Blog Inside Sneeze It · Inside Sneeze It
More than 90% of your ads will fail. Here is what we found in the ones that didn't.
Meta tries a lot of ads, backs a few and drops the rest. We graded 30,857 lead ads twice, once by what Meta spent and once by what a lead cost, and found four things every owner can check this week.
More than 90% of the ads you run will fail, and your team is not the reason. Meta tries a lot of ads, backs a few and drops the rest.
We graded every lead ad we could read in the accounts we work with: 30,857 ads across 179 accounts, from September 2023 to October 2026. Only 5.3% became what we call a spend winner, meaning Meta put at least ten times the account's typical ad spend behind it, and at least $500. The other 94.7% never got there.
Source: Sneeze It analysis of Meta lead campaigns, September 2023 to October 2026
The failure is in the count, not the money
Almost half of those ads, 47.2%, brought in zero leads. Together they used 13.2% of the spend. Meta stops spending on ads that don't work, and it does it quickly.
So most of your money does not go to losing ads. The bigger cost is good ads that never get the budget they earned.
Losing ads cost you thirteen cents of every dollar. Missing your good ones costs more.
What we built to find them
That is why we built Direction. It runs inside Studio, our creative tool, on every client account we manage. It reads every ad in the account and grades it twice: once by what Meta spent on it, and once by what a lead from it actually cost. When the two grades disagree, we go with the cost. Then it tells our team what to make more of, what to test and what to stop.
- 1ReadEvery ad in every connected Meta account
- 2Grade by spendHow much Meta chose to put behind it
- 3Grade by costWhat a lead from it cost against the account average
- 4DecideMake more, test, or stop, with the evidence beside it
We ran the same two grades across all 30,857 lead ads. Four things stood out.
One: Meta's favorite ad is not your cheapest lead
The ads Meta spent the most on brought leads at 0.98 times their account's average cost per lead, about average.
And most of the ads that did bring cheap leads never got Meta's backing. Of the ads that brought leads at or below their account's average cost, 86.8% never became a spend winner.
Two: the cheap leads were hiding in ads Meta starved
We looked for ads with at least five leads, a cost per lead at or below the account average, and less than three times the account's typical ad spend. We found 1,771 of them. Their leads cost 42% less than average, and the typical one had spent about $199 in its whole life.
139 of the 179 accounts had them, about three out of four, with four in a typical account. Together they brought 10.3% of all leads on 6.9% of the spend.
The ads Meta backed
Typical spend of $1,403 each. Leads at 0.98 times the account average.
The ads Meta starved
Typical spend of $199 each. Leads 42% below the account average.
Three: the first line decides it
This one comes from our Creative Benchmarks, where we have run over 18,920 Meta ads and labeled 2,126 of them by how they look, how they open and how they were made. Ads that opened with a bold claim became winners 14.9% of the time, more than twice the 6.2% rate across all our ads, and brought leads about 24% cheaper than average.
The opening is the first thing anyone sees. In the same data, ads that opened on a relatable moment cost about 42% more per lead than average, so the first line deserves the most work.
Four: more launches, more winners
Also from the benchmarks: across the 97 accounts with at least 50 leads, the more new ads an account launched each week, the more winners it found.
Source: Creative Benchmarks, median across 97 accounts with 50+ leads
Accounts launching four or more new ads a week found almost three times as many winners as accounts launching fewer than one. Launching more does not make any one ad better; it gives you more chances at a rare result.
The four checks we run on every account
You can run these in Ads Manager this week. Set the date range to the last 90 days and look at ads, not campaigns.
1. Find the few ads doing the work. Sort your ads by leads. In a typical account, 4 of 37 ads brought half of all its leads. Know which four are yours.
2. Find the cheap ads Meta starved, and fund them. Sort by cost per lead, lowest first. Any ad with five or more leads at or below your average cost, on a small budget, has earned real money. Give it some and watch the cost.
3. Stop paying for ads that bring nothing. An ad that has spent the price of two of your average leads and brought none is unlikely to turn around. Across our accounts, zero-lead ads that kept spending past that point used 5.3% of all spend. Pause them.
4. Judge new ads on cost per click, not cost per impression. Inside each account, among ads that spent at least $100, the cheapest quarter by cost per click brought cheap leads 53% of the time; the priciest quarter, 19%. Cost per thousand impressions told us almost nothing: 38%, 39%, 37% and 33% across the four quarters.
Sources
- Sneeze It analysis of Meta lead campaigns (objectives: leads, conversions, sales), 179 ad accounts, 30,857 ads with spend, 224,783 leads, September 2023 to October 2026. Read-only Meta Marketing API pull, October 12, 2026.
- Sneeze It Creative Benchmarks: 18,920 Meta ads, 2,126 labeled, 97 accounts with 50+ leads.
