The first 30 days, before I touch your ad account
The unglamorous diagnostic sequence. What gets read in week one, what gets built in week two, and why nothing gets spent until week three.
the short answer: the first month of a growth engagement should produce a diagnosis and a system, not results. spending in week one means spending before you know what the message is, and the cost of that mistake compounds for a quarter. here is the sequence i actually run.
Week one: read everything
the funnel, step by step. not aggregate conversion. the specific percentage completing each step between first click and first moment of value. the biggest cheap win in most engagements lives here and it is usually a step nobody had measured because it belonged to a different team.
the ad account history. every campaign ever run, what it cost, what it produced, when it was killed and by whom. the creative graveyard is the most useful document in the account, and almost nobody has looked at theirs.
the retention data. whether a cohort you did not pay for is still active a month later. for subscription products, specifically the share of first-month subscribers making the second payment. that number decides whether acquisition work is worth doing at all.
the support inbox and the reviews. the words customers use for the problem. this is ad copy research and it is free.
Week two: map the channels
where your users already are. subreddits, communities, search terms, competitor mentions. done manually, because reddit’s API is effectively closed to newcomers since late 2025 and the research tooling most guides assume no longer exists.
the creative angles worth testing. taken from the language collected in week one, not invented in a workshop.
the test order and the gates. which channel first, what pass looks like, what fail looks like, and the budget for each. all written down before anything spends.
that last part is the difference between testing and gambling, and it is the thing most engagements skip.
Week three: ship the first batch
briefs go out to the creator bench. first creatives get produced. the first tests go live against the thresholds agreed in week two.
only about 5% of creatives become winners, so this batch is deliberately wide rather than precious. the goal is not a winner in week three, it is a real read by week five.
Week four: the first honest report
what we tested, what won, what died, what changes next. numbers, not narrative. if the week went badly it says so.
by the end of month one you should have: a diagnosis of where users are actually lost, a documented testing system, one batch of results, and a decision about which channel gets month two.
what you should not have is a claim that acquisition cost has improved. one month is not long enough to know that honestly, and anyone reporting it is reporting noise.
Why not spend on day one
because paid amplifies a message and does not find one.
running ads before you know the language, the funnel leak, and the retention picture means paying to discover things that a week of reading would have told you. i have never seen the fast start pay for itself, and i have inherited several accounts where it cost a quarter.
What this means for judging the engagement
thirty days buys you a diagnosis and a system. sixty days buys a validated or invalidated channel. ninety days buys a defensible answer about whether this is scalable.
agreeing that timeline in advance is the single best thing a founder can do to make a growth hire work, and its absence is the most common reason they fail. i wrote about the rest of those patterns in why your first growth hire fails.
if you want the week one read done on your product, let’s talk. that first conversation is a compressed version of it and it is free.