Since TikTok changed the pace of content consumption, an idea has taken hold and gets repeated as if it were law: "video always outperforms static images". It's an oversimplification that has cost more than one business real money after pouring its entire budget into video production without first checking whether that was what its audience and product actually needed.
The reality, backed by real campaign data and not just the intuition of the moment, is more nuanced: video wins in specific contexts, static images win in others, and in plenty of cases performance is practically the same, with the difference that the image costs a fraction to produce.
Where video has a clear edge
Video works better when the product needs demonstration: an appliance whose value lies in how it's used, a service whose result is better understood by seeing it in action (a renovation, a cosmetic treatment, a fitness class), or anything where motion conveys information a photo can't. It also performs better in native video formats like Reels, Stories or TikTok, where the platform's own design favours moving content and penalises, with less added organic reach, whatever clearly reads as a static image dropped in among videos.
Another area where video tends to win is building trust quickly: a video with a real person talking to camera (founder, customer, employee) conveys authenticity that an image, however well produced, has a harder time getting across in three seconds of scrolling. For service businesses where trust is the main purchase barrier (consultancies, clinics, training), this matters a lot.
Where static images compete head-to-head (or win)
For products whose value is understood at a glance (fashion, home decor, food, anything essentially visual and static by nature), a well-lit, well-composed photograph can perform as well or better than a video, at a much lower production cost. Images also win in the main Facebook and Instagram feed for older audiences, where consumption is still more unhurried than in Reels-type formats, and where a clear image with legible text gets processed faster than a video that requires staying through the first few seconds to understand what it's about.
There's a third, less intuitive scenario where images clearly win: the bottom of the funnel, when someone already knows your brand and just needs the final nudge. There, a product carousel with clear price and availability tends to convert better than a video, because the person has already decided and wants quick information to close the purchase, not to be convinced all over again from scratch.
Production cost, the variable almost nobody factors in
Here's the nuance pure performance comparisons usually ignore: a well-produced video costs, on average, between five and fifteen times more than a similar-quality product photograph. If the video performs 10% better but costs ten times more to produce, the image can still be the more profitable option in terms of return on total investment, not just media spend. This is especially relevant for small businesses with a limited production budget: sometimes the right decision isn't "which format performs better" but "which format performs better per euro invested from start to finish, production included".
The middle ground almost nobody uses: low-cost video
Between professionally produced video and static photography there's an underused middle ground: phone-shot videos with simple editing, which on many platforms (especially TikTok and Instagram Reels) perform better than polished video precisely because they fit the native style of the content people are already consuming there. A thoughtfully made "homemade" video, with good natural light and a clear message, can outperform an expensive, very "ad-like" production, because it doesn't interrupt the scrolling experience the way something that "smells" like a professional ad does.
How to decide in practice, without guessing
The most honest way to settle this isn't to follow a general rule, it's to test both formats with the same message and the same audience, and let your specific business's data decide. As a reasonable starting point: if your product needs demonstration or your edge is personal trust, start by testing video. If your product is strongly visual and quick to decide on, start with images and add video as a complement, not a replacement. Either way, don't drop the "losing" format entirely too soon: ad fatigue means whatever performs worse today might perform better in a few weeks simply for being the thing seen least recently.
A case where the image won despite the team's intuition
An artisan candle brand was convinced video would be the winning format, given how visual the making process is, and invested in a 30-second video showing the wax being poured and the candle finally lit. As a comparison, they also uploaded a simple product photograph against a neutral background with good natural light, with brief copy about scent and burn time. Against the team's every expectation, the static image generated a cost per sale 34% lower than the video, with a production budget ten times smaller.
Analysing why, the likely explanation was twofold: the product (a candle) is fully understood at a glance, with no need for motion to convey its value, and the target audience (people aged 35 to 55 browsing mainly the main feed, not Reels) consumed content at a more unhurried pace, where a clear image with price and scent information visible got processed faster than waiting for the video to reach that same information. The lesson wasn't "images always win", it was that not even the team that knows its product best can reliably predict which format will perform better without testing it.
How to combine both formats within the same campaign
Instead of framing the decision as a single choice for the whole campaign, many accounts get better results by letting the platform itself split the budget across several formats within the same ad set: uploading both video and static images and letting the algorithm, with real data from that specific account and audience, decide how much budget to assign to each based on the performance it observes. This usually delivers better results than deciding upfront, based on the team's intuition, which of the two formats deserves the entire budget.
The hidden cost of producing badly, whether video or image
Beyond format choice, there's a factor that ruins performance for both equally: poor technical production quality. A video shot in bad light, with poor audio, or with an editing pace that gets boring within the first two seconds performs worse than a simple but well-lit, well-framed photograph. The honest comparison is never "video versus image" in the abstract, but "the best video I can produce with my current budget versus the best image I can produce with that same budget", and that comparison changes quite a bit depending on each business's actual resources.
For businesses with a very tight production budget, it's usually more profitable to invest in noticeably improving the quality of a few product photographs (natural light, good framing, basic colour editing) than to produce a mediocre video for lack of budget to do it properly. A simple, well-executed format almost always beats a more ambitious, poorly executed one.
How to turn the same photographic material into motion
For businesses with no video budget but good photography already available, there's a lesser-known middle ground: turning static photographs into small animations (slow zoom, transitions between several photos, text appearing progressively) using simple editing tools like Canva. This format, halfway between image and video, can capture part of the extra attention motion generates on platforms like Reels without needing to shoot anything new or take on the cost of a full video production.
Common mistakes when choosing between video and image without your own data
The first mistake is being guided by what works "in general" for the sector without checking it against your own business. Every account has its own audience, its own accumulated level of ad fatigue, and its own relationship with each format, and a broad sector trend rarely predicts exactly what will happen in a specific account. The second mistake is switching formats based on a single week of data, when any piece's performance can vary quite a bit week to week for reasons unrelated to the format itself (seasonality, a one-off spike in competition, changes to the platform's algorithm). The third mistake, perhaps the most common among small businesses, is assuming that because video is trendy in general marketing discourse, it has to be produced no matter what, even when that specific business's product and audience respond better to images. Following a general trend without checking it against your own data is the fastest way to spend production budget unnecessarily.
Frequently asked questions
How long should an ad video be to perform well?
In formats like Reels or TikTok, between 6 and 15 seconds usually captures attention best without losing the viewer before they reach the key message. In videos with more informational content or for older audiences in the main feed, up to 30-45 seconds can work well if the content holds interest.
Does the carousel (several swipeable images) count as a static format or is it its own category?
It's its own category, with its own advantages: it lets you tell a short sequence (several products, several steps, before and after) without video's production cost. It tends to perform especially well in ecommerce with a varied catalogue.
Is it necessary to have subtitles on ad videos?
Practically mandatory. A notable majority of users watch social media video with the sound off, especially on public transport or at work, so a video without subtitles loses a good part of its message for a large share of the audience.
Should I use the same format across every platform I advertise on?
No, each platform has its own visual language: what performs well on TikTok (very native, unpolished video) can perform worse on LinkedIn, where more polished content with more context tends to work better. Adapting the format to each platform gets better results than recycling the same piece everywhere.
Is it worth investing in animation or motion graphics for static ads?
It can be an interesting middle ground: it costs less than a video with real people or locations, and adds enough motion to catch attention on platforms that favour dynamic content, without the cost of a full audiovisual production.
How do I know if my product "needs demonstration" or not?
A simple way to check: if, when explaining the product to someone who doesn't know it, that person immediately understands what it's for with just a photo and a sentence, it probably doesn't need video. If you have to explain how it's used, how it's assembled, or what it looks like in motion for it to fully make sense, video adds value an image can't match.