Home/Insights/Meta can now turn your static ads into video. Keep the static, animate it, or shoot real video?

Meta can now turn your static ads into video. Keep the static, animate it, or shoot real video?

Static vs video ads now has a third option. What the sourced data says, why Meta's 3% figure cannot be checked in one account, and how to test it.

BY Esh, FOUNDER, ETHOS ENGINEPUBLISHED OCTOBER 16, 20267 MIN READ
KEY TAKEAWAYS
  1. Static vs video is now three options. The new one, an AI-animated static, costs almost nothing to make.
  2. The only method-stated dataset favours text-led statics. Most video-wins tables cite nothing.
  3. Meta's 3% figure is a conversion rate, not CPA. A CPM rise above 3% cancels it.
  4. Run the animation as its own ad beside the static and decide on CPA, not views.

The static versus video question used to be about production budget. Video cost a shoot and an editor. A static cost a designer and an afternoon. On 6 October 2026 Meta removed most of that difference. This article sets out what changed, what the evidence on formats actually supports, and how to decide for one ad account.

What did Meta switch on?

At Advertising Week New York on 6 October 2026, Meta said video generation from static images in Advantage+ creative is now generally available. Relevant Audience and Common Thread Collective both report the same status. Ads Creative Studio, which adds editing for hooks, audio, clip length, text overlays, captions and end scenes, is rolling out gradually. Turning catalog product images into video is a beta for eligible advertisers. Meta named no countries for any of it.

The developer documentation shows the mechanics. The Advantage+ creative feature called image animation turns a static image into a short animated video with subtle motion. It applies to image ads only, is marked as AI-generated, and is switched on or off per ad. Ads that use AI-generated features must be created paused, previewed, and then activated. That preview step is the one control you should not skip.

TikTok offers a similar tool. Symphony Creative Studio can generate video from an uploaded image and a short motion prompt. Its help page says every exported video carries an AI-generated label, and its Seedance 2.0 model cannot use images that show real faces. If the label worries you, our piece on what an AI label costs on ads covers the evidence.

So static vs video is now three choices

  1. Keep the static. Cheapest to make and fastest to change.
  2. Animate the static. Near-zero production cost, video placements, but the same picture underneath.
  3. Produce real video. A new person, scene, demonstration or story. Costly, slow, and the only option that adds a new idea.

The second option is new. Most of the advice ranking for static vs video ads was written before it existed, so it compares the first and third only.

What does the format evidence actually show?

Less than the confident tables suggest. We traced the figures that appear most often.

ATTN Agency's March 2026 table puts Meta video at a 1.9% click-through rate against 1.1% for static, and video CPA at $48.20 against $34.50. It says the analysis is backed by $47 million in spend but gives no source, sample or method. AdAmigo states that video has 34% lower CPA than static, while its own table implies about 13%. Segwise relays an analysis by Curtis Howland of 67,000 ads from more than 100 top accounts. The median account ran about 61% static and 39% video. That is a count of ads, not spend, and it comes only from successful accounts.

Motion's Creative Benchmarks 2026 is the exception. It covers 578,750 creatives from 6,015 Meta accounts and $1.29 billion of spend, launched between 1 September 2025 and 1 January 2026. A winner is a creative that spent at least 10 times its account's median and at least $500. On that basis, text only and product image with text ranked first and second among more than 15 asset types, on both hit rate and spend relative to volume. Animation and carousel sat near the bottom. High production ranked fifth on hit rate and tenth on spend use.

Motion lists its own limits. The window includes Black Friday, which favours offer-led creative. Winners are defined by spend, not profit. The findings are associations. Even so, it is the only dataset here that tells you how it was built.

Where do the sources disagree, and who is right?

The video side rests on engagement. Meta said at Advertising Week that more than 60% of the time people spend on Facebook and Instagram goes to video, without a method. ATTN and AdAmigo report higher click-through rates for video.

The static side rests on cost. ATTN's own CPA figures favour static. Motion's spend winners lean towards text-led statics.

Both can be true at once. Video may earn more attention and clicks while static wins on cost per purchase, because cheaper impressions and a direct offer outweigh the extra clicks. Which one wins for you depends on the job. The traceable evidence does not support a fixed ratio such as 60/40 or 70/30.

How much is Meta's 3% worth?

On the 29 April 2026 earnings call, CFO Susan Li said advertisers using Meta's video generation saw "more than 3% higher conversion rates in tests." Meta did not describe the tests or the comparison group. Advertisers who choose a new tool may differ from those who do not.

Take the number at face value and run the cost arithmetic. CPA equals CPM divided by the product of 1,000, click-through rate and conversion rate. A 3% rise in conversion rate lowers CPA by about 3% only if CPM and click-through rate stay the same. If the animated version is shown more in Reels and Stories and its CPM rises by more than 3%, the gain is gone.

There is also a detection problem. To tell two versions apart at 80% power and 5% significance, you need roughly 15.7 divided by the squared relative difference in conversions per version. That is our approximation for count data. For a 3% difference it comes to about 17,400 conversions per version. For a 25% difference it is about 250. In practice you can only see large effects in one account. Our article on how many concepts your budget can test works through the same constraint.

A worked example (hypothetical)

A hypothetical skincare brand runs a product-with-text static at a $12 CPM and a $30 CPA. It switches on image animation as a separate ad.

VersionCPMConversion rate changeCPA
Original static$12.00none$30.00
Animated, same CPM$12.00+3%$29.13
Animated, CPM at break-even$12.36+3%$30.00
Animated, Reels-heavy CPM$13.00+3%$31.55

The break-even CPM is $12.36. Any delivery shift that raises CPM by more than 36 cents erases Meta's claimed lift. The brand would also need hundreds of conversions per version to see anything smaller than a 25% gap. If the animation's CPA lands within that noise, the honest reading is no proven difference, so keep whichever is cheaper to maintain.

Does an animated static count as a new creative?

Probably not as a new concept. A Meta spokesperson told Marketing Brew on 28 September 2026 that Meta does not discuss how its models represent individual creative assets, and that its advice is simply to vary creative, test it, and leave optimisation to the system. In the same piece, Olivia Cripps of Billion Dollar Boy said two videos of the same person with different on-screen text count as the same creative.

An animation keeps the same product, framing and colours as its source. Our reading, which Meta has not confirmed, is that it adds a format and video placements but not a new idea. Jon Loomer reports testing 25 and 50 ads per ad set with no clear benefit. Animating every static doubles your ad count without doubling your concepts.

Which option fits which job?

SituationChoiceWhy
Offer, price or sale messageKeep the static, test animation as a second adText-led statics lead the only method-stated data
Product needs showing in useProduce real videoA demo is a new idea; motion on a still is not
Ad set with only statics and no Reels spendAnimate the best staticCheap entry to video placements
Concepts all look alikeProduce new video or new staticsAnimation adds format, not a new concept
Under about 250 conversions per version in the test windowDo not test animation aloneYou cannot detect a gap smaller than 25%
Real faces in the image, on TikTokUse another Symphony model or skipSeedance 2.0 cannot use real faces
Strict brand or claims reviewAnimate only after previewMeta requires a paused preview for AI features

A seven-step method

  1. Pick your best static by CPA over the last 28 days. Do not animate losers.
  2. Duplicate it as a separate ad with image animation on. Leave the original untouched.
  3. Preview the generated video on every placement. Reject warped products, wrong text or invented details.
  4. Run both in the same ad set for at least two weeks, or until each has enough conversions to show a 25% gap.
  5. Read CPM and placement mix first. A rising CPM tells you whether delivery moved to Reels and Stories.
  6. Decide on CPA and, if you have it, cost per new customer. Ignore views and thruplays for this decision.
  7. Spend the saved production budget on one genuinely new video concept. Check the opening against the hook measurements each platform uses.

What this changes

The cost of video has collapsed for one specific kind of video: a still image that moves. That makes it worth testing, cheaply and separately. It does not make it a substitute for a new idea, and the evidence that it beats a good static is a single unexplained platform figure. If you want a second view of your own creative mix, join the waitlist or write to esh@ethosengine.io.

Questions

Do video ads beat static ads on Meta?

Not on any evidence we could trace. Most comparison tables give no sample or method. Motion's study of 578,750 Meta creatives, the largest with a stated method, ranked text only and product image with text as the asset types most likely to become spend winners. Its window included Black Friday, which favours offer-led statics.

What does Meta's static-to-video feature do?

Advantage+ creative can generate video from your existing still images, and Meta called this generally available on 6 October 2026. In the Marketing API the related image animation feature adds subtle motion to a static image, and it is opt in or opt out per ad. Ads using AI-generated features must be previewed before they run.

Is the 3% conversion lift from Meta's video generation real?

It is Meta's claim, made by CFO Susan Li on the 29 April 2026 earnings call, about tests Meta has not described. It concerns conversion rate, not cost per result. Detecting a 3% difference in your own account would need roughly 17,000 conversions per version, so treat it as a prior, not a promise.

Does an animated static count as a new creative?

Meta has not said. An agency told Marketing Brew in September 2026 that the same person with different on-screen text is treated as one creative. An animation of one image keeps the same look, so plan on it adding a format, not a new concept.

Sources

  1. Relevant Audience, Meta's Advertising Week 2026 ad tools, what is live, in test or coming relevantaudience.com
  2. Common Thread Collective, Meta's New AI Agent Can Now Run Your Ads commonthreadco.com
  3. Manual Focus, Meta and TikTok AI ad tools from Advertising Week, a UK guide manual-focus.co.uk
  4. Meta for Developers, Get Started with Advantage+ Creative developers.facebook.com
  5. VectorShift, Meta Platforms Q1 2026 earnings call vectorshift.ai
  6. PPC Land, Meta Q1 2026, $56.3B revenue as AI tools double advertiser adoption ppc.land
  7. Motion, Methodology and definitions, Creative Benchmarks 2026 motionapp.com
  8. Motion, Top asset types, Creative Benchmarks 2026 motionapp.com
  9. Motion, Top visual formats, Creative Benchmarks 2026 motionapp.com
  10. Motion, Winning ads are rare, Creative Benchmarks 2026 motionapp.com
  11. ATTN Agency, Static vs Video Ads, Which Format Performs Better in 2026? attnagency.com
  12. AdAmigo, Meta Ads Benchmarks by Creative Format (2026) adamigo.ai
  13. Segwise, Static vs Video Ratio for Meta Ads, Data From 67,000 Ads segwise.ai
  14. Marketing Brew, How advertisers are adjusting to Meta's ad creative diversification best practices marketingbrew.com
  15. Jon Loomer Digital, The Truth About Meta Andromeda and Ad Retrieval jonloomer.com
  16. TikTok Business Help Center, How to generate video from an image with Symphony Creative Studio ads.tiktok.com
E
Esh

Founder, Ethos Engine. Builds the Digital Authority Score and runs the placement marketplace. esh@ethosengine.io

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →
MOREINSIGHTS