"How much do I start with?" is the first question almost everyone asks before their first campaign, and the answer that circulates — "a couple hundred pesos a day is enough to test" — is the one that has burned the most money in small businesses. Not because a couple hundred pesos is a small amount, but because below a certain threshold Meta's system never learns, and a campaign that doesn't learn spends exactly like one that does.
The good news is that the threshold is calculated, not guessed. It depends on one thing: what it costs you to get a result.
Why doesn't a small budget work on Meta Ads?
When you turn on an ad set, Meta doesn't yet know who to show it to. It tries different profiles, times and placements, and measures who responds. That stage is called the learning phase, and while it lasts, performance is unstable and the cost per result is higher than it will be later.
Notice what the rule says: 50 results per week, not 50 clicks and not 50 pesos. If you optimize for messages, that's 50 messages; if you optimize for purchases, 50 purchases. And that's the problem with small budgets: at $150 a day with a result that costs $60, you collect 17 in a week. A third of what's needed.
Meta does let you run a campaign on much lower amounts — it has its own technical daily minimum, which varies by billing type — and that's where the confusion comes from. The minimum you're allowed to spend is not the minimum that works. One is a billing-system limit; the other comes from how much data the algorithm needs to stop guessing.
How do you calculate the minimum Meta Ads budget?
The 50-event rule gives you a one-line calculation: your minimum daily budget is what a result costs you, times 50, divided by 7.
It's the same idea behind why your ads get clicks but no customers: the metric that matters isn't cost per click, it's the cost per result that actually pays you. Here that metric also decides your budget.
How much per month, based on your cost per result?
This table is the formula already solved. Find your cost per result in the first column — the rest is arithmetic, not a market estimate. (Figures in Mexican pesos, but the ratio holds in any currency.)
| If a result costs you… | Minimum daily budget | Per month (30 days) |
|---|---|---|
| $20 | $143 | ~$4,300 |
| $40 | $286 | ~$8,600 |
| $60 | $429 | ~$12,900 |
| $100 | $714 | ~$21,400 |
| $200 | $1,429 | ~$42,900 |
| $400 | $2,857 | ~$85,700 |
Two things worth taking from this. First: your budget doesn't depend on the size of your business, it depends on what you sell. A salon booking appointments at $40 needs less than a construction company whose lead costs $400, even if the construction company bills fifty times more.
Second: if the objective you picked is expensive, you can lower the threshold by picking one that's cheaper and more frequent. Optimizing for "purchase" when you sell three things a month is asking the algorithm to learn from three data points; optimizing for "message started" or "contact" gives you dozens a week on the same money.
What if you don't know what a customer costs you yet?
This is the most common case, and the most honest one to admit: the number you need to calculate your budget only comes from spending. The way out of that loop is to treat the first month as what it is — buying information — and not as a campaign that has to break even.
- Start with the cheapest event in the chain. A message or a contact happens far more often than a sale, and from that number you can work everything else upward.
- Let it run two weeks without touching it. The first week is pure learning and its numbers can't be used to decide anything; the second one is signal.
- Measure cost per result at the end, not on day three. A single expensive day means nothing; the full week does.
- With that number, go back to the table. Now your minimum budget is a calculation instead of a bet, and you know whether next month you raise it, lower it, or shut it off.
For that first month to be worth anything, the event has to be measured properly. If your pixel or Conversions API only records page views, you'll end up with a beautiful cost per visit and no usable data.
Four ways to burn the budget without noticing
Even with the right budget you can still throw money away. These four are the ones we see most, and all four have the same effect: keeping the ad set from reaching its 50 events.
- Splitting it across many ad sets. Five ad sets at $100 a day aren't "more tests," they're five campaigns stuck in learning limited at once. On a small budget, one ad set with a broad audience almost always wins.
- Changing something every day. Editing the budget, the audience, the objective or the creative resets the learning phase and throws away what was learned. If you're raising the budget, do it in moderate increments with days in between, not double overnight.
- Pausing and restarting. Turning it off for the weekend "to save" costs the same as a major edit: when it comes back, it starts from zero.
- Judging it on the wrong attribution window. If your sale takes a week to close and you're reading the 1-day-click report, you'll conclude nothing works. Check which window you're looking at before you decide.
What else do you spend on besides the ads?
The ad budget is only one piece. If you plan for it alone, the month goes entirely into ad spend and nothing is left for what makes it work:
- Creative. You need several pieces and replacements when they wear out. A single ad running for three months stops performing because of fatigue, not because of budget.
- Where the click lands. A page that loads fast and says the same thing as the ad. The ranges are in how much a website costs in Mexico.
- Whoever answers. The easiest cost to forget and the most expensive one: the messages nobody replies to are already paid for.
- Measurement. Pixel, events and, where possible, the Conversions API. It's one-time work, and without it everything else is guesswork.
A practical planning rule: if you only have money for the ads and nothing else, it isn't time to turn them on yet.
When is Meta Ads not worth it?
There are four situations where the right answer to "how much do I start with?" is "nothing, not yet":
- When your margin can't absorb the cost per customer. If you make $250 per sale and close one in five $60 messages, each customer costs you $300. Optimizing there just makes you lose faster. Raise the ticket, build packages, or improve the close first.
- When your total monthly budget is below the table's minimum. It's not that it returns less: it doesn't return. That money does more work on your Google Business listing, on your existing customers, or on fixing the page.
- When nobody is there to answer. Every peso in ads turns into a lead that goes cold. Fix the response first — with people or with a WhatsApp assistant — and then turn the ads on.
- When you can't measure where the customer came from. A week spent setting up measurement is worth more than a month of blind spend.
What to do if your budget is small
"Small" doesn't mean "do nothing." It means concentrate instead of spreading:
- One campaign, one ad set, one broad audience. Let Meta search; on little money, over-targeting only makes each result more expensive.
- Two weeks at a real budget, not two months on scraps. If you have $9,000, spend it over fourteen days at $640 a day rather than sixty days at $150. The first gives you an answer; the second gives you a receipt.
- Optimize for the most frequent event that's still useful. Lower your cost per result and the formula's threshold drops with it.
- Start with people who already know you. Someone who visited your site or messaged you before converts cheaper, and that moves the whole calculation in your favor.
If you want a second opinion on a campaign that's already running, the Meta Ads Audit in the AI Buffet has a fixed price and tells you where the budget is leaking and what to change first. And if you're looking for someone to run the whole campaign — structure, creative, measurement and follow-up — that's our Marketing service.
