A creative team finds a strong video concept. The hook works, the campaign starts giving it more spend, and the idea earns another round of production. One obvious next move is to turn it into a static ad: take the strongest frame, keep the message and CTA, and launch it alongside the video.
The two assets may look like versions of the same idea, yet the user experiences them very differently. Video has time to introduce a situation, build tension, show a product in action, or reveal an offer in stages. A static creative gets one frame and a fraction of a second to make its case.
AppsFlyer’s State of Creative Optimization 2026 puts numbers behind that difference. The analysis covers 1.2 million creative variations, around 1,400 apps, and up to 140 creative elements across gaming and non-gaming verticals. Across most of the categories studied, the creative attributes receiving the most spend in video differed from those leading in static images. Tone, CTA, selling point, and social proof could all change once the format changed.
For a media buying team, that changes the starting question. A successful video gives us a useful hypothesis, while the static version still needs its own logic.
A Format Changes How the Idea Reaches the User
Video can distribute information over several seconds. The opening creates context, the next scene develops it, and the CTA arrives after the user has already seen enough to understand what is being offered. Gameplay, testimonials, product demonstrations, transformations, and story-led hooks all benefit from that sequence.
A banner has to compress the same communication into one visual decision. The offer, product, emotion, and CTA compete for attention inside a single frame, so clarity becomes much more important.
AppsFlyer describes a similar split in its data. Attributes based on progression, urgency, authenticity, gameplay, or emotional build tend to have more room to work in video. Static images favor ideas that can be understood immediately: a clear offer, a direct CTA, or one strong visual message.
That distinction also matches what we see in campaign work at ROCKAPP. Static often gives the user an immediate choice: the visual catches attention and the message makes sense almost at once, or the impression passes. Video has more room to earn attention over time, but it also asks the user to stay with the creative long enough for the important part to appear.
The Same Concept Can Survive. The Execution Usually Changes
A strong creative idea can travel between formats. The useful part to carry over is usually the underlying trigger rather than the complete execution.
In our previous article, The Creative Worked. Then Everyone Copied It, we looked at how successful mechanics move between advertisers and even between verticals. A rescue scene, a familiar game mechanic, or a particular emotional trigger can inspire very different executions once another team adapts it to its own product.
The same principle applies when a concept moves from video to static. The idea may stay, while the way it communicates has to change.
Imagine a video built around a before-and-after transformation. The first scene establishes a problem, the middle shows the process, and the final frame reveals the result. A static adaptation has room for a single visual moment. The team has to decide whether the transformation itself, the final result, or the product benefit carries the strongest part of the original concept.
That creates a much more useful test than taking one frame from the video and treating it as the image version.
AppsFlyer Found Very Different Winners Inside the Same Verticals
The report becomes especially interesting when the comparison moves from general format behavior to individual creative attributes.
In Action & Strategy games, an “Exciting” tone led video across the channels AppsFlyer studied. On DSP static inventory, the pattern shifted: “Dramatic” accounted for 54.1% of spend, compared with 45.9% for “Exciting.”
Puzzle & Lifestyle games showed an even wider difference around social proof. High social-proof emphasis — reviews, ratings, “used by millions” messages and similar cues — reached up to 69% of video spend across channels. On one DSP image segment, high social proof accounted for only 3.2%, while low emphasis reached up to 80%.
Shopping apps offer another useful example. In video, “Download” led a fragmented CTA mix with 29.2% of spend. In static images, “Shop Now” reached 67.8%, compared with 6.4% in video.
Finance was the unusual case in the dataset. Feature-centric content led both formats on Social & Search, taking 40.8% of video spend and 43.2% of image spend. Even there, the consistency of the broader content approach still leaves tone, hook, composition, and CTA available for format-specific testing.
These examples make a literal video-to-banner adaptation much harder to justify as a default production process. The core concept can remain valuable, while the strongest expression of that concept changes with the format.
Static and Video Also Enter Different Auctions
Creative logic is only part of the difference. Media buying conditions matter as well.
Static and video can occupy different placements and reach different parts of available inventory. Some placements support video only, while others give image formats more room. That means two creatives built around the same idea may enter different competitive environments from the moment the campaign starts delivering.
At ROCKAPP, we also often see lower CPMs on static creatives compared with video. Static production is faster and cheaper as well, which makes it useful for testing more visual and messaging variations without putting the same production cost behind every hypothesis.
Those economics make static attractive, but they also make direct format comparisons easy to misread. A cheaper CPM can give the banner an advantage at the top of the funnel, while later conversion events determine how valuable those users become. Video may enter more expensive inventory and still justify that cost further down the funnel.
The placement mix matters too. If a campaign relies heavily on inventory where static has limited access, a banner winner can only cover part of the opportunity. Video often remains important simply because it opens inventory and user experiences that an image creative cannot reach.
This is why our teams usually prefer a balance of formats rather than choosing one permanent winner.
Video Winners Are Taking a Larger Share of Spend
AppsFlyer’s spend distribution adds another layer to the format split.
Among gaming apps, the top 2% of video creatives accounted for 53.5% of video spend in Q1 2025. By Q2 2026, their share had increased to 60.7%. Static moved in the opposite direction: the top 2% of image creatives went from 51.5% to 46.6% over the same period.
The pattern suggests different testing behavior for the two formats. Video spend increasingly concentrates around a smaller group of strong performers, which makes detailed analysis of those winners especially valuable. Static spend remains more distributed, leaving more room for a wider rotation of concepts and variations.
Creative volume is growing at the same time. Gaming apps spending more than $7 million per quarter ran an average of 3,977 creative variations by Q1 2026, up from 2,743 a year earlier. In non-gaming, the same spend tier reached 3,477 variations, compared with 2,365 in Q1 2025.
For production teams, higher variation counts make structure more important. Testing ten adaptations of one concept gives much less information when each version changes the format, hook, CTA, tone, and selling point at once.
What We Would Keep When Moving Between Formats
When a video idea earns another test in static, we first look for the element that made the concept interesting in the first place.
That may be an offer, an emotional trigger, a product benefit, a recognizable situation, a visual mechanic, or a particular user problem. Once that element is clear, the image can be built around it from scratch.
A video with a strong story may turn into a static focused on the final conflict. A product demo can become a clean product-benefit visual. UGC testimonial may translate better into a short claim or a single piece of social proof. A gameplay sequence may need one instantly readable moment rather than a screenshot chosen simply because it looked good in the video.
The CTA deserves the same treatment. AppsFlyer’s Shopping data shows how sharply CTA behavior can shift between formats, so keeping identical copy across video and image can remove one of the easiest variables to test.
The useful connection between formats therefore sits at the hypothesis level. Execution stays flexible.
How We Would Test Video and Static Together
For a new concept, we would define the shared hypothesis first: the user insight, product promise, emotion, or mechanic the team wants to test.
From there, video and static get separate executions. The video can use sequencing, pacing, sound, movement, and reveal. Static needs a clear hierarchy that works immediately in one frame.
Media buying then evaluates each format in the environment where it actually runs. CPM and CTR give the first signals, while CPI and deeper events show whether the creative is attracting users who continue through the funnel.
When one format performs well and the other struggles, the concept still has useful information to give. The team can inspect the format-specific execution before retiring the underlying idea. AppsFlyer’s data makes this especially relevant because the same creative attribute can move from a leading position in one format to a much smaller share of spend in another.
That approach gives creative teams more room to learn from a winner without turning every successful video into a banner copy, or every strong static into a storyboard.
The Best Mix Depends on the Campaign
Advertisers often lean toward video because it offers more room for storytelling, product demonstration, emotion, and native-looking formats such as UGC. Static remains valuable for a different reason: it is quick to produce, usually cheaper to run, and capable of communicating a clear idea immediately.
For ROCKAPP, the stronger setup usually includes both. Their roles differ across placements, auctions, GEOs, and campaign objectives, and the data from each format gives the creative team a different type of signal.
AppsFlyer’s 2026 research makes that split much easier to see. Video and static can start from the same strategic idea, while each deserves its own creative decisions, testing plan, and performance read. When teams treat them that way, one winning concept can generate a much broader set of useful tests without forcing every format to behave like the one that came before it.
