Ad Creative Testing Benchmarks 2026
Weekly ad volume and win rates
Compare new ads per week by budget using Motion's Meta research. Its win rate counts ads that receive more spend; it does not measure profit.
Key findings
Check the source and sample before you use a figure as a target.
Testing by budget
Find your budget group
Weekly volume counts newly launched ads. Win rate counts ads that meet the spend rule below. Each account has equal weight in its group's average win rate. [1]
Use the table to plan a pace you can fund and review. Launching more ads is useful only if you can learn from the results.
Meta creative testing volume and win rate by budget
[1]| Micro | $0 | 2.8 | 4.0% |
|---|---|---|---|
| Small | $10,000 | 4.1 | 6.4% |
| Medium | $50,000 | 6.6 | 8.1% |
| Large | $200,000 | 11.2 | 8.6% |
| Enterprise | $1,000,000 | 18.8 | 8.8% |
Each budget group ends where the next begins. Enterprise has no upper limit. These averages are not recommended quotas. Select a column heading to sort.
Winner definition
What counts as a winner?
Motion calls an ad a winner if its spend is at least 10 times the account's median ad spend and at least $500. The median is the middle value. This rule measures spend, not sales or profit. [2]
Example: if median ad spend is $200, an ad must spend at least $2,000 to qualify. It can meet that rule and still miss your acquisition-cost target.
- Record your winner rule before you compare rates.
- Compare profit-based rates only with rates that use the same profit rule.
- Separate image and video results when planning video production.
Measure your tests
Count tests consistently
Win rate = winners ÷ eligible ads tested × 100. Example: two winners from 25 tested ads gives an 8% win rate. With no eligible tests, mark the rate as unavailable.
Use the same review period for a fixed group of launched ads. Mark ads with too little spend as inconclusive so they do not look like proven failures.
Motion counts each separately launched variation as a new ad. Track concepts and variations separately to see how many different ideas you tested. [1]
Plan the next batch
Plan a batch you can fund
Write the test question first. Choose what will change: the concept, opening, proof, or call to action. Set a success measure, budget limit, and review date.
Example: three hooks × two bodies × one call to action = six variations. This counts combinations. It does not predict winners.
- Use one concept name to group related variations.
- Separate the test budget from spend on established ads.
- Choose the test budget from your costs and available funds.
- Review results before you produce a larger batch.
About this report
Data sources and limits
The figures come from one Motion dataset: 578,750 ads and $1.29 billion in Meta spend. It covers September 1, 2025 to January 1, 2026, including the holiday period.
Each account had at least 10 unique ads. The sample includes images and video. Keep those limits in mind when comparing a smaller account, a video-only plan, or another season. More ads did not necessarily cause better results.
- Testing volume: average new ads launched per account each week.
- Win rate: average of account rates within each budget group. Each account has equal weight.
- The figures cover Motion's sample, not Sovran customers.
- Examples explain the calculations. They are not research findings.
- Sources checked: September 29, 2026.
Sources
- [1] Motion. “Weekly testing volume and hit rate by advertiser spend tier.” 2026. https://motionapp.com/library/research/creative-benchmarks-2026/testing-volume-by-tier (opens in a new tab). Accessed September 29, 2026.
- [2] Motion. “Methodology and definitions — Creative Benchmarks 2026.” 2026. https://motionapp.com/library/research/creative-benchmarks-2026/methodology (opens in a new tab). Accessed September 29, 2026.
Common questions
More benchmark reports
Compare related costs, formats, and results.
Video Ad Production Cost in 2026 — Pricing Benchmarks by Method
Compare agency, editor, creator, and software costs. See what to include in your production budget.
Video Ad Hook Rate Benchmarks by Industry 2026
Compare how often viewers watch the first three seconds across 14 Meta ad categories.
Video Ad Structure Benchmarks 2026
See which ad sections appear most often in 527 classified video segments from 25 Sovran projects.
Plan your next test
Use these tools and guides to put the data to work.
Build your next batch of ads
Use Sovran to combine hooks, bodies, and calls to action from your footage. Give each variation a name so you can compare results.