A video ad variant is a version of the same ad with one thing changed: the opening, the length, the shape, the voice or the call to action. Running variants side by side is how ad creative testing works. Instead of guessing which ad will do best, you let the audience decide, then put more budget behind the winner. This guide covers what to change, how many variants to start with and how to read the results without fooling yourself.
A variant is not a new ad
The point of a variant is that it is almost the same. If you change the music, the opening, the offer and the voice all at once and the new version wins, you will not know why. Change one thing at a time and every test teaches you something you can use in the next ad. That is the difference between testing and just making more ads.
What to change, in order of impact
- The hook: the first one to three seconds. This decides whether anyone sees the rest, so it is nearly always the first thing to test.
- The length: a 6, 15 and 30 second cut of the same ad reach people differently.
- The shape: 9:16 for Reels, TikTok and Shorts, 4:5 or 1:1 for the feed, 16:9 for YouTube.
- The message: price, quality, speed or a story, as the main reason to buy.
- The person or voice: a different presenter, accent or language.
- The call to action: shop now, learn more, book a call, or an offer.
Hooks first, because the opening has the biggest effect on everything that follows. We look at openings in detail in promo videos that win the first three seconds.
How many variants do you need?
Fewer than most people expect. Every variant needs enough budget and views to show a real difference, so splitting a small budget across twenty versions usually means none of them gets a fair test. A sensible starting point:
- Small budget: three to four hooks on the same ad, in one shape.
- Medium budget: three hooks, each in two lengths, so six versions.
- Larger budget: start with hooks, then test lengths, shapes and languages on the winner, in rounds.
Ten versions is a good number for a first round across two or three platforms. That is why an Ad Variant Pack delivers one master ad as ten versions ready to test.
How to run a fair test
- Show each variant to the same audience, at the same time, with the same budget.
- Change one thing per round, and write down what it was.
- Let each version run long enough to get a meaningful number of views before you decide. A winner after one afternoon is often luck.
- Use the testing tools built into the ad platform where you can, so the audience is split for you.
Which numbers to watch
- Hook rate: how many people watch past the first three seconds. This tells you which opening works.
- Hold rate: how many watch to the end or to the halfway point. This tells you whether the rest of the ad keeps them.
- Click-through rate: how many click. This tells you whether the message and call to action land.
- Cost per result: what each sale, lead or booking costs. This is the number that pays the bills, and the one that decides the winner.
A variant with a great hook rate and a poor cost per sale is still the losing ad. Use the early numbers to understand why something worked, and the final number to choose.
Plan variants before the ad is made
Variants are much cheaper when they are planned at the start. If the studio knows you need three hooks and three shapes, it films or generates the extra openings and frames each shot so it can be reframed for vertical, square and widescreen. Asking for variants after the ad is finished often means reshooting. Our AI Commercial is built this way, with 15 and 6 second cutdowns from the same master, and AI UGC Ads are written with several hooks from day one.
Ads stop working, so keep testing
Even a winning ad gets tired. As the same people see it again and again, results slowly drop. The fix is not to start from nothing, but to keep the winning structure and test new hooks on top of it every few weeks. A monthly plan on the pricing page suits that rhythm: a set number of new videos every month, which can be fresh variants built from what already worked.
A simple starting plan
- Week 1: run three hooks on one 15 second vertical ad.
- Week 2: take the winning hook and test 6, 15 and 30 second versions.
- Week 3: test the winner in a second shape or a second language.
- Every month after: three new hooks on the best ad so far.
If you sell skincare or beauty products, we cover the formats worth testing in video ads for skincare and beauty brands. Or book a free 15 minute call and we will plan your first round of variants with you.



