Anyone seeing better results with stacked creations vs. single-ad testing?

[Fallon Desi1]

I've been running multiple creatives in a single ad set instead of isolating them, and performance seems better. Not sure if it's just campaign freshness or if the algo is actually optimizing smarter in 2025. Curious if others are seeing this too?

 

[Slorm31600]

Meta's algo in 2025 does handle mixed creatives better, it optimizes faster and spreads learning. If it’s working, stick with it. Just prune dead weight occasionally

 

[TDF Media]

I've been running multiple creatives in a single ad set instead of isolating them, and performance seems better. Not sure if it's just campaign freshness or if the algo is actually optimizing smarter in 2025. Curious if others are seeing this too?

Recently, FB's algorithm is smarter than before, gathering many creatives in 1 adset, it learns quickly, and automatically allocates budget to the best ones. But you also have to consider niche and target, sometimes it's new so it explodes, then after a few days it slows down again. I tested and found that 2025 FB optimizes better, but you should still leave some separate adsets to compare to be sure.

 

[Lana Wefun]

Recently, FB's algorithm is smarter than before, gathering many creatives in 1 adset, it learns quickly, and automatically allocates budget to the best ones. But you also have to consider niche and target, sometimes it's new so it explodes, then after a few days it slows down again. I tested and found that 2025 FB optimizes better, but you should still leave some separate adsets to compare to be sure.

Yeah I’ve noticed the same. It’s like the algo is finally smart enough to handle multiple creatives without killing delivery. Still, I agree keeping a few isolated ad sets as a benchmark makes the results more reliable.

 

[Petter Ly]

"Meta’s 2025 algorithm optimizes ad sets more intelligently, so stacking multiple ads in one set can speed up learning and budget allocation — but it’s still wise to keep a few separate ad sets for comparison and not rely on it entirely.

 

[HiddenGuyonBush]

Could be true that the algo is getting smarter, but I’m still skeptical. Results can also come from niche, audience, or simply the freshness effect. Some individual cases show improvement, but not sure it’s a broader trend.

 

[Dat2068]

I've been running multiple creatives in a single ad set instead of isolating them, and performance seems better. Not sure if it's just campaign freshness or if the algo is actually optimizing smarter in 2025. Curious if others are seeing this too?

- when you put multiple creatives in the same ad set, the system has more data to compare CTR, CVR, engagement level, etc. from there, it automatically distributes more to the most effective model
- if you split the creative into many ad sets, the data is fragmented, each sd set has few impressions, machine learning is difficult to optimize. Concentrating into 1 ad set helps the amount of data to be more stable, the learning phase will be faster and more stable
- meta has officially advised advertisers to reduce the number of campaigns/ad sets and focus on "creative testing in the same ad set. This goes along with features such as dynamic creative and advantage+ placements/campaigns

 

[Pippi Magnos]

Can i ask why does running multiple creatives in one ad set perform better than separating them? Does the algo really optimize smarter now?

 

[Gusbert]

Yeah, in 2025 the algo’s definitely gotten better at optimizing within ad sets. Running multiple creatives together can now help it find winners faster. Could be freshness too, but your results make sense.

 

[Elise Chloe]

In 2025, Meta’s algo is better at handling multiple creatives within a single ad set, which helps it learn faster and spread budget more efficiently. Still, it’s smart to keep a couple of ad sets split out, so you have benchmarks and don’t put all your testing into one basket.

 

[TBH Agency Marketing]

I've been running multiple creatives in a single ad set instead of isolating them, and performance seems better. Not sure if it's just campaign freshness or if the algo is actually optimizing smarter in 2025. Curious if others are seeing this too?

sure.it has a choice when it comes to good and bad campaigns it has the choice to keep the good campaigns

 

[SunshineAgency]

I've been running multiple creatives in a single ad set instead of isolating them, and performance seems better. Not sure if it's just campaign freshness or if the algo is actually optimizing smarter in 2025. Curious if others are seeing this too?

If it works keep it up and backtest a lot to create general statistics.

 

[Karo Tami]

- when you put multiple creatives in the same ad set, the system has more data to compare CTR, CVR, engagement level, etc. from there, it automatically distributes more to the most effective model
- if you split the creative into many ad sets, the data is fragmented, each sd set has few impressions, machine learning is difficult to optimize. Concentrating into 1 ad set helps the amount of data to be more stable, the learning phase will be faster and more stable
- meta has officially advised advertisers to reduce the number of campaigns/ad sets and focus on "creative testing in the same ad set. This goes along with features such as dynamic creative and advantage+ placements/campaigns

If we keep everything in one ad set, how does the system decide which creative to push more? Doesn't that risk some ads not getting tested properly?

 

[Willsbmize]

I’ve noticed the same , keeping creatives in one ad set seems to let the algorithm optimize better. Feels like it’s getting smarter in 2025.

 

[leisahorg]

Running multiple ads in an ad group isn't always a good thing, it also depends on your content, it will also take you a while to filter out quality campaigns to scale them up.

 

[leisahorg]

Running multiple ads in the same ad group also has the purpose of choosing the right ad, the possibility of it will be higher than running alone.

 

[TDF Media]

I think so too. I used to split adsets to test individual creatives, but recently, combining multiple creatives in the same adsets has given better results. It seems that the 2025 distribution algorithm is smarter, automatically focusing the budget on effective models without having to split it too much. It also saves micro-management.

 

[BBIM AGENCY]

The best interactive content format right now is short videos (Reels) because they are algorithmically preferred and easy to spread. You should also use live streams and polls to create direct interactions, as well as discussion-based content to encourage comments. It’s important to mix and match these types of content to find the format that works best for your audience.

 

[SEO NoraFJ]

Yes, Facebook’s algorithm in 2025 has gotten significantly better at optimizing ad sets for performance. Running multiple ads at once helps the system find winners faster, increasing overall performance. This may be due to the freshness effect of the content, but the results you get are reasonable.

 

[BET365DAYS AGENCY]

I see similar results. Combining multiple creatives into one ad set gives the algorithm enough data to learn and deliver more effectively, especially with small budgets. However, for more control, I still test them separately in the early stages and then combine them as I scale. It seems that by 2025, the algorithm will optimize faster, provided the account has a clear and stable performance history.