Every “UGC is dead” thread on X tries to convince you that video is winning by knockout. The actual data — from accounts spending real money — says something else. A 2025 analysis of 67,000 Facebook ads across DTC accounts found the median brand runs 61% static and 39% video. Static is the majority of the spend. It’s also the part of the creative system most agencies treat as an afterthought.
Here’s what’s actually verifiable about static ad performance right now — and how to design statics that beat your control.
The static-vs-video numbers, with sources
The honest summary: video has higher CTR, static has lower CPM. CPA depends on which side wins the math for your specific funnel.
- Median DTC mix: 61% static, 39% video — across 67,000 ads in the Segwise dataset. Variance is huge by vertical: Shapellx runs 97% video, SHEIN runs 1% — both scale profitably, just on different mixes.
- CPM differential: Static maintains 40–60% lower CPMs than video on Meta in DTC categories — same source.
- CTR differential: Video has 35–50% higher CTRs than static — also same source.
Translated: if your static can convert at 60% of your video’s CVR, your CPA is the same — at half the CPM. If your static is well- designed enough to convert at parity, your CPA is materially lower.
Carousels are the underrated weapon
The single most underused format in DTC right now is the carousel. Lebesgue’s benchmark data shows carousel ads generate 2.3x more engagement actions (reactions, comments, shares) than single-image ads — single-image at 0.94% engagement vs 2.17% for carousel.
On CTR, the two formats are closer than people assume — both around 0.9% median. But carousels win on conversion-from-engagement: viewers who engage with a carousel for 5+ seconds convert at 2.9x the rate of single-image viewers, per MHI Growth Engine’s analysis of 894 carousel ads (Q4 2025 – Q1 2026 data).
Same source: 4-card carousels are the optimal length. 41% of viewers see all four cards. CTR averages 1.47% (above the benchmark). Conversion rate averages 2.4%. Three cards ship faster but leave conversion on the table; five+ cards see drop-off after card 4.
Why static still works in 2026 (when video gets all the press)
1. Andromeda rewards modality variety
Meta’s Andromeda retrieval engine deduplicates near-identical creatives. A wall of UGC reskins clusters in embedding space and reads as a thin candidate set. A clean static product ad doesn’t cluster with any of them. Adding 3–5 distinct statics to a video-heavy account often lifts overall account ROAS — not because statics outperform, but because they expand the distinct-candidate count.
2. Static is the only sane format for some placements
Google Display, Meta Marketplace, certain LinkedIn placements, email, and most retargeting placements are still primarily image-driven. If you’re running across networks, static isn’t optional.
3. Static ad production cycles compress test velocity
A polished static ad ships in days, not weeks. A carousel pack ships in the same timeframe as a single video. Per the MHI data, you can run a 4-card carousel A/B/C/D test cycle in the time it takes to ship one video. That cycle compression is its own competitive advantage in a post-Andromeda world.
What “premium static” actually looks like in 2026
The brands ranking highest in the Meta Ad Library for static-ad creative — Glossier, AG1, Liquid Death, Our Place, Recess, Olipop, Magic Spoon, Caraway, Curology, Function of Beauty — share a consistent visual language. Look at any of their active campaigns and you’ll see the same five principles:
- Editorial composition, not product-on-white. The product sits in a curated environment (marble, velvet, linen, driftwood, travertine, polished steel) — not floating on a white background.
- Single hero element + breathing room. Negative space gives the eye a place to land. Cluttered statics fail at thumbnail size.
- High contrast between product and background. Meta’s own creative best practices flag this as the single most important variable for image ad CTR.
- Implied motion in static composition. Liquid Death tilts cans into action. Recess implies pour. Olipop angles the can mid-tilt. The static photographs energy even though nothing moves.
- Brand-consistent color and styling across the carousel. Each card looks like it came from the same campaign, not a Pinterest board. This is where most internal teams fall apart — design consistency under volume.
Thumbnails are the secret static play
If you only ship video, you still need static — for thumbnails. YouTube’s own creator data and TikTok/Reels A/B testing both find 15–40% CTR improvement from custom-designed thumbnails vs. auto-generated frames. On Reels and TikTok, the first frame is the thumbnail, and a poorly-chosen first frame caps the ceiling on every video metric downstream.
Most agencies don’t treat thumbnail design as a deliverable. The ones that do typically see 15–25% lift on video ad CTR with no other change.
The honest TL;DR
Static ads aren’t winning over video — they’re winning alongside video. The data shows DTC accounts spend more on static than video by margin. They convert better in carousel form by a 2.3–2.9x multiple on engagement metrics. They cost 40–60% less per impression. And they’re the only format that fits 30% of paid-social placements.
Most agencies don’t take statics seriously, which is exactly why we built Vibey Stills as a first-class engine — premium static ads, carousels, and thumbnail packs designed with the same care as the video work. Most clients stack it with Vibey Faces for the live-action spots, Vibey Toons for animation, and Vibey Volume for AI-built variants — the four-engine system Andromeda was rebuilt to reward.
Sources cited above: Segwise — Static vs Video Ratio analysis (67,000 ads) · Lebesgue — Facebook Carousel Ads benchmarks · MHI Growth Engine — Single Image vs Carousel (894 ads) · Meta Engineering — Andromeda announcement