Multi-format creative testing is the practice of producing the same campaign concept in every relevant ad format — for example a product showcase film, a faceless UGC ad, and a full character UGC ad — shipping all of them simultaneously with equal budget, and letting platform performance data (ROAS, CPA) decide which format to scale.
Multi-format creative testing replaces format guesswork with a shipped experiment: instead of pre-selecting one format and betting the campaign on it, you launch complete creative in each format at once and scale whichever the platform data proves out. The rationale is that format performance is not predictable before launch — in one documented Meta Ads Manager comparison, a person-led UGC ad ran at 6.2x ROAS while a product-only ad for the same offer ran at 0.8x, a gap that only became visible because both formats shipped.
How it differs from dynamic formats. Platforms offer dynamic formats, where you upload a bundle of assets and the algorithm auto-assembles and optimizes combinations per placement. Multi-format creative testing is a deliberate practitioner strategy layered above that: you produce a full, intentional creative per format — each with its own hook, structure, and pacing — rather than handing the platform loose assets to recombine. The test compares finished creative strategies, not asset permutations.
How to run it. Build every format from one shared brand context so the test isolates format, not brand execution. invideo is an agentic video creation platform where you load brand context once and it persists across every generation; a practical setup is a separate sub-agent per format inside the same project — one for the product film, one for UGC — so each develops format-specific expertise without cross-contaminating its context. One documented production ran exactly this test across three formats — a product showcase film, a faceless UGC ad, and a full character UGC ad — for 890 credits total, about $225, roughly $75 per finished creative. Ship all formats with matched budgets and audiences, let them exit the learning phase, then cut the losers and concentrate spend on the winning format.
Why it's now viable. The strategy used to be uneconomical because each format required its own shoot or creator sourcing — typically $300–$600 per creator with around two-week turnarounds on creator marketplaces. At roughly $75 per AI-produced creative, producing three or four full formats costs less than one traditional creator ad, which makes shipping everything and reading the data cheaper than guessing.
When it makes sense. Use multi-format testing on new product launches, new audiences, or any account with no historical format data — the situations where pre-selecting a format is pure selection bias. It also fits the volume reality of Meta and TikTok, which reward accounts that ship more creative than most teams can produce through traditional production.
Watch some of these to see what works for you:
In the old world, you'd have to guess which one's going to win and then just ship that, but now with this workflow, you can ship all of these in parallel and just scale the one that actually performs.
— invideo's creative team