Facebook Ads Budget Calculator

Set a daily budget sized to clear Meta's learning phase and hit your target cost per purchase, with a cushion built in.

Describe the result you need. AI will set up the tool and ask for any missing details.

Configure with AI

Enter to submit · Shift + Enter for a new line

Sent to AI. Leave out passwords and private customer data.

Your numbers

$

Meta says an ad set needs about 50 purchase events in a rolling 7-day window to exit the learning phase, shared across every ad in the set.

events

Actual cost per result is often higher while an ad set is still learning. This adds a buffer so the budget doesn't stall mid-week.

%

Your result · updates as you type

Recommended daily budget

$223

50 purchases a week at a $25.00 target CPA needs $1,250 a week, plus a 25% cushion while the ad set is learning.

Minimum daily budget
$179
Weekly budget to exit learning
$1,250
Events needed this week
50

This is per ad set, not per campaign: every ad inside one ad set shares the same 50-event pool. If you split budget across many ad sets, each one needs its own 50 events a week, so fewer, better-funded ad sets usually leave learning faster than many thin ones.

Breakdown

Target CPA × events needed per week
$1,250
÷ 7 days
$179
+ 25% learning-phase cushion
$223

Budgeting around the learning phase

Meta's algorithm needs a burst of results to learn who to show an ad to. It says an ad set needs about 50 optimization events (usually purchases) in a rolling 7-day window to exit the learning phase, shared across every ad inside that ad set. A budget too small to reach 50 events a week keeps resetting into learning, which usually means a worse cost per result.

Weekly budget = target CPA × events needed per week Daily budget = weekly budget ÷ 7, plus a cushion

Example: the default numbers

Target cost per purchase
$25
Events needed per week
50
Learning-phase cushion
25%

Weekly budget: 25 × 50 = $1,250. Minimum daily budget: 1,250 ÷ 7 = $179. With the cushion: $223 a day.

Using this with your ad sets

  • Per ad set, not per campaign. Every ad inside one ad set shares the same 50-event pool, but each ad set needs its own 50 a week.
  • Fewer, better-funded ad sets usually leave learning faster than many thin ones splitting the same total budget.
  • Real CPA often runs above target during learning. That's what the cushion is for: without it, the budget can run out before the week's events are collected.

Questions people ask

How many conversions does Meta need to exit the learning phase?

Meta says an ad set needs about 50 optimization events in a rolling 7-day window, shared across all ads in that ad set.

Why does my ad set keep re-entering the learning phase?

Big edits (budget, audience, creative) can reset it. Keep the ad set stable once it's learning, and avoid pausing and restarting it repeatedly.

Should I split budget across many ad sets?

Usually not while still learning. Splitting a small budget across many ad sets means each one struggles to reach 50 events a week on its own.

What if my target CPA is a guess?

Start with the break-even CPA from the break-even ROAS calculator, then tighten the target once you have real campaign data.

Want this builtinto your store?

I'm Emmanuel, a Shopify and WooCommerce developer. If this tool pointed at something you'd rather have fixed for you (a faster store, cleaner product data, a migration, a custom feature), send me a note.

Emmanuel Asika

Emmanuel Asika

Shopify & WooCommerce developer, Dublin

I build and fix online stores: new builds, migrations, speed, SEO and the custom bits apps can't do.

hello@emmanuelasika.com

Last reviewed 30 September 2026.

Sources: Meta Business Help Center: About the learning phase