How to calculate your ad account hit rate

What counts as a winning ad?

This is where most arguments about hit rate actually come from. Two brands can both say "our hit rate is 20%" and be measuring completely different things.

The performance definition

An ad wins when it comes in at or below your target CPA, or at or above your target ROAS, with a minimum number of conversions behind it. The minimum matters. One lucky purchase on a $40 spend is not a winner, it is noise.

If you have never set a target CPA from your own margins, do that first. It is the number everything else here depends on, and you can work it out with the breakeven ROAS and max CPA calculator.

The spend definition

Motion's 2026 Creative Benchmarks study takes a different route. An ad is a winner when it spends at least ten times the account median and at least $500 in total. The logic is that Meta already votes with budget, so sustained spend is the signal. If your median creative spends $200, an ad has to reach $2,000 to count.

Neither is wrong. The performance definition tells you whether an ad earns money. The spend definition tells you whether the algorithm backed it. Just be clear which one you are quoting.

Method 1: the new-ad hit rate

Media buyer Marin Istvanic tracks this weekly. Count the new ads you launched, count how many hit target, divide. His read on the number:

• Under 20%. The creative process is broken. The problem is usually upstream, in the angle or the offer, not in the edit.
• 20 to 40%. It is working. Keep going.
• Over 40%. The brief format is good enough to scale, so make more of what you are already making.

The useful part is what he does with the number rather than the number itself.

It picks your campaign structure. If 6 or 7 in 10 new ads hit target, testing in ABO is efficient, because most of what you put in works. If only 2 or 3 in 10 do, ABO is paying to discover losers, so CBO with a minimum spend does the sorting for you.

It sets your testing volume by season. His hit rate runs near 40% in Q1, so he tests 20 to 30 concepts a week. It falls to 10 or 15% over summer, so he drops back to about 5. Same process, different yield, so the effort follows the yield.

Method 2: the live-account hit rate

The Foundr Ad Testing Engine, built on the framework Nick Shackelford's team uses across more than 127 DTC brands and $200 million in Meta spend, asks a different question. Not "did my new ads work" but "what state is this account in today".

Every live ad goes into one of three buckets:

• Winning. At or below target CPA.
• Testing. Not enough data yet.
• Losing. Above target CPA.

Winners divided by total gives an account health score: strong, stable, thin or critical. That score then sets the budget split between scaling winners, testing new creative and exploring new angles. A strong account sits near 70/20/10. A weak one shifts to 40/40/20, which is the discipline most people skip, because a thin account is exactly when it feels tempting to pour more budget into an ad that is not really a winner.

The kill rules

An ad cannot sit in the testing bucket forever, so the thresholds are fixed in advance:

• Spent 1x your target CPA with no conversions. Kill it.
• Spent 1.5x with only one conversion. Kill it.
• Reached 2x your test budget. Kill it.

And nothing gets promoted to winner until it clears a minimum conversion count.

What is a good hit rate?

Motion analysed 578,750 creatives across 6,015 accounts and $1.29 billion in spend. On their spend-based definition:

Read that again if you have been beating yourself up about your creative. A 5% hit rate is not failure, it is the shape of the game. Winners are rare on purpose, because the algorithm is built to find the few and feed them.

Which makes hit rate a capacity problem more than a quality problem. Hold the rate steady and change the volume:

• 4 new ads a week is roughly one winner every five weeks.
• 18 new ads a week is roughly one winner a week.

Same hit rate. Completely different business.

Why the number can still lie to you

Barry Hott argues hit rate is a bad way to judge a creative source at all, and he has a point worth taking seriously. A high hit rate usually means safe, iterative ads and not enough big swings. It also scores a creative that lived six weeks the same as one that has carried spend for two years.

His alternative for judging an editor, an agency or a freelancer: how much ad spend can their ads carry? That rewards the rare ad that scales, which is the one that actually pays for everything else.

I use both. Hit rate to check my own process is producing, spend carried to judge who I keep working with. They are different jobs.

Work out yours in ten minutes

1. Set your target CPA from your margins, not from a number that feels nice.
2. Open Ads Manager, last 30 days, at ad level, and filter to active ads.
3. Sort every ad into winning, testing or losing against that target CPA.
4. Divide winners by total ads. That is your live-account hit rate.
5. Now filter to ads launched in the last 7 days and repeat. That is your new-ad hit rate.
6. Write both down with the date. One reading tells you almost nothing, the trend tells you everything.

If the live number is thin, stop scaling and go back to testing. If the new-ad number is under 20%, the fix is upstream in the angle and the offer, not in another edit of the same concept.