This table shows the portfolio shape of Meta advertisers across the five spend tiers. Each row is a tier; each column is the share of that tier's creatives that fell into the loser, mid-range, or winner category.
| Spend tier | Loser (%) | Mid-range (%) | Winner (%) |
|---|---|---|---|
| Micro (<$10K) | 50.2 | 46.0 | 3.7 |
| Small ($10K–$50K) | 49.3 | 44.6 | 6.2 |
| Medium ($50K–$200K) | 52.6 | 40.1 | 7.3 |
| Large ($200K–$1M) | 53.9 | 38.0 | 8.1 |
| Enterprise ($1M+) | 52.2 | 39.6 | 8.2 |
Definitions: Loser = turned off before 28 days. Mid-range = ≥28 days of spend, not winner. Winner = ≥10× account median spend and ≥$500.
What's stable
The loser share is remarkably consistent across tiers — between 49.3% and 53.9%. Roughly half of every advertiser's creative output gets turned off before reaching day 28, regardless of whether that advertiser is a Micro account testing 3 ads per week or an Enterprise account testing 18.
This stability is a signal about Meta's auction. It evaluates creatives against a consistent set of performance signals, and across the dataset, roughly half of all ads fail to clear the bar for extended spend. That ratio isn't a function of advertiser strategy or budget; it's a property of the auction.
What's not stable
Two things shift as you move up the tier ladder:
Winner share doubles. From 3.7% (Micro) to 8.2% (Enterprise). That's the same modest hit rate improvement described earlier — scale modestly raises the per-creative probability of becoming a winner, but not dramatically.
Mid-range share declines. From 46.0% (Micro) to around 39% at higher tiers. As advertisers scale, a slightly smaller share of their creatives become durable-but-not-breakout. This is partly the mirror of the winner-share rise: more of what would have been mid-range at Micro scale ends up as winners at Enterprise scale, because the auction has more signal to work with when volume is high.
What it means for portfolio diagnostics
Compare your account's shape to the tier reference:
- Loser share materially above ~53%. The auction is turning off a higher share of your creatives than is typical. Possible causes: creative quality (briefing or production), audience fit, or format mismatch for the placements you're running. Worth investigating.
- Loser share well below ~50%. Two possibilities. Either you're running a strong creative operation with genuinely high survivor rate, or you're not turning off enough underperformers — letting them accumulate mid-range spend that could be redirected.
- Winner share at or above tier average. Whatever your testing cadence is, the quality pipeline is working. Focus shifts to volume scaling (if you want more absolute winners).
- Winner share materially below tier average. Likely a creative quality problem rather than a testing volume problem. Diagnose at the creative level — format, hook, asset type — not at the cadence level.
Companion view
Spend allocation by tier shows a related but different distribution — how budget (not creative count) splits between losers, mid-range, and winners at each tier. That chart tells the opposite story from this one: where portfolio composition is stable across tiers, spend allocation shifts sharply.