Variations vs new concepts
Every ad is built from parts: the hook, the main visual, the headline, the primary text, the call to action, the format and the size. Change one or two parts and keep the idea, and you have a variation. Change the idea itself — the angle, the story, who it's for — and you have a new creative concept.
The two answer different questions:
| New concept | Variation | |
|---|---|---|
| The question | "What should we say?" | "How should we say it?" |
| Example for Dewy, a fictional serum | Morning routine vs ingredient story vs gifting | Three headlines on the morning-routine ad |
| When to use it | Early testing, or when winners fatigue | Once a concept has proved it can sell |
How variations multiply
Variations stack up fast. Four hooks × three visuals × two copy versions × two sizes is 48 combinations — more than most budgets can test one by one. The ad variation calculator shows the count and the budget needed to test it, which is usually the moment you decide to cut the list.
Two ways to keep it manageable:
- Change one thing at a time when you want a clean answer, such as "does the price headline beat the benefit headline?"
- Let Meta mix the parts with dynamic creative or flexible ads, when finding a good combination matters more than knowing exactly why it won.
Variations and Meta Andromeda
Meta's newer ad retrieval system is widely understood to reward genuinely different creative. Practitioners report that very similar ads — the same image with a new button colour, say — tend to be treated as largely the same ad, so they add little on their own.
A balance many teams use: spend most of the testing budget on distinct concepts, then make a handful of meaningful variations — a new hook, a new setting, a new format — of the concepts that win. Differences a viewer would actually notice are the ones worth a test slot.
Making variations of a winner
When an ad starts selling, variations are how you extend its life before creative fatigue sets in: the same angle and offer, with a fresh opening, a new scene or a different format. In BulkAds you can ask for variants of any winning ad, and on the tests it launches itself, Autopilot drafts four fresh variants of every winner automatically.
Questions people ask
How many variations should each concept get?
Enough to give the idea a fair chance, but not so many that the budget spreads too thin. A few meaningful variations is a common starting point. Add more once the concept proves it can sell.
Is changing the button colour a real variation?
Technically yes, but it's rarely worth a test slot. Changes people barely notice tend to produce results you can't tell apart. Test differences a viewer would describe out loud.
What's the difference between a variation and an iteration?
People use the words loosely. An iteration is usually a variation made in response to data — for example, a new hook on a winning ad because its hook rate has dropped.
Related terms
Put it to work
- Free tools
Ad variation calculator
How many ads your hooks, visuals and copy make, and what testing costs
- Guides
Product photo ad ideas: turn one packshot into 20 ads
One clean packshot is enough for a full creative test. Eleven angles, how to brief each one, the sizes and copy to make, and how to launch 20 ads and learn from them.
- Guides
How many ad creatives should you test? The number, and the maths.
A straight answer, then the simple sum that sets your real number, how many new ads to launch each week at your spend, and when to stop testing and scale.
BulkAds AI is an independent product and is not affiliated with, endorsed or sponsored by Meta Platforms, Inc. Facebook and Instagram are trademarks of Meta Platforms, Inc.