What is the minimum ad spend to work with a performance creative agency?

It is the question we get asked earliest and most often: “Do you have a minimum contract time, a minimum, you know, budget?”

Yes. $50,000 a month in ad spend. Not a contract minimum, a spend minimum, and the distinction matters. We are not saying smaller businesses are less worth serving. We are saying that below roughly that number, the thing we do cannot work, and taking the money anyway would be taking money for a process that cannot run.

Here is the arithmetic, so you can check it against your own numbers rather than take our word for it.

What a creative programme actually needs

The work is not making ads. It is finding out which ads work, which means running tests that produce a readable answer.

A creative under test needs its own ad set, and that ad set needs enough budget to generate two to three conversions a day. Below that you are not reading performance, you are reading noise: the day-to-day variance is larger than the difference you are trying to detect, and you will confidently cut winners and scale losers.

So the budget one test needs is:

Daily budget per test = your CPA x 3

And the number of tests you can run at once is:

Concurrent tests = daily testing budget / (CPA x 3)

That is the whole model. Everything else follows from it. We set out where the three-conversions-a-day floor comes from in how many ad creatives you should test per week, and why it sits deliberately below Meta’s fifty-conversion threshold in do you need 50 conversions per week.

Why that lands around $50,000

Work it through at a $40 CPA, which is ordinary for a subscription app.

One test needs $120 a day, about $3,650 a month. To learn anything week to week you want several concepts running at once, not one, because a single test tells you whether one idea worked and nothing about which direction to go next. Call it eight to ten concurrent tests as a working floor for a real programme.

That is roughly $30,000 to $36,000 a month in testing budget alone. Then you still need budget to scale the winners, which is the entire point of finding them. Add that and you are at $50,000 before anyone has been paid a fee.

Below that, one of two things happens. Either you run fewer tests and learn too slowly for the programme to pay for itself, or you underfund each test and get answers you cannot trust. The second is worse, because it feels like progress.

The number moves with your CPA

This is the part a flat minimum hides, and it is why you should not simply take our $50,000 as gospel for your own situation.

A hypercasual game at a $2 CPA can run a serious creative programme on a fraction of $50,000. A fintech app at a $200 CPA needs considerably more than it. Our floor reflects the CPAs our clients typically run at, not a universal law.

So the honest version of the minimum is: your testing budget needs to support roughly ten concurrent creative tests. Multiply your CPA by three, multiply by ten, multiply by thirty. If that number is comfortably under your monthly spend, a creative programme can work for you. If it is not, it cannot, whoever you hire.

What we would tell you to do instead

If you are below the line, we would rather say so on the first call. What actually helps:

• Spend less on more concentrated tests. Four well-funded tests beat twelve starved ones. Starved tests do not produce cheaper answers, they produce wrong ones.
• Test bigger swings. With few shots, spend them on genuinely different concepts rather than variations of the same idea. Small differences need large samples to detect. Large differences do not.
• Buy tools, not people. Below the threshold, your constraint is budget, not production capacity.
• Be wary of a cheap retainer. Below the line, the offers that will accept you tend to be staffed by more junior people. The budget problem does not go away; it just gets managed by someone with less pattern recognition.
• Come back when the arithmetic works. Several of our clients did exactly that.

And on pilots, since it is the next question people ask: two months is the shortest engagement that tells you anything. Shorter looks appealing and does not work, because a creative programme needs a full cycle to test, read, iterate and scale. A one-month pilot buys you the testing and none of the learning.

Why we publish this

Because the alternative is a discovery call where we work out in minute forty that the numbers were never going to add up. You have lost an hour and so have we.

And because the reasoning is more useful than the number. If you take one thing from this page, take the formula rather than the $50,000: CPA x 3 x the number of tests you want running. That tells you what a creative programme costs to run properly, whether you hire us, hire someone else, or do it yourself.

This post is part of our guide to how to choose a mobile UA agency.

Frequently asked questions

What is the minimum ad spend to work with a performance creative agency? Ours is $50,000 a month. The real threshold is set by your CPA: your testing budget has to support roughly ten concurrent creative tests, where each test needs about three times your CPA per day. At a $5 CPA that is far less than $50,000. At a $200 CPA it is far more.

Why do performance creative agencies have minimum spends? For creative work, because testing needs enough budget per test to produce a readable result. Below roughly two to three conversions per day per test, results are noise, and the programme cannot do its job however good the creative is.

Can I work with a performance creative agency on a $10,000 monthly budget? For a full testing programme, usually not, unless your CPA is very low. You would be better served concentrating budget into a handful of well-funded tests, or buying production help rather than a managed programme. Agencies that accept a budget that size for a full programme are usually staffing it with junior people, which is the real cost.

How do I calculate my own minimum? Multiply your CPA by three for a daily per-test budget, multiply by the number of tests you want running at once, then by thirty. That is your monthly testing budget, before any spend on scaling the winners.

Is the minimum about contract length or spend? Spend. It is about whether the account can generate enough signal to run a creative programme, not about locking you in.