How to set a starting ad budget

A good starting budget is the amount that lets Meta collect enough conversions to learn from, at a cost you can afford to lose while you find out what works. It is a function of your target cost per acquisition, not a round number picked from a blog post.

Start from the conversion event

Meta’s delivery system needs a steady flow of optimization events to exit the learning phase. Meta’s current guidance cites roughly 50 events per ad set per week. Treat that as a guide and check the figure in Meta’s documentation.

That gives a floor for each ad set:

Daily budget = target CPA x 50 events / 7 days

With a $20 target CPA, that is $20 x 50 / 7, or about $143 a day per ad set. With a $60 target CPA it is about $429 a day.

Most new advertisers cannot fund that on every ad set. The formula shows the trade-off you are making when you spend less.

What to do if the floor is too high

You have three honest options.

1. Run fewer ad sets. One ad set at $50 a day learns faster than five at $10. Our campaign structure guide covers consolidation.
2. Optimize for an earlier event. Add to cart happens more often than purchase, so it can reach the event count on a smaller budget. The cost is a weaker link to revenue.
3. Accept a slower learn. A smaller budget still works. It takes longer to produce a stable signal, and results will vary more in the meantime.

Fix your target CPA first

Your target CPA should come from margin, not hope. Take your average order value, subtract product and fulfilment costs, and the remainder is the most you can pay for a sale and break even. Break-even ROAS walks through the arithmetic. If your break-even CPA is $25, planning around a $15 target leaves room for profit after ad spend. See cost caps and target CPA for how Meta uses that number in bidding.

Budget for testing separately

A testing phase has a cost. If you plan to compare four creatives and each needs enough spend to produce a clear read, multiply that per-creative amount by four before you launch. Spend on creatives that lose is part of the cost of finding the one that wins.

A simple rule of thumb: decide the total amount you can lose over the first two to four weeks, and divide it across the days. If that figure is below the floor above, reduce the number of things you test.

Scaling after the start

Once an ad set converts steadily at or below target CPA, raise the budget in modest steps and watch cost per result. Large jumps can send delivery back into learning. See scaling ads for how to do this safely, and automated rules for guardrails.

Common mistakes

• Splitting a small budget across many ad sets.
• Changing the budget every day during the first week.
• Choosing a budget before knowing your break-even CPA.
• Judging results after one or two days.

How YieldBI helps

YieldBI’s growth controls let you set a Profit Goal and Growth Priority, and use them to drive daily scale, test, and pause recommendations. That gives a budget decision a basis in your margin instead of a guess. The guided campaign wizard also helps you structure a new campaign before you spend.