Meta Ads Winner Criteria: How to Know an Ad Is Worth Scaling

A winner is not an ad with a great ROAS on day two. It is an ad Meta chose to spend on, that held up against your account's own baseline over a full week.

Why a high ROAS alone does not make a winner

Open any ad set after two days and you will find a few ads with a beautiful ROAS. Look at the spend next to it. Often it is $40 and two purchases. That is not a result. It is noise with a nice number on it.

The reverse happens too. An ad takes most of the budget, ROAS is merely okay, and you pause it because it looks mediocre. But Meta had a reason to keep feeding it, and your account may have needed exactly that volume.

So a winner needs two things at once: Meta put real money behind it, and it earned that money at a return you can live with. Spend tells you about scale. ROAS tells you about quality. Judge on one and you will scale the wrong ads.

The four outcomes of a creative test

After a test has run for seven days, every batch lands in one of four boxes. This is the rule Hitrate uses, and you can run it in a spreadsheet today. It is judged per campaign, because a $300/day campaign and a $10,000/day campaign are not the same game.

First, define a big share of spend. The share depends on campaign size, and you should change it to fit your account:

• Campaign under $1,000/day: a big share is 30% or more of the campaign's spend.
• Campaign under $5,000/day: a big share is 20% or more.
• Campaign above $5,000/day: a big share is 10% or more.

Scaler, Spender, Efficient, Miss

Now sort every tested batch with four labels:

• Scaler: took a big share of its campaign's spend, and the campaign spent at least 10% more per day than the week before. Meta liked it, and the campaign grew with it. This is your winner.
• Spender: took a big share, but the campaign did not grow. It holds a lot of budget, yet it may only be taking it from ads that were already working. Treat it as a strong candidate, not a proven scale driver.
• Efficient: beat the campaign's own ROAS but got little spend. The ad is good, and Meta did not give it room. Test it again with a different hook or a different audience before you give up on it.
• Miss: neither big spend nor better ROAS. Log it, learn from it, move on.

A worked example (illustrative numbers)

Say a prospecting campaign spends $2,000 a day. Last week it averaged $1,700 a day. A big share at that size is 20%, so $400 a day.

Batch 14 gets $620 a day, which is 31% of the campaign. The campaign grew about 18% week over week. That is a Scaler. Batch 15 has a 2.4 ROAS against the campaign's 1.9, but only $60 a day. That is Efficient. Batch 16 holds 25% of spend while the campaign stays flat. That is a Spender. Batch 17 gets little spend and a ROAS below baseline. That is a Miss.

These numbers are an example, not a benchmark. The point is the shape: a verdict comes from spend share, spend growth and return together, and each one has a written threshold you decided before you looked.

Judge ROAS against your own baseline, or a target CPA

Do not compare an ad to a number from a blog post. Compare it to the campaign it ran in. A 1.8 ROAS is a win in a campaign averaging 1.3 and a loss in one averaging 2.5.

If you run on a hard unit economics limit, use a target CPA instead. Pick it from your margin before the test starts, write it down, and judge every batch against it. The rule only works if the bar is set in advance. Moving it after you see the results is how a weak ad becomes a winner on paper.

Wait seven days, but do not close the book at seven

Delivery needs time to settle. Seven days gives you weekday and weekend behavior and gets past the early swings. Give the verdict then, not on day three.

Seven days is the first read, not the last. A batch that looks like a Miss in week one can still become a Scaler in week two or three as Meta finds its audience. Keep batches live that are not hurting the account, and re-check them against the same rule.

The test unit matters too. Judge a concept, not a single ad. Put one concept in one ad set, with a few variations, and the verdict says something about the idea. One ad in a pile of ten tells you very little.

Mistakes that produce fake winners

Most false winners come from the same few habits:

• Calling it on day one or two, when spend is tiny and purchases are single digits.
• Changing the bar after the results are in.
• Comparing ROAS across campaigns with different audiences and budgets.
• Testing five ideas inside one ad, so you cannot tell what worked.
• Naming ads loosely, so a month later nobody can say which desire, angle or format a result belongs to.
• Never writing the verdict down, so the same loser gets re-briefed in three months.

From verdicts to an ad creative hit rate

One winner is a lucky break. A pattern is a strategy. Once every batch has a label, you can count your ad creative hit rate: the share of tested batches that became a Scaler or a Spender.

Then split it. Hit rate by desire, by angle, by awareness stage and by format. You might find that problem-aware ads on one desire hit far more often than the rest. That tells you what to brief next, and it costs nothing but the discipline of tagging every test.

Hitrate does the tagging and the labeling for you. It writes the ad names so each result maps back to its batch, desire, angle, awareness stage, format and creator. It syncs results daily from Meta, read-only, and applies the rule above after seven days. The labels and numbers are computed by rule, never by an AI model.

Related guides

• The creative strategist workflow, from research to learnings
• How to identify creative fatigue in Meta ads (with thresholds)
• Meta hook rate benchmark: what counts as a good hook rate
• How to improve hook rate on Meta video ads
• A Meta ads creative testing framework you can run weekly

Updated 2026-10-02