What Meta Isn't Telling You About Your Creative

It’s time for Meta to share this information.

So one of advertisers’ longest-running complaints relates to an embarrassing lack of transparency about creative performance. There’s no granularity.

One example is when using flexible format. When you create an ad using flexible format, you can provide up to 10 images or videos. Meta will mix and match to show different creative with your text to get the best performance.

The problem is that Meta provides no details about the images and videos that were used. The breakdown by dynamic creative element has never worked for flexible format.

I initially thought this was just a bug or oversight, but it’s been several years now. This lack of functionality is clearly intentional at this point.

Meta does offer a breakdown by flexible format now, but that doesn’t fix this problem either. This breakdown only creates rows for whether single media or carousels created from your creative were used. There’s no detailed information about the specific images or videos that were used.

The same problem exists when using related media. Related media allows you to select images and videos that you’ve previously used to promote the same product. That way, you can give Meta many different options of creative within a single ad.

But once again, the breakdown by related media doesn’t share important details. It generates rows of results for when your original media or related media were used.

And this same issue exists for AI-generated creative, too. If you select AI-generated images or videos, which are available to some advertisers, there’s very little transparency. The breakdown by image generation will only provide results for your original media and all AI-generated creative.

Now, understand that I mostly don’t care about all of this.

The danger of getting these details is that you’ll overreact to the results. You’ll see that one image is getting you the best results, so you’ll edit your ads to only include that one version.

Advertisers are notorious for making these micromanagement mistakes related to small sample results. It assumes that Meta wastes money on the low-performing options. This leads to the breakdown effect, which is where advertisers look at low-volume results and assume they’ll scale.

But advertisers shouldn’t obsess over the winning image, video, or copy these days. A successful ad will have many different winning combinations for different people and placements.

While I say that, I also understand the desire to get this detailed information.

I discussed the possibility of creating one monster ad with lots of creative options in my private community recently. You can do this using related media. One member asked how you’d know which images performed best. My response was that you can’t, but it doesn’t really matter anyway.

Look at results in aggregate and worry less about finding winning creative.

But another member made a good point. If you’re employing a creative team, they need feedback about what works and what doesn’t. If they create 20 images and Meta’s only using two of them, that’s important information to know.

It’s not about finding the winner in this case, but instead about having information to better direct resources. And that’s an argument that makes a whole lot of sense to me.

Yes, Meta needs to provide this level of detail in their results, preferably using breakdowns. And yes, there’s a very real risk that advertisers will misuse that information, as they tend to do.

But it can also be used for an actual benefit that Meta could surely get behind. It would help us provide direction so that we can design images and videos most likely to make an impact. Otherwise, we’re mostly flying blind.

So here’s the bottom of the glass.

Meta desperately needs to fill this information gap. They have the data, but they’ve chosen not to share it with us.

While it’s true that advertisers will misuse this information, Meta can use that as an opportunity to educate them. Help advertisers understand what they can do with this information, rather than keeping it from them entirely.

And that can help us make better ads that resonate with our audience, rather than guessing about what works and what doesn’t.