Why your first growth hire fails, and I am usually the second one
Four failure patterns I keep finding when I inherit an account, and what the founder could have done differently in each.
the short answer: most first growth hires fail for reasons that have nothing to do with the person hired. they get brought in to fix something that is not a growth problem, given no authority to change what is actually broken, judged on a timeline shorter than the feedback loop, or hired for a channel the founder had already decided on. i have inherited enough of these accounts to see the same four patterns.
Pattern 1: hired to fix a retention problem with acquisition
the most common by far.
users are not coming back, revenue is flat, and the conclusion is that the top of the funnel needs work. a growth marketer arrives, spends money efficiently, and the numbers do not move because more people entering a leaky product produces more people leaving it.
what it looks like six months later: the hire is blamed for poor performance while the retention number nobody was watching sits unchanged.
what to do instead: check whether a cohort you did not pay for is still active a month later. if not, that is the job, and it is mostly a product job.
Pattern 2: no authority over the thing that needs changing
the growth hire can change ads. they cannot change the onboarding flow, the pricing, the signup form, or the first-run experience, because those belong to product and product has a roadmap.
so they optimise the only surface available to them, which is usually the least leveraged one.
what it looks like: creative gets better, cost per acquisition improves slightly, conversion stays flat because the drop-off is on a screen they cannot touch.
what to do instead: before hiring, decide what this person is allowed to change. if the answer is only ads, hire a media buyer and be clear that is what you bought.
Pattern 3: judged on a timeline shorter than the feedback loop
only about 5% of paid creatives become winners, so finding three means testing twenty to twenty five. at a realistic production rate that is weeks of work before the first reliable signal, and a channel needs about ten days of honest testing to produce a verdict.
a ninety day review is fair. a thirty day one measures luck.
what it looks like: the hire spends month one building a system, gets asked in month two why the numbers have not moved, and starts optimising for visible activity instead of results.
what to do instead: agree what evidence looks like at thirty, sixty and ninety days before they start. write it down.
Pattern 4: hired for a channel that was already decided
the founder read that TikTok works, hires someone for TikTok, and the person’s actual skill, working out which channel fits this product, never gets used.
sometimes the channel is right. often the hire discovers in week three that it is wrong and now has to argue against the premise of their own job.
what it looks like: quiet underperformance in a channel nobody wants to abandon because it was the reason for the hire.
what to do instead: hire for the diagnosis, not the channel. if you already know the channel, you want an executor and should say so.
What I usually find when I inherit the account
a creative library too small to have found a winner. no written kill thresholds, so losing ads ran on sentiment. spend spread thin across four channels, none validated properly. and almost always, a funnel step losing a large share of users that nobody had measured because it belonged to a different team.
none of that is the previous hire’s fault. it is what happens when the work is scoped as “get us users” rather than “find out what is stopping users”.
How to avoid needing a second one
decide what problem you are buying. diagnosis, execution, or capacity. they are three different hires.
give them the funnel, not just the ads. or accept that you are buying media management.
write the evidence milestones before day one.
let them tell you the channel. that is the expensive part of the skill.
if you want an honest read on whether your situation needs a hire at all, let’s talk. the four gates i apply are in do not hire a growth marketer yet.
sources: published 2026 meta creative win-rate benchmarks · adliftr analysis of 500,000+ meta ads.