Published benchmarks put video ahead on click-through and static ahead on cost per acquisition, at the same time. What is actually going on, and a decision rule that does not depend on picking a winner.
HighQualityUGC Team
Editorial
We run UGC ad tests daily and publish what holds up: real credit costs, real hook rates, no vendor fluff.
If you go looking for a definitive answer to this you will find one, twice, pointing in opposite directions. One source has Meta video CPA at $48.20 against static at $34.50. Another reports video delivering 34% lower CPA. Both are 2026 figures from credible-looking analyses.
Rather than picking whichever supports the decision you already made, it is worth understanding why the numbers disagree. The explanation is more useful than either number.
Three reasons, and all of them apply to your account too.
Once you see the third one, a lot of published comparisons look shakier.
| UGC video | Static image | |
|---|---|---|
| Cold prospecting | Strong. Can carry an argument to a stranger. | Weak alone, cheap to test angles with |
| CPM | Higher; roughly 38% above static in cold | Lower, which buys more reach per dollar |
| Click-through | Higher, around 27% in aggregate reporting | Lower, but clicks are often higher intent |
| Retargeting | Good for objection handling | Excellent for reminders and offers |
| Catalogue and dynamic ads | Rarely applicable | The native format |
| Production cost per variant | High unless generated | Very low |
The row that decides most cases is the last one. If the viewer does not already understand what the product is or why it matters, a static image cannot teach them. If they do, a static image is a much cheaper way to close.
Forget the format comparison and ask what the ad has to accomplish.
That fourth point is the underused one. Static ads are a cheap research instrument for video production. Test the claims as images, then produce video only for the claim that won. It removes most of the risk from a video budget.
The usual framing is that video costs more to make. That is true per asset and it is not the number that matters.
What matters is cost per performing ad, which is cost per asset divided by hit rate, plus the cost of the tests that failed. Static wins on the first term by a wide margin. Video often wins on hit rate in cold traffic. Which one wins overall is genuinely account-specific, and it changes when your production cost per video changes.
That last clause is the thing that has actually moved recently. Much of the received wisdom about static being the efficient choice was formed when a video meant a shoot. If a video variant costs a fraction of what it used to, the arithmetic that made static the sensible default no longer holds in the same way, and the honest position is that this comparison should be recalculated rather than inherited.
Not one format. A rough shape that works for most DTC accounts:
That last instruction is the one most likely to change something. Most accounts compare video and static on a blended cost per acquisition without controlling for which audiences each was serving, which reproduces exactly the confound that makes the published benchmarks disagree.
Both, at different things. Reported benchmarks put video around 27% higher on click-through while static acquires customers around 28% more cheaply with roughly 38% lower CPMs in cold audiences. The conflict comes from comparing formats that are usually doing different jobs in different parts of the funnel.
Three reasons: the datasets contain different mixes of advertisers and categories, static is disproportionately measured in high intent retargeting and catalogue placements, and static is cheap enough that only the best ones survive while an expensive video runs even when it is mediocre.
When the viewer already understands the product and needs the offer, price or a reminder; in catalogue and dynamic placements where static is the native format; and for cheap angle testing before committing to video production.
| Explaining something new |
| The only real option |
| Cannot do it |
When the viewer has to be convinced of something they do not currently believe, or when the product needs to be seen working. A static image cannot demonstrate, and it has no time in which to build an argument.
Separately by funnel position, never blended. A single cost per acquisition across both formats reproduces the exact confound that makes published benchmarks disagree, because the two formats are usually serving different audiences at different intent levels.
