How to use the calculator
Fill in the pieces you have, or plan to make, for one product or one test:
- Visuals or concepts: the distinct images or video ideas. A new scene or angle counts as a new visual, the same photo in a new colour barely does.
- Hooks or headlines: the opening lines you want to try.
- Primary texts: the body copy options.
- Sizes per ad: one to three, such as 4:5 for feeds plus 9:16 for Stories and Reels.
- Formats: image only, or an image and a video of every visual.
- Target cost per purchase: the most you can pay for a sale and still be happy with it.
- Spend per ad before judging: how many times your target cost per purchase each ad gets before you call it — 2× by default, anywhere from 1× to 3×.
- Daily test budget: what you'll spend a day on this test.
The calculator shows how many ads your pieces make, the files and lines you'd need, and two budgets with the days each one takes: one for testing every combination, one for testing concepts first.
The formulas and assumptions
| Result | Formula |
|---|---|
| Every combination | visuals × hooks × texts × formats |
| Visual files to make | visuals × sizes × formats |
| Copy pieces to write | hooks + texts |
| Budget to test everything | combinations × multiple × target cost per purchase |
| Budget to test concepts first | visuals × formats × multiple × target cost per purchase |
| Days to reach verdicts | budget ÷ daily test budget |
Three assumptions sit behind these numbers:
- Sizes don't multiply ads. Ads Manager lets one ad carry a feed image and a Stories and Reels version, so a 4:5 and a 9:16 of the same idea share one ad and one set of results. They're more files to make, not more ads to test. Meta renames this option from time to time. Look for placement or asset customisation.
- Every ad needs its own verdict budget. The multiple is the rule of thumb from our guide to how many ad creatives to test: at about 2× target, a clear loser has shown itself, by 3×, a winner selling on target has had room for three sales.
- Concepts first means one ad per visual and format, each carrying your best hook and text. You test the hooks and texts later, on the visual that wins.
A worked example
Say you sell a cold brew concentrate and your target cost per purchase is $5.15. You have 4 visuals, 3 hooks and 2 primary texts. You want each ad in 4:5 and 9:16, and an image and a video of every visual. You'll judge each ad at 2× target, or $10, and you can spend $41 a day.
| Result | Sum | Answer |
|---|---|---|
| Every combination | 4 × 3 × 2 × 2 | 48 ads |
| Visual files to make | 4 × 2 × 2 | 16 files |
| Copy pieces to write | 3 + 2 | 5 lines |
| Test everything | 48 × $10 | $495, about 12 days |
| Concepts first | 4 × 2 × $10 | $82, about 2 days |
Nine pieces of creative — four visuals, three hooks, two texts — turned into 48 ads. Testing all of them costs six times as much as testing the concepts, and takes nearly two weeks at $41 a day.
Treat the days as a floor: Meta doesn't split an ad set's budget evenly, so some ads take longer than others to reach their share.
Test everything or concepts first?
Most of those 48 ads differ by a single line of copy. They share visuals, they compete for the same people, and Meta will push most of the budget towards a few of them anyway, so many will hardly spend. Practitioners also report that Meta treats very similar ads as largely the same, which is part of why genuinely different ideas count for more under its current delivery — see our guide to Andromeda and creative diversity.
Concepts first splits the job into two rounds:
- Find the visual. Launch one ad per visual and format — 8 ads in the example — each with your best hook and text.
- Tune the words. Test your hooks and texts on the winning visual. That's at most hooks × texts ads: 6 in the example, or $62 at 2× target.
The two rounds come to about $144 instead of $495, and they answer the bigger question first: the gap between two ideas is usually much wider than the gap between two headlines.
Testing every combination still has its place when you have one proven visual and want the best copy for it. And if you let dynamic creative mix the pieces inside one ad, as the 3-2-2 method does, the combinations aren't separate ads: you budget per image or video, which is the concepts-first figure.
How to act on the result
- If the concepts-first test runs past a week, drop a visual or a format, or raise the daily budget. Tests that drag on get muddied by fatigue, sales and changes to your site.
- If making files is the bottleneck, start with one size per visual and make the second size for the ads that win. Likewise, you can make videos only of visuals that have already sold as images.
- If you have more hooks than budget, keep the three that differ most — say a problem, a result and an offer — rather than three wordings of one idea.
- Then plan the month. The creative test calculator shows how many ads your monthly budget can give a verdict to.
If producing the files is what holds you back, BulkAds' AI ad generator makes batches of up to 12 image or video ads from your real product photos, in 1:1, 4:5 and 9:16.
Questions people ask
How do I calculate the number of ad combinations?
Multiply the options you're mixing: visuals × hooks × primary texts × formats. Four visuals, three hooks, two texts and two formats make 48 combinations. Sizes don't multiply ads, because one ad can carry a feed size and a Stories and Reels size.
How many ad variations should I test at once?
As many as your budget can give a verdict to, at roughly 2–3× your target cost per purchase each. For most brands that means 3–6 genuinely different ads per test, then variations of whichever one wins.
Do different sizes of the same ad count as separate ads?
No. Ads Manager's placement customisation lets a 4:5 feed image and a 9:16 Stories and Reels version sit in one ad, so they share one set of results. They are separate files to make, though.
Should I test every combination of hooks, visuals and copy?
Rarely. Testing one ad per visual with your best hook and text, then testing hooks and texts on the winner, usually costs a fraction as much and answers the bigger question first.
What is a creative testing matrix?
A grid of the elements you want to test — visuals, hooks, texts and formats — and the ads their combinations make. It shows how quickly the number of ads grows, so you can pick the combinations that answer a real question.
Keep going
- Features
AI ad generator
On-brand image ads in batches of up to 12
- Free tools
Creative test calculator
How many new ads you can test, and how long until every one has a verdict
- Free tools
Ad hook generator
About 30 fill-in hooks for your product, across eight angles
- 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.
- Guides
The 3-2-2 method: twelve ad combinations, one ad set
Three creatives, two primary texts, two headlines: how the 3-2-2 method works, how to set it up today, how to read the results and when to move beyond it.
- Glossary
Ad variation
An ad variation is a version of an ad that changes one or more elements — hook, visual, copy, call to action or size — while keeping the same core idea. Variations help you find the best execution of a concept.
The tool runs entirely in your browser. Results are estimates to plan with — your own account’s numbers always win.
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.