On this pageJump to a section
Open an ad you have been running. Click preview. Does every word and every pixel on that screen come from you?
If you have creative enhancements switched on, the honest answer may be no. That is not automatically bad. Some of those changes are harmless and helpful. Others can alter your message, change how your product looks or put someone else's words under your brand. The trick is knowing which is which before the ad runs.
What are Advantage+ creative enhancements?
They are adjustments Meta can make to your ads automatically to fit placements or try to improve performance.
The standard bundle
Meta's developer documentation lists the sub-features of standard enhancements by ad type. For single image ads: image touch-ups, text optimizations, inline comments and image templates. For single video ads: video auto crop, text optimizations and inline comments. Inline comments surface relevant comments under the ad, image templates apply a template so the image fits certain placements better, and video auto crop adjusts the video's aspect ratio.1
The generative AI layer
In May 2024 Meta announced generative AI features for ads: full image variations based on an original creative, text overlays with a dozen typefaces, image expansion to different aspect ratios, and text generation to create headline variations in addition to primary text. It also said it was testing text generation that reflects a brand's voice and tone.2
The controls
The documentation shows opt-in and opt-out controls for standard enhancements, and warns that starting with Marketing API version 22.0, opting in or out of standard enhancements will no longer be available. Instead you opt in or out of individual Advantage+ creative features.1 In Ads Manager, look for the creative options on each ad and check what is switched on.
Which enhancements carry the most risk?
Think in traffic lights. This is my judgment, not Meta's.
| Enhancement | My default, and why |
|---|---|
| Video auto crop, image expansion | Green. Format fitting. Check that faces and text are not cut off. |
| Image templates, text overlays | Yellow. They change the look. Review previews against your brand. |
| Image touch-ups | Yellow. Meta's documentation does not describe the changes, so compare previews before and after. |
| Inline comments (other people's comments under your ad) | Red in regulated categories. You do not control what is displayed. |
| Text optimizations, text generation | Red for claims. The words your buyer reads may not be yours. |
| Generated image variations and backgrounds | Red when the product must be shown accurately. |
Why text is the highest risk
Your copy is where your promises live. If a system rewrites a headline, you can end up with a claim you did not write and did not review. In a category with claims rules, like health, finance, income or weight loss, that is more than a branding issue. You are responsible for what your ad says.
Why product depictions come next
Meta describes image generation as producing variations with new scenery and adjustments to the product.2 If your product has to be shown exactly, as with a device, a package, a before-and-after or a clinical service, any change to its appearance is a risk (the wider picture is in AI Ad Images Got Good). Generated pictures can drift from the truth in small ways that matter.
Why comments are a wild card
Inline comments show relevant comments beneath your ad.1 Comments are public and uncontrolled. In a calm category that can add social proof. In a sensitive one, a stray comment can make a claim you could never make yourself.
If Meta can change what your ad says, then your compliance review has to cover what Meta might say, not just what you wrote.
Do the enhancements actually lift results?
Sometimes, and Meta's evidence is advertiser-reported.
Meta's announcement cited Casetify seeing a 13% increase in return on ad spend when testing background generation, and Fresh reporting 5x incremental return on ad spend and 42% incremental conversions using Advantage+ shopping campaigns with Shops ads and text variations.2 Those are individual advertisers' results, reported by Meta. Your account is not theirs.
Meta has also been pushing automation more broadly. Social Media Today reported in February 2025 that Advantage+ optimizations switch on when advertisers use Advantage+ audience, placements or campaign budget.3 I broke that setup down in Advantage+ Sales Campaigns. The direction of travel is toward more automation, which makes it more important to know what is on.
How do you audit your ads and test an enhancement?
- List your live ads and open the creative options on each. Write down which enhancements are on.
- Preview every ad on every placement. Screenshot what you see.
- Mark each enhancement green, yellow or red for your business using the table above.
- Turn off the red ones in claim-sensitive ads until you can review every output.
- Test the rest one at a time, on the testing cadence you already run. Duplicate the ad, change one setting, and compare results at a meaningful volume.
- Record the decision and the date, so you know why a setting is on or off.
A rule for regulated advertisers
If your ads make health, financial or income-related claims, assume that any feature that rewrites words or alters how a result or product looks needs a human review of every variation, or needs to be off. The lift is not worth an account problem or a regulator's letter.
How do you keep a record you can rely on later?
If an ad is questioned by a platform, a customer or a regulator, you want to show what ran. Enhanced ads can differ from what you uploaded, so keep your own record.
- Save the originals you submitted, with the date.
- Screenshot previews of each placement as the ad appeared in review.
- Log which enhancements were on for each ad and who approved them.
- Note any change to the setting and why.
It sounds like paperwork. It is the cheapest insurance you can buy, and it makes your testing cleaner too, because you will know exactly what each ad contained when it produced a result.
Where does this leave you?
Automation is not the enemy. Unreviewed automation is. Let Meta fit your video to placements, test the safe options against a control, and keep your hands on the words and the proof.
Your words, your claims, your call. Everything else is a setting you can test. If you want that review habit built into your ad operation, see the system I build for clients.
Keep building.Ekai
Questions people ask
What are Advantage+ creative standard enhancements?
They are optional adjustments Meta can apply to your ads automatically. In Meta's developer documentation the bundle includes image touch-ups, text optimizations, inline comments, image templates for single image ads, and video auto crop for single video ads.
Can I turn off Meta's creative enhancements?
Yes. Meta's documentation shows opt-in and opt-out controls for standard enhancements, and notes that starting with Marketing API version 22.0 you opt in or out of individual Advantage+ creative features instead of the whole bundle. Check Ads Manager for the options in your account.
Which creative enhancements are riskiest for brands?
The ones that change words or the depiction of your product: text optimizations and text generation, generated image variations, and showing other people's comments under your ad. In regulated categories, I would keep those off unless every possible output has been reviewed.
Do creative enhancements improve results?
Sometimes. Meta reports advertiser examples, such as a 13% ROAS increase for one brand testing background generation, but those are single advertiser results. Test enhancements against a control in your own account.
Sources
- 1
- 2
- 3
