Thinking · Newsletter
What commerce brands can take from mobile gaming user acquisition
Gaming advertisers run tens of thousands of creatives per campaign while ecommerce advertisers run about a thousand at the high end. The gap is not budget. It is four operating habits, three of which transfer cleanly.
This piece is now part of a longer guide: What commerce brands should take from mobile gaming UA.
Mobile gaming has been buying performance media at brutal efficiency for a decade longer than most consumer brands. The habits that came out of that pressure are worth stealing, and a couple of them are worth leaving behind.
The gap, stated plainly
Gaming advertisers routinely run tens of thousands of creatives against a single campaign. Ecommerce advertisers run around a thousand at the high end. That is roughly a 50x difference, and it is not explained by budget.
It is explained by how the two disciplines think about what an ad is. Gaming treats creative as a supply problem. Commerce treats it as a craft problem. Both are right, and only one of them scales on a channel that rewards volume.
Transfers cleanly
1. Concept systems instead of finished films. A gaming team does not commission an ad, they commission a concept and define what can vary inside it before anything is shot: opening, performer, problem frame, demonstration, ending, length. One approved idea becomes thirty assets. A commerce team that commissions a film gets one asset and a schedule.
2. Reading below the asset. A single ROAS figure cannot tell you what to make next. Gaming teams decompose performance by hook, format, visual device and message, so a result becomes a reusable pattern rather than a lucky asset.
3. Designing the first second as its own variable. Not as part of the edit. As a thing you produce five versions of and test independently, because it carries more variance than anything else in the asset and it is the cheapest part to change.
Transfers with care
Exaggeration. Gaming's "cheesy" convention works because it compresses stakes into two seconds with the sound off. The compression transfers. The tone does not, and in supplements, finance or anything where scepticism is the barrier, exaggeration reads as a warning.
The fail-state device. Deliberately doing the thing wrong so the viewer feels superior. Brilliant in puzzle games. It only works in commerce if your product has an obvious wrong way to use it.
Does not transfer
Leading with the mechanic. In gaming, the mechanic is the product, so showing it is showing the thing. In commerce it almost never is. This is the mistake gaming people make on their first consumer account, and we have made it ourselves more than once: an ad that opens on how the product works rather than why it matters.
What the volume habit actually buys
Specificity. This is the part most often missed.
Volume is not about flooding the auction. It is that thirty assets can each address one motivation, while three assets have to address the average of all of them. Nobody buys on the average. The reason gaming teams out-convert is not that they made more ads, it is that they could afford to stop generalising.
AppLovin's June 2026 data makes the same point from the other direction: campaigns using every available creative beat manually curated sets by 7 to 23 percent, because different creatives win in different pockets of inventory.
The operational change this requires
If you take one thing, take this: change what a brief produces.
Today most commerce briefs produce a deliverable. A gaming-style brief produces a system, with the axis of variation stated up front and the number of expected outputs written down. That single change moves a team from producing four assets a month to forty, with no increase in shoot days.
It also changes who you can work with. A partner who thinks in films will quote per film. A partner who thinks in systems will quote per concept and tell you how many finished assets a shoot day yields. Ask.
Questions people ask
Does gaming-style creative look wrong for a premium brand? The tone does. The operating discipline does not, and the two are separable. You can run a restrained, well-art-directed library of ninety assets.
Do we need a bigger team? Usually not. You need a different brief and a production partner set up for variation rather than for delivery.
Where does AI fit? In the variation layer, not the idea layer. Generative tools are genuinely good at the fifteenth framing of a working concept, and still weak at hands, products and anything that has to be believed as testimony.
Is any of this specific to AppLovin? The volume argument is sharpest there because the inventory is so heterogeneous. The concept-system habit pays off on every channel.