Meta adds product browsing to existing ads on 28 October. Which Advantage+ creative enhancements should stay on?

KEY TAKEAWAYS

1. Product browsing turns on for existing ads on 28 October unless you opt out. Meta's 3% lift claim comes without cost or revenue data.
2. Ads Manager cannot show which enhancement helped or hurt. Defaults can differ ad by ad and return after edits.
3. Placement fitting is low risk. Message and destination changes need a test before they run on scaled ads.
4. Test one class at a time against a protected control, judged on orders and cost in your own data.

PPC Land reported on 15 October 2026 that Meta had told advertisers it will add product browsing to their existing Facebook and Instagram ads from 28 October, unless they opt out. Product browsing is one Advantage+ creative enhancement among 31 that Meta lists. Most guides treat these as toggles to flip once. The harder problem is that Ads Manager cannot tell you which ones helped, so any decision needs a test you design yourself.

What changes on 28 October?

When someone taps an ad with product browsing on, they may see an in-app page with product information and personalised suggestions before reaching your site. The feature already existed as an option. The change is the default, applied to ads that are already running.

Meta's notice claims 3% more conversions on average. The supporting test ran from 6 to 21 August 2026 across about 117,700 advertisers, and Meta says the result held with over 99% confidence. PPC Land lists what the notice leaves out: which conversions were counted and where, the attribution window, cost per conversion, revenue, how the test was split, and whether new ads will also start with it on. It found no Help Center page, changelog entry or newsroom post on the change.

Relevant Audience noted the timing. The switch lands 30 days before Black Friday on 27 November. It also reported a practitioner, David Herrmann, asking Meta to delay to January because Ads Manager does not show when product browsing contributes to a result. PPC Land could not verify his client cases.

There are two ways out. The notice has an opt-out button, which works before 28 October. Each ad also has a product browsing toggle in its Advantage+ creative and destination settings, which Meta says can be changed at any time. It is not clear whether the button covers one ad account or several, so check each account.

Why can't Ads Manager tell you which enhancement worked?

Three reasons, and they compound.

Reporting stops at the ad. Leapbuzz points out that Ads Manager reports performance per ad, not per enhancement. adlibrary found no Meta figure that isolates the effect of turning off a single enhancement.

Defaults differ ad by ad. Terrads, citing Meta's Help Center, reports that Meta turns enhancements on when it predicts they will help. Two ads in the same account can receive different defaults. A Review defaults tab lists what Meta switched on, grouped into seven categories, but it does not show features you opted into yourself.

Settings come back. JumpFly reports that replacing a creative, changing the format, duplicating an ad or editing it may select recommended enhancements again. Terrads adds that some enhancements can only be turned off in Advanced preview, which is collapsed by default.

So two ads that look identical in your plan may run with different edits. A creative test between them can end up measuring Meta's enhancement choices rather than your concepts.

What do Meta's own numbers show?

Meta publishes averages, and they point upward.

• Product browsing. 3% more conversions, as above, with no cost or revenue data.
• Video generation. In its first quarter 2026 results, as reported by PPC Land, Meta said advertisers using video generation saw a 3% gain in conversion rate in a large test. It also said more than 8 million advertisers use at least one AI creative tool.
• Generated ad text. A 2025 paper on Meta's Text Generation feature for Facebook ads tested a model trained on ad performance against a model trained to imitate curated ads. Across nearly 35,000 advertisers and 640,000 ad variations over 10 weeks, click-through rate rose 6.7%. It did not report conversions, and the comparison was between two models, not against advertisers' own copy.

These are reasonable signals that the features help on average. They are not forecasts for your account. An average lift is compatible with many advertisers losing, and click-through rate says nothing about the cost of a sale.

Which enhancements carry which risk?

Practitioners disagree. Jon Loomer, in a guide last updated in March 2025, is comfortable leaving most enhancements on and sometimes turns off image templates. Digital Ambreen, writing in October 2026, switches off anything that changes the message until its output has been checked on the account. The difference makes sense once you sort the enhancements by what they change.

The fit class mostly adapts a square or 4:5 asset to taller placements. Meta's developer documentation marks placement adaptation as opted in by default, with 4:5 and 9:16 enabled. If you have no 9:16 version, cropping and expansion fill the gap. Check the output in the tallest placements, where cropping causes the most problems.

The message class changes the thing you meant to test. If a winning concept ran with a generated headline or added music, you cannot tell which one won. Meta's documentation also says ads using AI-generated features should be created paused and previewed before going live.

The destination class changes where the conversion happens. That makes your own store data the referee.

How do you test one class at a time?

1. Keep a protected control. Pick a proven ad and turn every enhancement off. Record the settings with a screenshot and, if you use the Marketing API, a read of the ad's creative features, where each one shows as opted in or out.
2. Duplicate it once per class. Turn on one class only. Then reopen the full enhancement menu, including Advanced preview, because duplication can reselect others.
3. Split delivery evenly. Run the control and the variant as a split test with equal budgets and the same audience.
4. Judge on your own numbers. Use orders and revenue from your store or analytics, and cost from Ads Manager. For product browsing especially, Meta-reported conversions may include actions inside Meta's apps that your store never sees.
5. Size the test before you start. A variant needs enough conversions to beat noise. Our guide on how many ad concepts to test sets that budget by target cost per acquisition.
6. Adopt one class, not the bundle. If a class wins, turn on that class across scaled ads and recheck after every edit.

A worked example

Take a hypothetical apparel store running one evergreen Reels ad at about 400 purchases a month. It runs the control against a product browsing variant for three weeks, splitting spend evenly.

Ads Manager reports 212 conversions for the variant against 200 for the control, a 6% rise. The store's own order data shows 196 orders attributed to the variant's campaign against 198 for the control, at nearly the same spend.

The Meta-reported lift does not appear in the store. The likely gap is actions counted inside Meta's in-app experience that never became a site order. The store keeps product browsing off on that ad and runs the same test on a catalogue ad, where in-app browsing has more products to show. Another advertiser could see the opposite result, which is why the test runs per account.

Are TikTok and ChatGPT moving the same way?

Yes, toward automated changes by default.

TikTok's Auto-fix for Smart+ campaigns flags music without commercial rights and suggests replacements, and suggests fixes for video dimensions and resolution. A fix replaces the asset for that ad group, and the original stays in the Creative Library. Smart+ auto-selected creatives, available for app promotion and some lead generation campaigns, can pull creator content and generate or refresh creatives with AI. Choosing custom selection mode turns that off.

OpenAI told ChatGPT advertisers that creative enhancements will switch on for new and existing self-serve campaigns from 30 October 2026, without approval of each version, according to Search Engine Roundtable. Existing opt-outs are kept.

The pattern is the same on each platform. The platform edits the asset after you upload it, and its reports do not split results by edit. Your test design has to make up the difference. For more on what counts as a new creative when automated variants rotate, see our note on creative fatigue, and on Meta's AI animation of stills, our static versus video guide.

What to do before 28 October

• Decide per account whether to opt out of the product browsing default. If unsure, opt out now and test it on one ad.
• Note your current orders, cost per order and the gap between Ads Manager and store data. That baseline makes any shift visible.
• Open each scaled ad and record which enhancements are on, including Advanced preview.
• Set up one control and one product browsing variant on a catalogue ad.

If you want your ads, search presence and AI visibility read from one place, the waitlist is open.

Questions

What changes in Meta ads on 28 October 2026?

According to a notice reported by PPC Land on 15 October, Meta will add product browsing to existing Facebook and Instagram ads from 28 October unless the advertiser opts out. A tap may then open an in-app page with product information and personalised suggestions. Meta had published no Help Center page on the change at the time.

Should I turn off Advantage+ creative enhancements?

Not all of them. Enhancements that crop or expand an asset to fit a placement carry little brand risk and help where you lack a 9:16 version. Enhancements that change your words, audio or imagery, and ones that change where a tap lands, change what you are testing. Turn those off until you have tested them.

Can I see which enhancement improved my ad's performance?

No. Ads Manager reports results at the ad level, not by enhancement, and Meta can predict different defaults for two ads in the same account. The only reliable read is a controlled comparison where one class of enhancement differs and everything else is held the same.

Is there an account-level switch for creative enhancements?

Not for generally available ones. The account setting for testing new creative features covers only features still in testing. Generally available enhancements are set per ad, and some sit in Advanced preview. Editing, duplicating or replacing an asset can select recommended enhancements again, so recheck every new ad.

Sources

1. PPC Land, Meta will turn on product browsing in existing ads on October 28 ppc.land
2. Relevant Audience, Meta will turn on product browsing in existing ads on 28 October 2026 relevantaudience.com
3. Meta for Developers, Advantage+ creative. Get started developers.facebook.com
4. Meta for Developers, Marketing API v22 impacts to Advantage+ creative enhancements developers.facebook.com
5. arXiv, Improving Generative Ad Text on Facebook using Reinforcement Learning arxiv.org
6. TikTok Ads Help Center, How to use Auto-fix to detect and fix creative issues for Smart+ campaigns ads.tiktok.com
7. TikTok Ads Help Center, About auto-selected creatives for Smart+ ads.tiktok.com
8. Search Engine Roundtable, OpenAI Auto Enabling Creative Enhancements In ChatGPT Ads seroundtable.com
9. PPC Land, Meta Q1 2026 revenue as AI tools double advertiser adoption ppc.land
10. Relevant Audience, Meta's Advertising Week 2026 ad tools. What is live, in test or coming relevantaudience.com
11. Terrads, There is no account-level switch that turns Meta's creative enhancements off terrads.com
12. Ads Uploader, Advantage+ Creative Enhancements. The Complete 2026 Guide adsuploader.com
13. JumpFly, Meta's Advantage+ Settings. Why They Keep Reappearing jumpfly.com
14. Jon Loomer Digital, Which Advantage+ Creative Enhancements Should You Turn On? jonloomer.com
15. Digital Ambreen, Advantage+ Creative Enhancements. What to Turn Off digitalambreen.com
16. Leapbuzz, Meta Advantage+ Creative in 2026. What the Default-On AI Enhancements Are Doing to Your Ads leapbuzz.com
17. adlibrary, Advantage+ Creative. Why It's Editing Your Ads adlibrary.com

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →