What D2C Brands Get Wrong About Creative Testing
Written By
Armend Meha
Most brands think they're testing creatives. They're actually just launching random ads and hoping. Here's how to structure your creative production for real testing.
Testing creative is one of those things every D2C brand says they do, but very few actually set up properly from the production side. Most of what passes for "testing" is producing one new ad, sending it to the media buyer, and hoping it works.
Real creative testing requires structured creative production — variations designed intentionally so the performance team can isolate what's working and why.
The mistake: random variation
The most common problem we see is brands requesting creatives that are all completely different from each other. Creative A has a UGC-style photo with a benefit-driven headline. Creative B has a product-on-white shot with a curiosity-driven headline. Creative C is a carousel with a totally different angle.
When the media buyer tests these and B wins, what has anyone learned? Was it the photo style? The headline approach? The format? Nobody knows, because everything changed at once. The media buyer found a winner but has no insight into why it won — so the next production brief is another round of guessing.
How to produce creatives for proper testing
The key is designing creative batches where one variable changes at a time. This is a production discipline, not a media buying discipline. It starts with how the creatives are built.
When a client wants to test headlines, we produce 5 variations using the same image, same layout, same colors — only the headline changes. When the media buyer runs these, the winning headline is clear because it was the only variable.
When a client wants to test visual styles, we produce variations using the same headline and copy — only the image or photo style changes. Same logic: the performance data points directly to which visual approach resonates.
This sounds simple, but it requires discipline in production. It's tempting to make each ad feel "unique." We resist that urge. Structured creative batches produce insights. Random variation produces noise.
The creative variable hierarchy
When we plan a production batch, we recommend clients test in this order, from highest impact to lowest:
Hook or headline — this is the single biggest lever. The headline determines whether someone stops scrolling. We typically produce 5 to 10 headline variations first before testing anything else. It's the fastest way to find what resonates.
Visual style — once the media buyer has flagged winning headlines, we test visuals. Product shot vs. lifestyle. UGC-style vs. designed. Bright vs. muted. Clean vs. busy. Same winning headline, different visual treatments.
Format — same concept, different ad format. We'll take a winning static and produce it as a carousel, a story placement, and a square format. This tells the media buyer where the audience prefers to engage.
Layout and design details — colors, typography, text placement, copy length. These matter, but less than people think. We test them last, once the hook and visual approach are locked in.
Give the media buyer enough to work with
The number one complaint we hear from media buyers and performance managers is: "I don't have enough creatives to test properly." They need volume — not random volume, but structured batches designed for testing.
If you're only producing 3 to 5 new creatives a month, the media buyer can't run meaningful tests. By the time they have data, the creatives are already fatiguing and they're back to asking for more. A healthy production pipeline delivers 10 to 20+ new creatives per month, structured into testable batches.
The monthly production rhythm
We structure creative production on a monthly cycle that aligns with how media buyers test:
Week 1 — Deliver a new batch of creatives. Focus on new hooks and concepts for the media buyer to test.
Week 2 — The media buyer shares early signals: which creatives are promising, which aren't working. We produce iterations on the winners.
Week 3 — Deliver iterations plus a few brand new concepts. The media buyer has fresh assets before the first batch fatigues.
Week 4 — The performance team shares what worked and what angles resonated. We use those insights to plan the next month's production batch.
This cycle means the media buyer always has fresh creatives, every production batch is informed by real data, and the creative quality improves month over month because we're building on what actually works — not guessing.
