is what makes it dangerous. Nobody's line item moves. What's happening is a slow bleed under the number: MER drifts, CAC creeps, and decisions that used to take an afternoon now take a week. The new hire isn't incompetent. She's doing, for the first time and in public, the exact learning her predecessor did quietly over three years. By the time the metrics show it, the quarter is gone.

"We Document Things" Isn't the Answer

The first objection comes from mature teams: we have runbooks, we have a wiki, we run QBRs. Documentation exists for exactly this reason, and it helps. It just doesn't solve the problem, because documentation captures decisions, not the judgment that produced them. A wiki page can say "we cut prospecting budget 30% in September." It can't say how the person knew, in the moment, that the signal was real rather than noise. That pattern recognition is the layer that doesn't survive the handoff to a document, which is why a fully documented onboarding still takes a quarter to get to full speed. The runbook tells the new hire what happened. It can't tell her what to do the next time something happens that isn't in it, which, in a channel mix that changes every quarter, is most of the work.

The second objection is structural: hire two people instead of one, spread the exposure. It sounds like redundancy. It's the same problem, multiplied. Now the context is split across two people who can each still leave, and coordination costs real time every week. More headcount doesn't change the underlying property: the knowledge lives inside humans who can hand in notice on a Tuesday. It adds more Tuesdays.

The Fix Isn't Another Hire

The actual fix isn't a person at all. Once the loss is named this precisely, not the seat but the context, the alternative sounds less like a pitch and more like the obvious next move: a continuity layer that holds the reasoning, not just the dashboards, with no resignation date.

That's the gap Sutton is built to close: one compounding brain across your whole funnel, worth more in month 12 than month 1, nothing re-onboards and knowledge doesn't leak out the sides. It carries the encoded judgment of $150M in DTC sales driving 6 exits across our founding team, always with a human on the spend. Not a replacement hire. Not an autonomous media buyer. A second brain for the function that never hands in its notice.

Back to 1:52 a.m.

The phone's still lit, and the honest answer to "what happens if she leaves" was never a faster req or a better job posting. Two people, backfilled fast, still lose the quarter, because pattern recognition isn't a thing you can source in a job interview.

The better question, worth asking long before there's a notice letter on anyone's desk, was never who do we hire if she leaves. It's what did we let live in one person's head that should have belonged to the brand all along.

She puts the phone down. The MER is still 3.4. It just doesn't feel like an answer anymore.