Meta ads: the first thirty days
What we do in each of the first four weeks of a Meta ads engagement, why we start with ten creatives, and what the numbers can and cannot tell you by week four.
A new ads engagement is judged on its first month, which is fair, as long as both sides agree in advance what a month can prove. This is what the first thirty days look like with us: week by week, what we do, what you see, and what we refuse to promise.
Week one: what happens on the call?
The first call is thirty minutes and there is no deck. We ask for the number you need, the margin on what you sell, and what a sale actually is for you: a checkout, a booking, a message, a cash-on-delivery order that may be refused at the door. Then we ask to look at the ad account, if there is one, before we say anything about what we think we can do.
The rest of week one is access and tracking. Your ad account, in your business manager, with us added as a user you can remove. The pixel and the conversion events checked against real orders, because an account optimising toward an event that fires twice per purchase is optimising toward fiction. If there is no account, we set one up in your name, on your card, and hand you the keys the same day.
We also agree, in writing, what the first report will measure. Cost per result against the number from the call. Not reach, not clicks, not video views, unless one of those is the result.
Week two: why ten creatives?
Because we do not know which one will work, and neither does anyone else. An agency that launches with one hero ad is betting your first month on a guess. Ten creatives, each saying one thing in one format, let the platform and the audience tell us which argument lands. Some are a product on a plain background with one line. Some are a short video. Some are a question. The point is range, not polish.
- The product, plainly, with the price.
- The product in use, filmed on a phone, with captions for muted viewing.
- The objection, answered: what it costs to ship, how long it takes, what happens if it is wrong.
- The question that starts a conversation, for a brand that sells by message.
- The comparison, when the buyer is choosing between two things that look alike.
The second job of week two is the page. The ad is a promise and the page has to keep it, on a phone, in a few seconds. If the sale closes in a conversation, as it does for a brand that sells personalised products, the page is the inbox, and the creative is built to start a message rather than a click. Either way, nothing goes live until the place the ad sends people has been looked at by someone who is not us.
Week three: what goes live, and how much do you spend?
The ads go live in week three, on a small budget. Small means enough to produce results we can read, not enough to hurt if the first week is wrong. We would rather spend a modest amount learning which three creatives deserve the budget than spend the whole month proving that seven of them did not.
The structure is simple on purpose: a handful of ad sets, broad audiences where the platform is good at finding buyers on its own, a retargeting set for people who have already met the brand, and the same ten creatives running against each. We check the account every morning your time. We do not touch it every morning, because the platform needs a few days of stable settings to learn, and an account that is adjusted daily never gets to.
What you see in week three is the ads themselves, in your own account, and a short note on what is running and why. What you do not see yet is a verdict. Three days of data is not a verdict, and anyone who gives you one is reading noise.
Week four: what do the first numbers say?
By the end of week four there is a real report. It says what was spent, what came back, which creatives earned their budget and which were cut, and the cost per result against the number we agreed in week one. It is written in sentences, and the numbers in it come from your account, which you can open and check.
What it cannot say is whether the engagement works. A month is enough to find the first creatives that sell and the first that do not. It is not enough to know what the account does in a holiday week, or whether the cost of attention rises when more advertisers arrive, or how a second batch of creatives performs against the first. We say that on the call, and we say it again in the report, so that a good first month is not mistaken for a finished one.
Why are losers cut every week?
From week four on, the account runs on a weekly rhythm. The creatives that cost more per result than the number we agreed are paused. The ones that beat it get more budget and get rebuilt: the same argument, a new opening, a new format. New creatives enter on a schedule, because every winning ad eventually tires, and the account needs its replacement ready before that happens rather than after.
This is the part of ads management that is unglamorous and that matters most. The launch is one week. The weekly cut is every week for as long as you are a client. Ask any agency how often it pauses an ad and what rule it uses to decide. The answer should be a threshold and a day of the week.
| Week | What we do | What you see |
|---|---|---|
| One | The call, the number, the access, the tracking | A written note on what the first report will measure |
| Two | Ten creatives, and the page or the inbox | The creatives, for approval, in your voice |
| Three | Live on a small budget, broad plus retargeting | The ads running in your own account |
| Four | The first report; the first weekly cut | Spend, return, cost per result, what was paused and why |
What did this produce for The Gallery?
The Gallery sells personalised newborn wear in Pakistan. Every order needs a conversation: a name, a spelling, a date, sometimes a photograph of the nursery. The checkout was already the direct message, so the campaign was built to start conversations rather than to send people to a product page, and it was optimised toward the inbox.
Ten creatives, cut weekly. The result, counted where the sale actually closed, was $5.20–$6 back per $1 of ad spend, every campaign to July 2026. Not an average across one good quarter: every campaign. The spend behind that return is not published, and we will not quote a return for your account before we have seen it. Nobody who guarantees a number is telling you the truth.
Why does the spend stay in your own account?
Because the account is where the value accumulates. The pixel data, the audiences, the creative history and the results all live in it, and whoever owns the account owns them. If the agency runs your ads from its account, leaving means starting from nothing. If you own it, leaving means removing one user.
It also keeps the money honest. Your card pays the platform directly. Our fee is flat and separate, and we never take a percentage of what you spend, because an agency paid on spend is paid to spend. The cost to you is that you hold the keys: you approve the budget, and you will see every dollar leave. We think that is how it should feel.
The Starter package, at $2,000 a month, is Meta ads on one account with four new creatives a month, optimised weekly, with a monthly written report and a monthly call. Larger budgets and more than one platform are priced to the number. Either way, the thirty days above are the same thirty days, and the report at the end of them is the one you judge us on.