What share of top DtC brands' Meta ads are video? (November 2026)
Video is 47% of the median top DtC brand's live Meta ads (Nov 5, 2026, 100 brands). The middle half sit between 36% and 63%, and 42 of 100 brands run mostly video. Static images are still the majority format.
Static or video: what do top DtC brands run on Meta?
Mostly static. The median brand runs 53% static ads and 47% video. Only 42 of 100 brands are majority video.
Video share of Meta ads by sector
| Sector | Brands | Median video share |
|---|---|---|
| Apparel & Accessories | 27 | 33% |
| Food & Beverage | 20 | 47% |
| Home & Kitchen | 14 | 63% |
| Beauty & Personal Care | 14 | 47% |
| Health & Wellness | 10 | 63% |
| Outdoor & Travel | 7 | 63% |
| Pet | 4 | 61% |
| Kids & Baby | 3 | 59% |
Do higher-priced products use more video?
Yes. Video share rises with order value: a static image can't justify a big purchase on its own.
| AOV band | Brands | Median video share |
|---|---|---|
| Under $100 | 38 | 47% |
| $100–299 | 41 | 47% |
| $300–799 | 9 | 50% |
| $800+ | 5 | 62% |
Who this benchmark is not for
The share of live ads is not the share of spend. A brand can run mostly static ads and still put most of its budget behind a few videos. Check your own spend split before changing your mix.
How this benchmark is built
FORGE Operators tracks 100 best-in-class direct-to-consumer brands across apparel, beauty, food and beverage, home, health and more. Once a month we pull every brand's live ads from the public US Meta Ad Library and record live ad count, ads launched in the last 30 days, video share, creator and partnership ads, offers, social proof, CTAs and the age of the oldest live ad. Mix percentages cover every live ad in each brand's library. Revenue and AOV are FORGE estimates from public sources where not disclosed. Figures are medians unless noted, so one outlier (Quince runs over 5,000 ads) doesn't skew them.
Snapshot: Nov 5, 2026. Ads pulled 2026-11-05. This is a full census: every live US ad for every brand. Brand-level month-over-month comparisons start with the next pull measured the same way.