Recent post

Every veteran buyer had the same reaction to this news: didn’t we already have this? The Meta creative breakdown, the newly publicized way to see how individual images, videos, and text variations perform inside one ad, sounds like something Ads Manager has done for years. Part of it has. Part of it hasn’t. The useful question is which part is which, and what you can safely do with the numbers.

Our short version: the Meta creative breakdown is helpful, it’s less new than the headlines make it sound, and it doesn’t solve iOS measurement. Use it to decide what to cut and what to brief next. Use your MMP to judge which ads bring in valuable users. Neither source gets to be the whole truth, which is why creative and UA have to be read together.

The News, and Why It Sounds Familiar

On October 27, 2026, Social Media Today reported that Ads Manager now shows the performance of individual creative elements through its Breakdown menu. Choose Media and you see how each image or video performs when it’s delivered as single media. Text variations get their own view. Other coverage describes the path as Breakdown, then Creative, then Media, with text variations under the Dynamic creative element option. Earlier coverage said the rollout was gradual, so you may not see it yet.

Now the déjà vu. Dynamic Creative dates back to 2017, and older guides show breakdowns by image, video, headline, text, and call to action, with Ads Manager screenshots from 2018. The text half of this new feature lives in that same Dynamic creative element menu. In July 2025, Meta’s Creative breakdown for Flexible formats was covered as a per-element view. Asset breakdowns have also been unreliable at times. Advertisers have reported Image and Video breakdowns returning nothing on multi-image Flexible ads while Text and Headline worked.

So is anything in the Meta creative breakdown new? One piece, yes.

What Is New: The Media View

The new piece is the Media option, which splits results by each image or video inside a single ad. It’s the view buyers have wanted since ads started carrying several assets at once, and it makes the Meta creative breakdown a real improvement over guessing which asset did the work.

Here’s our read, and it’s an interpretation. Coverage in July 2026 described the same Breakdown, then Creative, then Media path for Multi-Media Ads, which hold up to 10 images and videos in one ad, and said it was arriving gradually with a new ad creative experience. If that’s right, October looks more like a second wave of coverage for a feature already showing up in accounts than a brand-new launch.

Also our read: packing up to ten assets into one ad is part of what blurred the ad-level view in the first place. Meta is partly fixing a problem its own formats created. That doesn’t make the fix less useful. A tool that arrives late is still a tool, and plenty of buyers will use it this week.

Five Limits of the Meta Creative Breakdown

Before you build a process around it, know where it stops.

  1. Single media only. It covers images and videos delivered as single media. Ads shown as a carousel or a collection don’t appear.
  2. Some actions are missing. Custom conversions and some actions aren’t supported. Social Media Today names Donate and Subscribe as examples.
  3. It isn’t a controlled test. Meta decides which asset gets served to whom, so every comparison is shaped by delivery. That’s our view, and we think it matters more than anything else on this list.
  4. The breakdown effect. This is Meta’s own term. A segment can look expensive on average and still be the right place for budget. Cut it on average cost alone and you can make the whole account less efficient.
  5. Low-volume winners. An asset with a handful of results and a great cost per result deserves suspicion until the volume shows up.

What the Meta Creative Breakdown Can’t Tell You on iOS

We found no source that explains how this breakdown treats iOS conversions. That’s worth saying out loud. The likely answer is that it inherits Meta’s existing attribution and modeling, but that’s inference, and we’d be happy to be wrong in the direction of more data.

Here’s what we can say. Meta Ads Manager reports modeled conversions using its own attribution logic. SKAN data from Meta is aggregate. According to one major MMP’s documentation, the dimensions Meta shares through SKAN are install, conversion value, campaign name, and campaign ID. There’s no ad or creative dimension on that list. Meta may also redact values when volumes get too low.

Meta’s AEM runs alongside SKAN and passes ad and creative IDs to that MMP, according to that same documentation, along with ad set ID, campaign ID, and placement. It supports a limited set of campaign types, so check whether yours qualify. How much of that reaches your reports depends on the MMP, so ask yours which IDs it receives for iOS install campaigns.

There’s also Meta’s Advanced Mobile Measurement program. One MMP’s product notes list its re-launch on June 18, 2025, as an opt-in program that gives device-level data, without saying whether that data carries ad or creative detail. Another MMP’s help page still describes the program as deprecated in 2021. Those two don’t agree, so ask your MMP what is available to you today.

What does that mean for the Media breakdown? Maybe the asset-level numbers on iOS are modeled. Maybe they lean on AEM. Maybe it’s a mix. Nobody outside Meta can say from public documentation. Until someone can, treat the iOS rows as directional.

Meta vs. MMP: Who to Believe, and for What

Why do Meta’s numbers and your MMP’s numbers disagree? Attribution windows, view-through credit, modeling, and deduplication all drive differences. Meta reports modeled conversions with its own logic. An MMP measures cross-channel attributed installs and the downstream events that follow. Each is answering a slightly different question, so each can be right and still disagree with the other.

Here’s what each source is good for:

Meta’s Media breakdown works at the individual asset level, using Meta’s own attribution. It’s best for deciding what to cut and what to brief.

An MMP with AEM data can work at the ad or creative ID level, depending on the MMP and the campaign type. It’s best for comparing ads on downstream events and value, where that data exists.

SKAN through an MMP works at the campaign level, in aggregate. It’s best for checking trends.

There’s one wrinkle, and it’s an inference, so treat it that way. Ten assets packed into one ad may look like one creative in an MMP’s attributed data, because the ID the MMP receives belongs to the ad. Some MMPs also offer asset-level numbers, and from what we can tell those are the network’s own figures passed through. One MMP guide notes that Meta reports at the asset level for Flexible and Advantage+ formats and that the combinations served may be unknown. Check how your own MMP handles it before you rely on any of this.

So which creative won? Depends who you ask. Meta may point to the third video. Your MMP may point to the ad that contains it and say those players left after a week. Both can be true. A cheap result tells you an asset got attention. A valuable user tells you the attention was worth buying. You need the pair.

How to Read the Media Breakdown Without Fooling Yourself

The Meta creative breakdown is a good shortlist maker and a poor judge. Here’s how we’d sort what it shows you.

What to Trust

Trust the big, boring signals. An asset that keeps spending with nothing to show for it across several days. An asset Meta barely delivers after a fair run. A difference between assets wide enough that noise can’t explain it. These are the findings we’d act on, usually by cutting, since cutting an obvious dud is the cheapest decision to undo. You can always put the asset back in a new ad.

What to Treat as Directional

Small differences between assets. Anything in the first days after launch. Results on events Meta has to model, which on iOS may be most of them. Comparisons between assets that reached different people or placements. Any cost number sitting on a handful of results.

Use these to form a hypothesis for the next brief. Don’t use them as a verdict. If the second video’s hook seems to pull better, that’s a good reason to brief three more hooks in that direction. It’s a weak reason to kill the other video.

When to Isolate a Creative in Its Own Ad

Isolate when the decision is big: putting serious budget behind a concept, killing one, or explaining to a client why something won. Put the asset in its own ad with the same audience and budget settings as its siblings. It still won’t be a lab test, but delivery has one less variable to hide behind.

If you need your MMP to see a creative separately, a separate ad is the likely route, since the IDs it receives are tied to the ad. That part is our inference. Try it in your own account before you bet a reporting process on it.

Isolating has a cost. You split volume and you can slow down learning, so don’t isolate every asset. Pick the ones that matter. Which ones would you argue about in a meeting?

What We Still Don’t Know

A few open items, and we’d rather list them than guess. How does the breakdown treat iOS conversions? How fast is it reaching every account? How stable are the asset-level numbers once the feature is everywhere? What will MMPs show for ads that carry several assets? We’ll revisit this when Meta’s documentation or MMP guides answer those questions.

Meanwhile, the working rhythm is simple. Read the Media view weekly to build a cut-and-brief list. Read your MMP at a longer window to see whether the players behind those ads are worth the cost. Then put both in the same room.

The Fetch

Use Meta to brief creative. Use your MMP to judge value. If your creative team and your media buyers read those two reports in separate rooms, reach out and let’s put them on the same page.