How much should I spend on Facebook ads?

Start with the business target

The first budgeting question is not “What can I afford?” It is “What does a customer or lead allow me to spend profitably?” That means working backward from break-even ROAS, target CPA, and contribution margin.

Why small budgets fail so often

Many accounts do not fail because the ad is bad. They fail because the budget is split across too many ad sets, too many tests, or too difficult an optimization event. The result is not enough data for any one path to stabilize—this is closely related to learning limited issues.

A simple way to estimate budget

• Start with the result you want. If your target CPA is $5 and you want 2 to 4 conversions per week, you are roughly looking at $10 to $20 per week.
• If your target CPA is $10 and you want 5 conversions per week, you are roughly looking at $50 per week.
• If your target CPA is $20 and you want 5 conversions per week, you are roughly looking at $100 per week.
• The formula is simple: target CPA multiplied by the number of conversions you want to generate in a week.
• If your budget is small, keep the setup simple so the little data you do get is not fragmented across too many ad sets or tests.

How business type changes the answer

Ecommerce brands with broader demand can usually support more testing earlier. High-ticket, local service, or niche B2B advertisers usually need fewer ad sets and more budget concentration because event volume is lower.

What to do if your budget is small

• Use fewer ad sets.
• Reduce creative complexity.
• Choose the closest realistic optimization event to revenue.
• Judge performance over a longer window instead of reacting daily.

The working rule

The right Facebook ads budget is the amount you can afford to spend consistently while still giving the campaign a chance to produce useful signal. If the budget is small, simplify before you scale and measure over longer windows instead of expecting fast certainty.

The floor is set by arithmetic, not by ambition

There is a minimum below which a campaign cannot produce a readable result, and it comes from the learning threshold. An ad set needs roughly 50 optimisation events in seven days — about seven a day — to leave the learning phase. Multiply your realistic cost per conversion by seven and that is the daily budget one ad set needs. A £20 cost per purchase implies about £140 a day in a single ad set.

If your total budget is well below that, the right structure is not a smaller version of a big account. It is one campaign, one ad set, and every pound in it — because the same money split three ways produces three ad sets that all learn nothing.

Budget for the learning cost, separately

The first hundred to two hundred clean conversion events are an expense, not a profit target. Delivery during learning is deliberately exploratory and genuinely worse, and several operators describe budgeting for that phase as a cost of doing business. An account that kills every campaign in its first three days pays this cost repeatedly and never gets the benefit.

Practically: decide before launching how much you are willing to spend to find out whether this works, and treat that figure as the test's price. If it is less than roughly ten times your expected cost per purchase, the test cannot answer the question.

When to add more, and how fast

Do not scale from just above break-even. Every budget increase costs some efficiency, so a campaign at 1.8x against a 1.6x break-even can fall through the floor as soon as you add money. Operators want a real gap — break-even 1.6 against an actual 3.0 — before increasing anything.

On the method, practitioners genuinely disagree, and it is worth knowing both positions. Large agencies raise budgets on a schedule, commonly around 20% every 48 hours, relying on having several campaigns so no single one carries the risk. Many mid-sized operators refuse to touch a working campaign at all and duplicate it instead, launching the duplicate at roughly 1.6–2x the original's budget and running both. Our guide on judging ROAS covers how much headroom you need before either is safe.

Where the budget should not go

Two common ways small budgets are wasted before they ever reach the auction. Spreading a small budget across several ad sets or campaigns, which guarantees none of them learn. And spending on traffic or engagement objectives "to warm up the pixel", which trains Meta to find people who click and do not buy — an expensive mistake to undo later.