Ask five media buyers whether static or video works better on Meta and you will get five confident answers, most of them shaped by the last account they ran. The format question feels like a matter of taste. It isn't, quite. There is now enough published data to say which format does which job, and in October 2026 Meta changed the economics again by letting any advertiser turn a static into a video inside Ads Manager.
Most DTC teams treat format choice as a creative preference. They pick video because video feels more modern, or they default to static because it's cheaper and easier to brief. Neither of those is a strategy. The data is clear enough now to make format decisions on evidence, not instinct, and the numbers point somewhere that surprises a lot of operators.
What the numbers actually show
Two public sources do most of the work here. ATTN Agency's 2026 benchmarks give the headline finding: video earns 73% more clicks, but static acquires customers 28% more cheaply. Segwise's write-up of a 67,000-ad Meta dataset covers the other half of the question, which is how real accounts actually mix the two. That CTR and CPA split tells you most of what you need to know about how the formats behave.
| Metric | Static | Video | Edge |
|---|---|---|---|
| Average CTR | ~1.1% | ~1.9% | Video (+73%) |
| Average CPA | $34.50 | $48.20 | Static (28% lower) |
| CPM (cold audience) | ~$15.80 | ~$21.80 | Static (~28% lower) |
| Engagement rate | ~1.4% | ~5.2% | Video |
| Creative fatigue window | 20 to 30 days | 40 to 60 days | Video (runs longer) |
| Production cost (relative) | Baseline | 5 to 10x static | Static |
Source: CTR, CPA and engagement figures via ATTN Agency's 2026 benchmarks; fatigue windows and production multiple via MHI Growth Engine, as cited by Segwise; CPM data via Needle.
Why does video pull more clicks but land a worse CPA? The mechanism isn't complicated. Video earns attention broadly. It's harder to scroll past, it generates curiosity, and the engagement numbers reflect that. But a lot of those clicks are people who were entertained, not people who planned to buy anything. Static, by contrast, communicates a specific message to a narrower audience who already has some purchase intent. Fewer people click, but a higher share of them actually convert.
Video finds the audience. Static closes them. Running only one format is leaving half the job undone.
Does this hold across all product types and funnel stages?
Not uniformly. The Segwise data shows the format split varies substantially by vertical and by where someone is in their buying journey.
| Context | Format that wins | Why |
|---|---|---|
| Cold prospecting (brand unknown) | Video | Earns attention; tells a story to someone who's never heard of you |
| Retargeting (seen the brand) | Static | Faster message delivery; lower CPM; converts intent that's already warm |
| Products requiring demonstration | Video | 20 to 40% higher conversion rate when the product needs to be seen in use |
| Simple offer / price-led creative | Static | Single clear message reads instantly; no watch-through required |
| Beauty and shapewear brands | Video (80 to 97%) | Visual transformation is the product; video shows it better |
| SaaS and tech products | Static (70%) | UI screenshots and feature callouts communicate faster than explainer video |
| Brands spending under $20K/month | Static-heavy (70/30) | Budget constraint; static gives more test surfaces per dollar |
| Brands spending $100K+/month | Video-heavy (65 to 70%) | Scale needs format diversity; algorithm rewards mixed creative pools |
The median account in that 67,000-ad dataset ran 61% static, 39% video by volume. That median is useful context. It tells you that the brands posting the most ads at scale are still leaning static, even though the prevailing wisdom tends to favour video.
Key takeaway
Static wins on CPA and cost efficiency. Video wins on engagement and cold reach. The question is never "which format is better" but "which format for this message, this audience, this stage of the funnel."
Why reliability favours static
The practical difference in production is reliability. A creator can misread a video brief and deliver footage you can't use. A static brief leaves far less room for that kind of drift, and when you are running several concepts in one batch, the gap adds up. It's one reason we almost always validate the message in static first, then brief video once we know the angle is working.
Speed matters too. A static concept can go from brief to live test in a fraction of the time, because a video concept has to pass through a creator, a shoot or generation, and an edit first. If you want to learn what's working before the next budget cycle, that gap decides how many ideas you get to test.
Why the "static vs video" framing is slightly the wrong question
The smarter question is sequence. Video earns attention with cold audiences; static closes efficiently once the message is proven. The formats aren't competing; they're doing different jobs.
The fastest way to test whether an angle converts is to strip it to its most direct form. A static card with a clean headline and product shot tests the message with almost nothing else in the way. A 30-second video with the same angle costs more, takes longer and carries more variables. If the message fails in static, you have saved the video budget.
Meta's own numbers point the same way. On its Advantage+ catalogue ads page, Meta reports that ad sets with catalogue product video saw 20% more conversions per dollar, and that letting the system choose image or video for each person delivers a lower cost per incremental conversion than single-media ads. That's not an argument for going all-in on video. It's an argument for always having some video in the mix. The algorithm benefits from format variety at the ad set level. A proper creative system keeps both in rotation by default.
Meta can now turn your static into a video. Should you let it?
This is the newest wrinkle in the debate, and most format guides haven't caught up with it. At Advertising Week New York on 6 October 2026, Meta announced that video generation from existing static assets is now generally available in Advantage+ creative, with catalogue advertisers in beta able to turn product images into product video at scale. At the March 2026 IAB NewFronts, Meta reported that beta advertisers who used its AI video generation tool for most of their campaign ads saw average gains of 10% in CTR and 8% in conversion rate, per eMarketer's coverage.
Our take: switch it on as a test, never as a substitute. A generated video built from a winning static is a cheap way to find out whether motion adds anything to a message you have already proved. That makes the static-first sequence stronger, not weaker, because your best statics now double as raw material for video variants. What it can't do is replace a creator. An animated product shot with music has none of the face, voice or lived-in proof that makes UGC video convert for low-trust products. Treat it as one more format variant in the test, give it its own ad so you can read it, and keep briefing real video for the concepts that earn it.
| Video source | Best used for | Weak spot |
|---|---|---|
| Meta-generated from a static | Testing whether motion lifts a proven message, cheaply | No human proof; looks like every other generated ad |
| Edited from existing footage | New cuts, lengths and hooks on a proven concept | Limited by what was shot |
| Creator-led UGC | Cold audiences, demonstration, low-trust products | Slowest and costliest to make, so brief it on proven angles |
Does static outperform video for UGC specifically?
Static and video are both viable for UGC creative, but they use the format differently. Static UGC tends to be a screenshot of a review or a creator quote card, something that borrows social proof rather than personality. Video UGC gives you the creator's voice, expressions, and energy, which is harder to replicate in a single frame.
In our experience, video UGC outperforms static UGC for new products where trust is low, because the creator's on-camera authenticity does work that a cropped review screenshot can't. For a brand with strong review volume and recognisable positioning, static proof-based creative often holds its own. That said, the CPA advantage of static still applies here: AI-generated UGC in static format in particular can be produced and tested fast enough to run many more variants against a single video concept, which typically surfaces a winner sooner.
A good testing loop treats both formats as inputs, not alternatives, and gives each one enough spend per creative to reach a verdict. You're not picking a format. You're building a test matrix that uses each format where it fits.
How should I split my budget between static and video on Meta?
Here's a straightforward starting framework based on Segwise's format-mix data and our own recommendations:
| Monthly Meta spend | Suggested static/video split | Logic |
|---|---|---|
| Under $5,000/month | 80% static, 20% video | Budget is too thin to run proper video tests; static gives more test surfaces per pound/dollar |
| $5,000 to $20,000/month | 70% static, 30% video | Static to find messages that convert; video for the 2 to 3 strongest concepts |
| $20,000 to $100,000/month | 50 to 60% video, 40 to 50% static | Enough budget to run meaningful video tests; cold traffic benefits from video reach |
| $100,000+/month | 65 to 70% video, 30 to 35% static | At scale, video diversity feeds the algorithm; static handles retargeting |
These splits aren't ratios to lock in forever. They're starting points. Your account's actual winner rate by format, over four to six weeks of testing, should override any generic benchmark. The point of the split is to make sure you're generating enough data from both formats to know which is working for your specific offer.
The production reality nobody talks about
The most persuasive case for static isn't a benchmark. It's iteration speed. When a static stops performing, you can have new copy and a new visual live the same day. When a video stops, you go back to a creator, wait for a reshoot, edit and test, and that cycle can take weeks while CPA climbs.
That is why "static to validate, video to scale" works in practice. Static finds the three or four messages that actually convert, and the video budget goes into those proven angles rather than ideas that felt good in a brainstorm.
The same logic applies to hooks. Six static headline variations cost a fraction of six video hooks, given video's 5 to 10x production multiple. If four of those hooks were never going to work, static found that out before you spent video money on ideas the market was going to reject anyway. Once a winner exists, the iteration playbook for scaling a winning ad covers what to change next, in which format.
None of this makes video the lesser format. It makes video the format you use once you know what you're scaling.
How to read a static vs video test without fooling yourself
A lot of format "tests" aren't tests. Static and video are dropped into the same ad set, Meta sends the video to Reels and the static to Feed, and three weeks later someone concludes that one format wins. What they actually measured was two placements, two audiences and two sets of CPMs. Before you trust a format verdict, check for these.
| What skews the result | Why it matters | How to control for it |
|---|---|---|
| Different messages | A strong static against a weak video is a message test, not a format test | Same angle, same claim, same offer in both formats |
| Placement mix | Video leans into Reels and Stories, static into Feed, so you are comparing placements | Break results down by placement before comparing |
| Attribution | Since March 2026, click-through only counts link clicks, and the engaged-view window for video dropped from 10 seconds to 5 under the renamed engage-through attribution | Compare both formats on the same attribution setting, and look at link-click conversions separately |
| Spend imbalance | Meta often pours budget into one format early, starving the other of a fair read | Give each format its own ad set or a minimum spend before judging |
| Funnel stage | Static in retargeting will usually beat video in prospecting on CPA, whatever the creative | Compare formats within the same audience, never across |
The attribution row is the one that has changed most recently. Search Engine Land reported that Meta now limits click-through attribution to link clicks and moves conversions after shares, saves and other interactions into engage-through attribution. Meta also cut the engaged-view threshold for video ads to 5 seconds, citing data that 46% of Reels purchase conversions happen within the first 2 seconds of attention (Social Media Today). An engaged view only exists for video. Statics can still earn engage-through credit from shares and saves, but not from watch time. So if your attribution setting counts engage-through conversions, video has an extra route to credit, and a CPA comparison on that view can flatter video. That doesn't make the credit fake. It means you should know which number you are reading.
Frequently asked questions
Do static ads or video ads perform better on Meta in 2026?
On Meta, static delivers about a 28% lower cost per acquisition on average, according to ATTN Agency's 2026 benchmarks. Video earns more clicks (CTR around 1.9% vs 1.1%) but those clicks convert less reliably. For most DTC brands, static wins on CPA and video wins on reach and engagement with cold audiences.
Should I test static or video first on Meta?
Test static first. It's cheaper to produce, generates usable assets much faster than video, and validates whether a message actually converts before you commit video production budget to it. Once a static concept proves the message works, video is the right format to scale reach.
What is a good static vs video split for Meta ads?
For brands spending under $20,000 per month on Meta, 70% static and 30% video is a solid starting point. At $20,000 to $100,000 per month, shift toward 50 to 60% video as you have budget to run proper video tests. Above $100,000 per month, 65 to 70% video makes sense because the algorithm needs diverse formats at scale.
Does Meta's algorithm favour video over static ads?
Meta doesn't publish a format preference, but it does favour having both. On its Advantage+ catalogue ads page, Meta reports that letting the system choose image or video for each person delivers a lower cost per incremental conversion than single-media ads. Run both formats on the same message and let delivery sort out who sees which.
How long do static ads run before creative fatigue on Meta?
Static typically shows performance decline after 20 to 30 days at moderate spend. Video lasts 40 to 60 days before significant fatigue sets in. That means you need to refresh static roughly twice as often, but each refresh costs a fraction of a new video.