Meta

Six formats, one Low score

Priya runs Meta ads for a skincare brand. Her ad set has six active creatives: two polished statics shot on a light marble surface, a 30-second UGC video from a founder testimonial, a carousel of before-and-after results, a short-form Reel cut from the same UGC clip, and an animated text-on-colour asset. By most media buyers' standards, that's a solid creative mix. Different formats, different lengths, different visual treatments.

She opens Ads Manager on September 1, 2026, adds the new Creative Diversity column, and sees: Low.

If that scenario sounds familiar — or if you're about to run the same experiment — this post explains what the metric actually measures, why a Low rating is more consequential than a label, and what you'd need to change to move the needle.

What Meta launched on August 26, 2026

Meta added Creative Diversity to Ads Manager on August 26, 2026 as a column you can add at the ad set level. It's not a notification or an alert — you have to go looking for it under Columns > Customize Columns, search "Creative Diversity," and add it to your view.

The metric rates each ad set Low, Medium or High based on how visually similar your active creatives are to one another. Meta hasn't published the exact formula, and the metric itself carries a label that matters: "estimated and in development." That phrase means the thresholds will change, the model is being refined, and you shouldn't treat today's score as a permanent verdict. But it also means Meta is actively iterating on this, which makes it worth taking seriously now rather than waiting for a polished version.

The two companions you also need to add

Creative Diversity doesn't sit alone. Two related columns were surfaced at the same time:

Creative Fatigue — flags specific creatives that have been shown to your target audience so often that performance is starting to decline. This is an existing concept in Meta's ecosystem, but having it as an explicit column makes the audit faster.

Top Creative Themes — groups your active creatives into thematic clusters based on visual similarity. This is what makes the Creative Diversity score legible: instead of just telling you your variety is Low, it shows you which themes dominate your ad set, so you can see that four of your six creatives share the same "product on marble, bright lighting, no talent" visual signature.

Used together, these three columns give you a live creative health dashboard inside Ads Manager without needing a separate tool.

Why most accounts score Low — and why format variety isn't the answer

Here's the myth the metric is busting: running multiple formats does not equal creative diversity.

Early reports from accounts that tested the column show ad sets with statics, videos, UGC, and carousels all active simultaneously still landing a Low rating. That outcome surprised a lot of buyers, because the instinct in the industry has been to treat format variety as a proxy for creative variety. "We have a Reel, a static, and a carousel" has been shorthand for "we're not hammering the same creative."

What Meta's algorithm appears to be reading is visual similarity at a deeper level — colour palette, framing, scene composition, talent presence, and probably more signals from its computer vision models. Two videos can be completely different formats (a 15-second Reel and a 60-second testimonial) and still look visually near-identical if they both feature the same person, same background, same product positioning, and same lighting setup. From Meta's model perspective, those aren't two diverse creatives — they're the same visual concept in different wrappers.

The same logic applies to statics. A product-on-marble shot and a product-on-linen shot might feel meaningfully different to you because you spent a day on the photoshoot. To a computer vision model trained on hundreds of millions of ads, they're both "clean background, centred product, no talent, warm tones." That's one visual theme, not two.

Why your Creative Diversity score affects delivery

This is where the metric moves from interesting to operational.

Meta's auction is built around giving the delivery system options. When you run multiple creatives in an ad set, Meta's system tests them against different audience segments, times of day, placements, and contexts to find where each creative performs. The more visually distinct your creatives are, the more genuine options the system has to match the right look to the right person.

A Low Creative Diversity score signals that your options aren't really options — they're variations of the same choice. That constrains the system in a few ways:

• Reach efficiency drops because there's less creative surface area to rotate against different sub-audiences
• Frequency climbs faster because fewer genuinely different looks are available to show to the same person
• CPMs tend to rise over time as the system exhausts your effective creative range and you're effectively competing with yourself

None of this is unique to the new metric — it's how Meta's auction has always worked. What Creative Diversity does is make the diagnosis explicit rather than leaving you to infer it from declining ROAS and rising frequency numbers.

Creative Diversity vs Creative Fatigue: what each one tells you

You can have a High Creative Diversity score and still have individual creatives hitting fatigue — for example, an ad set with five genuinely different visual themes, but one strong performer that's been running for three months and is now exhausted. You can also have zero fatigue signals and still score Low on diversity, if you launched a batch of fresh creatives last week that all look the same. Both columns are necessary.

How to audit your creative mix properly

The Creative Diversity column will tell you that you have a problem. Top Creative Themes will show you what the problem looks like. Here's how to use them together.

Step 1: Add all three columns at the ad set level

Go to Ads Manager > Ad Sets view > Columns > Customize Columns. Add Creative Diversity, Creative Fatigue (at the ad level for the second one), and Top Creative Themes. Sort by Creative Diversity ascending to surface your Low-rated ad sets first.

Step 2: Open Top Creative Themes for each Low-rated ad set

This view clusters your creatives into visual themes. Look for ad sets where one theme contains three or more of your active creatives. That cluster is your problem. It's not about how many creatives you have — it's about how many visually distinct themes you're running.

Step 3: Map your creative inventory honestly

Pull your active creatives and ask for each one: what does it look like to someone who has never seen your brand? Describe it in five words: "marble background, product centred, no talent." Now see how many of your creatives share that same five-word description. If more than two share it, you don't have creative diversity — you have creative variations.

For performance marketing teams managing multiple D2C clients, this audit becomes especially valuable because visual sameness tends to compound when multiple rounds of creative briefing pull from the same reference library or the same photographer.

Step 4: Brief genuinely different visual territory

"Different" in Meta's context likely means different along several dimensions simultaneously:

• Scene and environment: product-in-use in a kitchen vs. an aspirational outdoor setting vs. a messy, real-world bathroom counter
• Talent and human presence: no talent vs. close-up face vs. lifestyle shot with full body vs. diverse demographic
• Colour temperature and palette: warm neutrals vs. high-contrast bold colours vs. cool clinical whites
• Camera treatment: static flat lay vs. handheld UGC texture vs. motion-designed text animation
• Narrative angle: problem-first vs. transformation vs. ingredient focus vs. social proof

If your next batch of creatives differs on only one of these dimensions, Meta's model will likely cluster them with your existing creatives. Briefing for genuine visual diversity means changing multiple dimensions at once.

What "estimated and in development" actually means

The caveat matters. Thresholds that produce a Low rating today may produce a Medium in three months — not because your creative mix changed, but because Meta's calibration did. Use the score directionally, not as a target to chase. If your ad set scores Low, the underlying problem (visual sameness constraining delivery) is real regardless of where the threshold sits.

Audit checklist before your next creative review

• Add Creative Diversity, Creative Fatigue, and Top Creative Themes columns to your Ads Manager ad set and ad views
• Sort ad sets by Creative Diversity ascending — start with your Low-rated ad sets
• Open Top Creative Themes for each Low-rated ad set and identify which visual theme dominates
• For every active ad, write a five-word visual description; group creatives that share the same description
• Pause any creatives flagged under Creative Fatigue that aren't contributing to spend or results
• Count how many genuinely different visual themes your ad set covers — if it's fewer than three, your next creative brief needs to solve that
• When briefing new creative, specify different scene environments, talent presence, colour palettes, and camera treatments — not just different formats or aspect ratios
• After launching new creatives, re-check Creative Diversity within 48–72 hours once Meta has enough data to classify them
• Treat the score as directional — optimise for real visual variety, not for the label

A Low score isn't a punishment. It's a signal that the system has fewer genuine options to work with than you think it does. The accounts that address that have more creative surface area for Meta to test against more audiences, which compounds in delivery efficiency over time.