Open Ads Manager in almost any small business account and you will find the same situation: a long list of active ads, a few with purchases, many with a handful of clicks, and no clear idea which ones deserve more money.
Most advertisers decide by feel. They pause whatever looked bad yesterday, keep whatever has the nicest creative, and leave the rest running “to see what happens.” The result is predictable: good ads get cut too early, weak ads keep spending for weeks, and the budget ends up spread across too many things to learn anything.
The fix is not a smarter guess. It is a rule you apply the same way to every ad, every time.
The two numbers you need before judging any ad
Your target CPA
You cannot call an ad a loser without knowing what an acceptable cost per purchase is for your business. That number comes from your margins, not from industry benchmarks. If a product leaves you $30 of gross profit, an ad at a $45 CPA loses money on every sale regardless of how “good” its click-through rate looks.
If you have not set a target yet, the Breakeven ROAS Calculator gives you your maximum affordable CPA in a minute. Your target should sit below that ceiling.
Your test spend threshold
This is the amount an ad must spend before you are allowed to judge it. Without it, you end up comparing an ad that spent $8 with one that spent $300, which tells you nothing.
A practical approach is to set the threshold at a multiple of your target CPA, somewhere between one and three times, depending on how much budget you can afford to spend on testing. The exact number matters less than applying it consistently.
The four groups every ad falls into
With those two numbers, every active ad lands in one of four groups.
1. Not enough data yet. The ad has spent less than your threshold. Leave it alone. Pausing it now means you never learn whether it works.
2. Underperformers. The ad has passed the threshold and its CPA is clearly above your target. These are the ads to turn off. If an ad has spent well past the threshold with no purchases at all, it belongs here too.
3. Winners. The ad has passed the threshold and its CPA is at or below your target. Keep them running and protect them from unnecessary edits.
4. Scaling candidates. Winners that are consistent over time and are not yet receiving a large share of the budget. These are where extra budget should go.
This sounds simple, and it is. The difficulty is doing it for every ad, every week, without shortcuts.
Mistakes that lead to the wrong decisions
Judging by click-through rate alone
A high CTR means people find the ad interesting. It does not mean they buy. Some of the most clicked ads in an account attract curious visitors rather than buyers. Always judge on cost per purchase once the ad has enough spend. If you have many ads with strong clicks and weak sales, read Meta Ads Getting Clicks but No Sales.
Looking at the wrong time window
An ad that was a winner over its lifetime may have been declining for two weeks. An ad that looks terrible over 3 days may just have had a slow start. Check both: lifetime results to see if the ad ever worked, and the last 7 days against the previous 7 to see where it is heading.
Ignoring ads that Meta barely spends on
In campaigns where Meta distributes the budget, some ads receive almost no spend. They never reach your threshold, so they cannot be judged, but they also stay “active” forever. If an ad has had little delivery for a couple of weeks, Meta has already made its choice. Replacing it with a new idea is usually more useful than waiting.
Judging the same creative separately in each ad set
If the same creative runs in several ad sets, look at its combined results as well as each copy. A creative that underperforms everywhere is a creative problem. One that works in one ad set and fails in another points to an audience difference.
Pausing everything at once
Turning off half the account in one afternoon makes the next week very hard to read. Act on the clear cases first: underperformers well past the threshold and scaling candidates with consistent results.
How to run the check
Doing this by hand looks like this:
- Export ad-level results for the last 7 days, the previous 7 days and lifetime.
- Add your target CPA and threshold to the sheet.
- Filter out ads below the threshold.
- Mark the remaining ads above target as underperformers and those below as winners.
- Among the winners, check which ones are stable across both 7-day periods and how much of the budget they receive.
For an account with 30 or 40 active ads, this takes a while, which is exactly why most people stop doing it after the first few weeks.
Let the classification happen automatically. Our Meta Ads Analyzer evaluates every ad against your own CPA target and test spend threshold, and shows you winners, underperformers and scaling candidates as soon as you connect your account. Start your 3-day free trial
What to do with each group
- Underperformers: pause them. Before you do, look at what they have in common (same angle, same format, same audience). That pattern is worth more than the individual ads.
- Not enough data: give them time or budget until they reach the threshold. If an ad is not getting delivery at all, replace it.
- Winners: keep them running and avoid editing them. Edits can disrupt delivery on ads that are doing their job.
- Scaling candidates: move budget toward them gradually. If scaling has hurt performance before, read Why Meta Ad Performance Drops After You Increase the Budget.
If your winners are losing strength after a few days, the problem is probably fatigue rather than the ads themselves. See Facebook Ads Worked for a Few Days, Then Stopped.
Frequently asked questions
How long should I let a Facebook ad run before turning it off?
Judge by spend, not by days. An ad should reach your test spend threshold (typically one to three times your target CPA) before you decide. For low budgets, that might take a week or more. For higher budgets, a couple of days.
Should I turn off an ad with a high CTR but no sales?
If it has passed your spend threshold without purchases, yes. A high CTR with no sales usually means the ad attracts attention from people who are not buyers, or that something after the click is blocking the purchase. Check whether other ads convert on the same landing page before blaming the ad.
Does pausing ads reset the learning phase?
Pausing individual ads is generally less disruptive than changing budgets, targeting or optimization settings, but many changes at once still make results harder to interpret. Make changes in small batches and give the ad set a few days to settle.
How many ads should I run per ad set?
Enough that Meta has options, few enough that each ad can reach your spend threshold. With a small budget, a handful of ads per ad set is usually more useful than a dozen that never get enough spend to be judged.
What if all my ads are underperforming?
Then the problem is probably not the ads. When every ad fails at once, look at the offer, the website, the audience or a change in the market. Start with Why Your Meta Ads Stopped Performing.
Stop deciding by feel. Connect your ad account, set your CPA target, and see exactly which ads to pause and which to scale. Try our Meta Ads Analyzer free for 3 days