The first 90 days: how we approach a foundation rebuild.

There is pressure on every new engagement to show a quick win. Most agencies respond to it by reaching for tactics in the first week, a bidding tweak here, a new ad variant there, something visible they can point to. We have learned to resist that, because tactics applied to an account with broken foundations do not compound. They just move the noise around and buy a month of false comfort.

Our first 90 days on any account are foundation work. It is less immediately impressive and far more valuable, because everything we do afterwards lands on solid ground. Here is roughly how that sequence runs.

Days 1 to 30: see the account clearly

The first month is diagnosis. We separate brand from non-brand so we can see the true cost of acquisition. We audit every conversion action and write down exactly what each one counts. We map where spend is actually going, which almost always surfaces a meaningful slice of budget that nobody can defend. We do not make big changes yet. We make the account legible.

• →Brand and non-brand performance separated and quantified
• →Every conversion action audited and documented
• →Wasted spend identified across geographies, audiences, and stale campaigns
• →Measurement gaps mapped, including whether server-side tracking exists

Days 30 to 60: fix the foundations

The second month is repair. We correct conversion definitions, which often makes the reported numbers temporarily worse and the real numbers better, because the optimization is finally pointing at something true. We cut the indefensible spend. We restructure campaigns so the architecture matches the business. We get server-side tracking in place if it is missing. This is the month that determines whether anything afterwards will work.

Days 60 to 90: begin to build

Only in the third month do we start the work most people think of as the job. With clean data and a sound structure, bidding strategy becomes meaningful, creative testing produces signal you can trust, and audience expansion compounds instead of leaking. The same tactical moves that would have done nothing in week one start to work, because the ground beneath them is finally solid.

This sequence is not glamorous and it is not what most accounts are sold. It is, however, why the accounts we take on tend to keep improving long after the initial cleanup, rather than plateauing once the easy wins are gone. Foundations first is slower to look impressive and far faster to actually compound.