Can your budget actually work?
Meta has a floor. Below it, its system cannot optimise properly no matter how good your ad is, and it never tells you where the floor is. Most founders spend a year failing underneath it and conclude they are bad at ads. Here is the arithmetic, in ninety seconds.
"You cannot coach someone across a gap that is arithmetic, not skill."
Two answers, one verdict
Nothing is stored unless you ask. Nothing is sent anywhere.
This is your optimisation event: the thing you tell Meta to go and find. It is the highest leverage decision on a small budget and almost nobody realises they are making it.
Media budget only. Not what you pay an agency, not what your funnel cost to build.
Where the floor comes from
Meta's own help centre says an ad set leaves the learning phase after about 50 results in a week, and that until it does, delivery is less stable, the cost per action is higher, and your budget is spent faster. That is not opinion. It is Meta describing its own system. Turn it into money and you get a floor.
Everything on this page is that one sentence, done as arithmetic.
Three ways across the gap, and only three
Every honest answer to a small budget on Meta is one of these three. The first costs nothing and is the least talked about.
Change what you optimise for
Meta's own documented remedy. Move the event further up the funnel and the required budget collapses, because cheaper events happen more often. This costs nothing and it is the single most valuable move available to you.
Raise the budget
Honest, unpopular, and often the wrong answer for a founder rebuilding after a corporate career. If you can genuinely fund it, this is what the current event demands every month, with nothing left over for creative or landing pages.
Pool it with others
The same money from more people, in one campaign for one shared event, clears the floor together. Not data sharing: one advertiser, one campaign, the way a franchise co-operative fund works. This is what The Foundry runs.
The arithmetic is only half of why founders stop
The other half is fear, and it is specific rather than general. Six of them, and most people have one that dominates. Rate each one honestly. This is your day zero baseline, and the only way to prove later that it moved.
Finish the check above and a full brief appears here, ready to paste into Claude or ChatGPT.
Knowing the floor is step one. Not panicking on day three is step two.
Kavach is the instrument that runs after this one: a hard limit you can never exceed, and one sentence every morning telling you to keep, kill or wait, with the arithmetic shown. It opens with the AI Business Amplifier.