If you’re running ads for the first time, the goal of week one is to learn, not to win. Almost every first campaign loses money; the difference between people who eventually make ads work and people who rage-quit is what they do with that first week. A measured, deliberate start beats a rushed one.
Here’s a calm, day-by-day shape for it.
Before you spend: three decisions
Make these three decisions before the campaign exists. They take an hour and prevent most week-one mistakes.
- Pick one clear goal. One conversion that matters — a trial start, a purchase, an install that opens the app. Not five at once. Everything you’ll read in your reports next week only makes sense relative to a single goal.
- Set a tuition budget. An amount you would be fine losing entirely, spread across the week. Write it down and don’t touch it mid-week. This week’s spend is tuition, not an investment with an expected return.
- Know your break-even before you start. Decide the ROAS at which a campaign pays for itself for your business — from your margin and, if you’re subscription-based, your retention. Without this line, you’ll judge next week’s numbers by vibes.
The week, day by day
Day 0 — set up properly. Connect your conversion tracking and make sure the event you care about actually fires (do a test signup yourself). Write one or two simple ads with an honest promise — what your product does, for whom. Launch to one audience on one platform.
Days 1–2 — leave it alone. Seriously. The platform is in its learning phase, delivery is unstable, and any number you see is noise. The urge to check hourly is real; checking changes nothing. Build something instead.
Days 3–4 — first read, no edits. Open your reports with the field guide next to you. Walk the funnel: are people clicking (CTR)? Are clicks converting (CPA)? You’re not judging success yet — you’re locating the leak, if there is one.
Days 5–7 — change one thing. If the data clearly points at a leak, fix that single thing: a new creative if CTR is dead, a landing-page tweak if clicks don’t convert. One change, so next week you know what caused any difference. If nothing is clearly broken, change nothing and let it run.
What not to do in week one
- Don’t edit daily. Every significant edit resets the platform’s learning. Panic edits are how small budgets evaporate without teaching anything.
- Don’t add platforms. One platform is a lesson; three platforms is three half-lessons for triple the budget.
- Don’t chase cheap clicks. A falling CPC feels like winning. If those clicks don’t convert, you’re just buying noise at a discount — tie every number back to money.
- Don’t trust the platform’s revenue estimate blindly. Judge against what actually landed in your revenue tool.
At the end of the week: decide
Sit down once, with the week’s data, and answer three questions:
- Did the tracking work? If you can’t trust your conversion data, fixing that is the next week’s plan.
- Where was the leak? No clicks, no conversions, or no profit — each points at a different fix.
- What’s the one change for week two? New creative, new audience, new landing page — one variable, so the next read stays clean.
That loop — read, decide, change one thing — is the whole game. Repeat it weekly and you’ll be making decisions most small advertisers never learn to make.
Where Flowjat fits
A copilot that does the reading for you and proposes the one change is the point of Flowjat: it watches your campaigns against your real revenue and says, in plain words, what’s above your break-even, what’s below, and what to adjust next. Start free and connect your first platform — or read how we keep that connection safe first.