Business · 6 min
What a playable ad actually costs
Agency quotes, in-house hours, per-network variants and the cost of a rejection. Where the money goes and which parts are genuinely fixed.
Ask three suppliers for a playable and you will get three structures: a per-unit fee, a monthly retainer, or a rate card per network variant. The confusion is not accidental — the format's cost is dominated by rework, and rework is hard to quote.
Where the hours actually go
Concept and brief — small, and the part with the highest leverage on outcome.
Design and asset prep — moderate, and mostly reusable across variants.
Build — the bulk of it, and almost entirely mechanical once the mechanic is settled.
Network variants — a multiplier on the build, not on the concept. Four networks is not four ads.
Rejection and rework — unbudgeted, schedule-destroying, and usually caused by a compliance string or a WebView incompatibility rather than by the creative.
The cost that never appears on the invoice
A rejected unit costs the media plan, not the build fee. Creative review at the network is a human step measured in days, and it happens after your launch date has been set. That is why deterministic pre-export checks are worth more than another round of visual polish.
What changes when the build is generated
The concept and the judgement cost the same as they always did. The build, the four variants, the ES5 transpilation, the packaging and the analytics wiring collapse toward zero marginal cost. That flips the economics of testing: the sensible number of playable concepts per campaign stops being one and starts being four, because the expensive thing is no longer the ad — it is finding out which mechanic the audience responds to.
Budget for iteration rather than for a masterpiece. Ship three mechanics, read completion and choice distribution, then put the media behind the one the audience picked.