BUDGET · Meta
Work out whether your Meta ads budget is big enough.
Enter your target cost per acquisition, your margin and the budget you have in mind. It tells you whether that budget clears the learning-phase minimum, how long learning will take, and what happens to the plan if your real CPA comes in higher than you hoped.
Built by Jonas Sluijs, former Meta growth lead, $500M+ in ad spend.
What this is not. It does not connect to your ad account and cannot see what you actually pay per conversion. It works from the CPA you type, so the output is only as good as that estimate. If you have never run ads, use a range rather than a single number and plan for the pessimistic end. For a budget built on your real account, book a free strategy call.
The arithmetic this runs
Meta wants roughly 50 conversions per ad set per week to leave the learning phase and deliver stable costs. That single rule sets a floor under every Meta budget, and the floor is your cost per conversion multiplied by 50, divided by seven, to get a daily figure per ad set.
At a €20 CPA that floor is about €143 a day for one ad set. At a €200 CPA it is €1,430 a day for the same single ad set. This is why the same budget that works comfortably for a low-priced product fails completely for a high-priced one, and why splitting a small budget across several ad sets is usually the worst thing you can do with it.
The second number is your allowable CPA, which is gross margin per sale rather than price per sale. If those two numbers cross, so the budget only clears the learning floor at a CPA above what your margin can pay for, there is no campaign structure that fixes it. The answer is a cheaper conversion event, a higher-margin offer, or waiting.
What the floor looks like at different CPAs
| Your cost per conversion | Rough daily floor for one ad set |
|---|---|
| €10 | About €70 a day |
| €25 | About €180 a day |
| €50 | About €360 a day |
| €100 | About €715 a day |
These are floors for a single ad set, not recommendations. Two ad sets need twice as much. This is the arithmetic behind learning limited.
Common questions
Does this connect to my ad account?
No. There is no login, no OAuth and no ad-account connection. It works from the numbers you type on this page. Nothing is read from your account and nothing is written to it.
What is the minimum budget for Meta ads?
There is no single figure, because the minimum is set by your cost per conversion rather than by Meta. An ad set needs roughly 50 conversions a week to leave the learning phase, so the floor is your CPA multiplied by 50 and divided by seven for a daily number. A cheap conversion makes that floor small; an expensive one makes it large. The minimum budget guide works through it.
My budget is below the floor. What should I do?
Concentrate rather than spread. One ad set at the floor will outperform three ad sets at a third of it, because only the first can finish learning. You can also optimise for a cheaper event higher in the funnel, such as add to cart rather than purchase, so conversions arrive often enough to learn from. Do not run several ad sets on a small budget and hope one wins.
Should I raise budget or duplicate the ad set?
Raising budget on a proven ad set keeps its learning if the increase is gradual. A large jump pushes it back into learning, where costs rise until it stabilises. Duplicating starts a fresh learning phase but leaves the original intact. Duplicate or increase budget covers when each is right, and scaling kills performance covers what goes wrong.
How long before the budget shows results?
Plan on the learning phase completing first, which takes as long as it takes to gather roughly 50 conversions in an ad set. Judging performance before that is reading noise. Attribution adds more delay, since conversions can be reported one to three days after the click. Attribution windows explains the reporting lag.