The first ninety days of an operating partnership are not a strategy phase. They are the window in which four systems get built, staffed, and reporting — or the engagement quietly reverts to a project that ends the day the site goes live, whatever the original pitch called it.
What ninety days is actually for
The number isn’t arbitrary. It’s roughly the time it takes a new engagement to answer, with evidence rather than intention, the four questions any launch has to keep answering after everyone stops watching: what arrived this week, what got published, what broke, and what changed in the pitch as a result. We’ve written before about why those four questions are the ones that actually determine whether a launch survives its first month — the first ninety days are where the answer to each one gets a system behind it, instead of relying on someone’s memory.
Ninety days is long enough to see a real pattern in each of those four areas — a channel that’s actually converting, a page that’s actually ranking, a form that’s actually routing — without waiting so long that a bad system has already cost a quarter of pipeline.
The four systems that get stood up first
Nothing exotic. In order:
- A number somebody looks at. Not a dashboard configured once and never opened again — a weekly figure, broken out by channel, that a specific person is accountable for reacting to.
- A publishing cadence. Search traffic compounds from surface area added over time, not from what shipped at launch. If nothing new goes up in month two, growth quietly stalls, on a lag long enough that the cause is easy to miss.
- A monitoring layer for the unglamorous failures. Forms that stop delivering, pages that fall out of the index, certificates that lapse — none of these show up in a strategy review, and all of them cost more over a year than any single positioning decision.
- A feedback loop from the market back into the copy. The first version of a pitch is a hypothesis. Whatever prospects actually ask, linger on, or stall at should reach the site within weeks, not at the next planning cycle.
Why ninety days, and not sixty or a year
Sixty days is rarely enough to tell a channel that’s working apart from one that’s just new. A year is too long to wait before discovering a system was never actually reporting — by then the cost has already compounded across a full quarter. Ninety days is roughly one full cycle of the four questions above, run enough times to trust the answer.
That’s also the window we commit to as a standard on the MedGenie side of the business: if what we launch isn’t outperforming the baseline it replaced within that period, measured on terms agreed before the work started, we keep working at no additional cost until it does.
What this first stretch isn’t
- Not a rebrand exercise repeated from the pitch deck.
- Not a set of KPIs chosen because they’re easy to hit.
- Not a hand-off document describing a system nobody has built yet.
What “done” looks like at day ninety
Not a finished company — an accountable one. By day ninety, someone can answer each of the four questions above without checking with anyone else, on a cadence shorter than a month. That’s a different bar than “the site is live” or “the campaign ran,” and it’s the bar our three operating rules are actually built to hold an engagement to, whether the deliverable is a go-to-market launch or an API estate.
Who actually walks through this
The same ninety-day structure applies whether the company entering the market has no revenue yet and the launch is the whole story so far, or already has paying customers and a site that’s been neglected for a year. The difference between an operating partner and a launch agency shows up exactly here — in whether anyone is still assigned to the four questions on day ninety-one.
If you’re evaluating what the first quarter of an engagement should actually produce, tell us what you’re building and we’ll walk through what week one, month one, and day ninety look like for it.