Meta Ads Creative Testing: What to Test, How Many and How Often

A lot of Indian D2C brands tell us the same thing about Meta. The first few ads do well, then after some weeks the cost goes up, and nobody knows why. Most of the time, the ad just gets old to the people seeing it, or the testing wasn’t set up well from the start.

This post gives you an overview of Meta advertising creative testing in layman’s terms: what to test first, how many ads to test and how often to refresh them. We have used ₹ in the calculations below as a substitute for your budget to give you a more realistic perspective.

Why creative matters more now

Meta decides the target audience for your ad in almost all cases. With both broad targeting and Advantage+ campaigns, Meta will see the ad and decide who would be interested.

A better ad reaches better people, and a weak ad is hard to fix with audience settings.

Because of this, Meta ad creative testing is usually worth more of your time than playing with interests. If a brand spends ₹50,000 a month and never tests new ads, it’s skipping the part that matters most.

What to test first

Some tests are not worth running. Changing a button colour usually changes nothing, while changing the idea of the ad can change a lot. This is the order we’d go in:

Within one test, change only one thing. If the idea, the hook and the creator are all different, you can’t tell which one made the difference. Across your account, though, you do want ads that look different from each other. Meta seems to treat ads that look almost the same as one ad, so ten copies of the same video with different text colours is really just one test.

When we plan tests at Influencer Bazaar, we start with the concept and the script, then pick creators, then optimise. Small edits come at the end.

How many ad creatives to test at one time

How many ad creatives to test depends on how much you spend. People online say anything from 3 or 5 per ad set up to 20 or 30, and none of these is a fixed rule.

Your budget is a better way to decide. Meta wants roughly 50 optimisation events in 7 days for an ad set to settle down. If one purchase costs you ₹500, that’s around ₹3,500 a day for one ad set. If you spend ₹1,000 a day, you get about 14 purchases a week, so the ad set stays in learning. Putting 20 ads into it just spreads the small budget even thinner.

These are rough starting points from me, not rules.

Under ₹1,500 a day: 3 to 5 ads that are clearly different, in one ad set. If you can’t get enough purchases, optimise for add-to-cart for now.

₹1,500 to ₹5,000 per day: 6 to 10 ads, either in one ad set or two.

Greater than that: 10 to 15 ads or even more, and refreshed on a routine basis.

How often to refresh Meta ad creatives

The frequency of refreshing Meta ad creatives is dependent on your budget and your audience size and not on any predetermined date. If your audience size is smaller and your budget is higher, then an ad will lose its effectiveness in just two weeks. Otherwise, it can continue longer than that.

Static images usually get tired faster than videos.

This is what works for a lot of accounts:

• Small spend: 2 or 3 new ads every 3 to 4 weeks
• Medium spend: 3 to 5 new ads every 2 weeks
• High spend: new ads every week

Don’t stop a winning ad only because it’s old. If the numbers are fine, keep it running and add the new ads next to it. Add them as a batch instead of one every day, because each change can send delivery back into learning.

Creative fatigue Meta ads: signs to watch:

Fatigue builds up over several days. Look for a few of these happening together:

• Frequency going up
• CTR going down
• CPM going up

Cost per purchase is slowly rising while your website conversion rate stays the same

If an ad drops after two or three days, it’s usually not fatigue. Check the learning status, any budget changes and your tracking first. People often mix up creative fatigue. signals in Meta ads with a stuck learning phase.

A simple weekly line chart of CTR and frequency for each ad is enough. When frequency goes up and CTR goes down, that ad is close to done.

Advantage+ creative: what it does and doesn’t do

Advantage+ Creative is a set of automatic changes Meta makes to your ad. This includes cropping, expanding the image for Reels and Stories, adding music, changing the text, and, in some accounts, AI-made versions. Some of these are often turned on by default, so it’s worth checking your account.

We’d turn it on for some ads and keep the same ad with it off, then compare the results. Look at the previews before launching, mainly the crops and any text on top of the image. Turn it off when the look has to be exact, like when the colour of the product matters.

Advantage+ creative only makes different versions of your idea. It can’t come up with a new idea. If all your ads say the same thing, they will still get tired.

UGC ads for Meta:

UGC ads for Meta look like a normal Reel, so people don’t skip them as fast as a studio ad. They’re also usually quicker and cheaper to make in batches, which helps when you’re testing a lot.

From one shoot, you can get 6 to 9 ads. Take 3 creators and allow them to experiment with 2 or 3 hooks. Keep the video between 20 and 35 seconds; include the hook in the first 2-3 seconds; feature the product use early in the video. Allow the creator to use the language that your customers use. Hinglish is fine.

We work with 100+ brands and 500+ creators at Influencer Bazaar, and the UGC ads for Meta that work best for us are usually simple, with one person talking about one problem and one product. Our packages follow the same idea. Starter is for first tests, Growth has 8 to 12 video creatives, and Scale has 15+. In one campaign based on user-generated content, the ROAS increased 3X in 30 days. It is heavily dependent on the product, its price, and the offer, so please don’t take it as a guarantee.

Here is a simple workflow that you can apply:

• Pick one product and one goal, usually purchases.
• Write 3 concepts using real customer reviews and questions.
• Make 2 or 3 hooks for each concept, so you have 6 to 9 ads.
• Launch them in one ad set with enough budget, and leave it alone for a few days.
• Judge each ad after it has spent about 2 to 3 times your target cost per purchase. If your target is ₹500, that’s ₹1,000 to ₹1,500. Also pay attention to the number of people watching past the first 3 seconds.
• Keep only the winners and abandon those that have failed.
• Write the next batch based on what won. If review-style ads did well, try more review hooks.
• Do it again every 2 to 3 weeks.

If you don’t want to run all this yourself, our Meta ads and creative teams at Influencer Bazaar can handle it together, from strategy and creators to editing and reporting. You can start with a free strategy call.

Where to start this week

Meta ad creative testing does not require much preparation. Choose the number of ads that fits the budget; vary one element at a time; analyse the creative fatigue in the Meta ads insights weekly; choose how frequently the Meta ad creatives should be refreshed. And then repeat the same with the next set.

FAQs

Q1. How many ad creatives to test simultaneously?

With a small budget, 3 to 5 ads that are clearly different. With a bigger budget, 10 to 15 or more. Small edits of the same ad don’t count as separate ideas.

Q2. How long should I run a test?

A few days at least. Judge each ad after it has spent about 2 to 3 times your target cost per result.

Q3. How often to refresh Meta ad creatives if my budget is small?

Every 3 to 4 weeks is usually fine, with 2 or 3 new ads. Do it sooner if frequency is going up and CTR is going down.

Q4. Is a sudden drop in results always creative fatigue?

No. Check the learning status, budget changes, tracking and your landing page first. Fatigue usually shows up slowly.

Q5. Should I keep Advantage+ creative on?

Test it. Keep it on for some ads and off for others, check the previews, and turn it off where the visuals have to be exact.

Q6. Do UGC ads for Meta work better than studio ads?

Often they do, but not always. Test both. Many brands find UGC easier to make in large numbers.

Q7. Is it possible to do Meta Ads creative testing with ₹500 per day?

Yes, but make sure you don’t overdo it. Use 3 ads, tweak only one variable at a time, and optimise for a more frequent event, like add to cart.

Q8. Should I pause an ad when frequency goes up?

Not only because of frequency. Pause it when frequency is going up, CTR is going down, and the cost per result is rising too.