A founder signs with an agency, pays the first invoice and waits. Six weeks later there is a kickoff deck, a brand questionnaire, a shared folder and a standing call, and nothing a customer has seen. That is the most common shape of the first quarter, and it is the one to refuse before it starts.
Ninety days is long enough to ship three things.
By day ninety a startup working with an agency should have three things it did not have on day one. A positioning sentence the founder can say out loud without notes. A website that states it and asks for the next step. One channel running with a number attached to it, even a small one. Everything else in the first quarter is either in service of those three or it is overhead.
That list is short on purpose. A founder with a seed round and a small team cannot absorb ten workstreams at once, and an agency that proposes ten is planning for its own utilisation, not the company's first year. Three finished things beat ten started ones, because a finished thing can be judged and a started one can only be discussed.
The first month is positioning, and it should end in a sentence.
Weeks one to four are for deciding what the company says. That means interviews with the founder, a few customers and, if there is one, the person who sells. It does not mean a sixty-question brand survey. The output is a sentence and a short argument behind it, the kind described in positioning is the sentence your startup must earn, and the founder signs off on it in writing.
If month one ends without that sentence, nothing after it will hold. The site will be written around a guess, the channel will be tested against a guess, and the second quarter will be spent rewriting both. A missed positioning deadline is the one delay worth stopping the whole engagement for.
The second month is the site, because the site is where the sentence lives.
Weeks five to eight turn the sentence into pages. For most early companies that is a handful of pages and one form, not a sitemap with forty entries. The site goes live in month two, not "when it is ready", because a live site with one weak page teaches more than a staging site with none.
This is also the month a founder learns how the agency handles disagreement. The first round of copy will be wrong somewhere. An agency that defends every line, or one that changes every line the moment it is questioned, is showing you how the next year will go. The useful version argues once, with a reason, and then moves.
The third month is one channel and one number.
Weeks nine to twelve pick a single channel and run it properly. Which one depends on the company: the order a startup should try channels in is a better guide than whatever the agency happens to be strongest at. The point of the month is not scale. It is a real number from a real test, read against the one number marketing is accountable to, so the second quarter starts from evidence rather than taste.
A founder should expect that number to be modest. Ninety days is enough to learn whether a channel is worth more money. It is rarely enough to make it pay back.
What day ninety should feel like for the founder.
At the end of the quarter the founder should be able to answer three questions without opening a deck. What do we say. Where do people read it. What happened when we told more of them. If any answer is "we are still working on it", the engagement is late, and a good agency will say so first.
The other sign is time. The founder's hours on marketing should be going down by month three, not up. The first month needs them most, for interviews and sign-off. By the third, the agency should be bringing decisions, not questions. If the founder is still writing briefs in week eleven, the reason a startup outsources marketing has not been met yet.
Hold an agency to those three outputs, not to a list of activities.
Tell us what you're building.