How do I opt out of Meta product browsing before October 28?

There are two routes. The first is the «Opt out» button inside the notice Meta sent to ad accounts, which has to be used before October 28, 2026; the notice does not state whether it covers one account or every account you manage, so do not rely on it alone. The second is the per-ad toggle: in Ads Manager, open the ad, go to Advantage+ creative, and switch off «Add product browsing» under Destination (Personalized destinations). You can change it at any time, including after October 28, but any spend between the default flip and your adjustment has already run through the new experience.

Does product browsing change my creative or my landing page?

It does not change your creative, but it changes the first screen after the tap. Instead of opening your site directly, Meta may show a product page inside Facebook or Instagram built from your catalog with personalized suggestions, from which the person can still navigate to your web. That extra step moves the start of your landing page views, your GA4 sessions, your consent banner and any post-click test.

Should I keep it on because Meta reports 3% more conversions?

That 3% is Meta's own figure: an average from its August 6 to 21, 2026 experiment across roughly 117,700 advertisers in 3+ verticals (e-commerce, CPG and retail), with no definition of what a conversion is, no published comparison design and no statement of what it costs. Conversions are a count, not a margin: if the suggestion carousel pushes buyers toward lower-margin products, purchases rise while the money falls. Leave it on only where your catalog is strong and your P&L, not Ads Manager, makes the call.

Meta turns on product browsing in your live ads on October 28: opt out or test it on purpose

On October 28, 2026, Meta moves product browsing from opt-in to on-by-default for Facebook and Instagram ads that are already running. On a tap, the person may see a product page inside the app first — your catalog data plus personalized suggestions — instead of landing directly on your landing page. You have two ways to refuse it: the «Opt out» button in Meta's notice before October 28, or the per-ad toggle under Advantage+ creative → Destination. Our position at Mintec is simple: opt out by default when the landing page carries the sale, and turn it on deliberately when a healthy catalog and real margin are what decide. Meta backs the change with 3% more conversions; that number is a count, not a gain, and that difference is the whole article.

What changes on October 28 — and what does not

The notice circulating among media buyers since mid-October is short: "Starting October 28, 2026, we'll add product browsing to your existing ads," followed by the performance claim in bold and a single blue «Opt out» button. Two sentences below describe the mechanism: when people tap your ad, they may see an experience on Facebook or Instagram with product info and personalized suggestions.

Three details separate this from an ordinary feature launch:

The feature is not new; the default is. Product browsing already existed as an Advantage+ creative enhancement, inside Personalized destinations. What October 28 changes is that ads already running move from off to on, and anyone who wants the old behavior has to act.
The extra step is the product. The tap does not land on your site first: it lands on a Meta-hosted product page, a scrollable row of images, prices and descriptions pulled from your catalog, plus suggestions Meta picks. Your web is still reachable, but it is no longer the first screen.
Meta has published nothing official. As of October 15-16 there was no Help Center article or newsroom post; the only official trace is the help page on ad destinations, which lists product browsing among personalized destinations for the Sales objective. The date comes from the notice, not the documentation.

And what the notice does not say matters as much: which formats and objectives are in scope; whether ads created after the 28th also default to on; whether the «Opt out» button covers one account or all of them; and whether boosted posts are affected. With so many answers unpublished, the safe operation is an audit, not trust.

Read the 3% before you quote it

The number Meta wants you to remember: 3% more conversions, on average, with in-app product browsing experiences. The footnote backs it with an experiment run August 6 to 21, 2026 across roughly 117,700 advertisers globally, spanning 3+ verticals (including e-commerce, CPG and retail), at greater than 99% confidence.

Three readings we take before a client repeats it:

It is a count, not a margin. Meta does not define what a conversion is, where it is counted, or what it costs. If the suggestions move a buyer to an entry-level product instead of your hero product, the order still counts, the average order value drops, and the campaign looks better in Ads Manager while getting worse in the P&L.
It is an average across catalog verticals. E-commerce, CPG and retail. If you sell services, generate leads or close on WhatsApp, that average does not describe you — there is no way to know whether accounts like yours landed above or below it.
Nobody has replicated the comparison design. A confidence interval over 117,700 advertisers says nothing about what happened to the accounts that got worse. The average is a claim about the market, not a promise about your account.

Meta's commercial logic is direct: every minute the shopper spends inside the app is a minute not spent on someone else's website. Which is why the real question is not "does it work?" but "does it work for whom, and who pays for the difference?"

The cost nobody priced in: your measurement

Here the change stops being a preference and becomes an instrumentation problem. The landing page is not only where the sale happens: it is where your analytics, your consent banner, your promotions, your post-click tests and your session baseline live. Putting a Meta layer in front of that click changes what all of those tools see.

Your landing page views and GA4 sessions drop even if conversions do not. If the tap no longer opens your site directly, fewer sessions arrive, and the ones that do arrive with product information already seen. Without a baseline taken now, in November you will not know whether you lost traffic or changed definition — which is exactly the problem we covered in why Meta and GA4 numbers do not match.
Landing page view optimization changes meaning. An ad set optimized for that event was bidding for a screen that now sits one tap further away.
Landing page tests lose sample — and lose it from a different visitor. Fewer sessions reach the page, and the ones that do come from people who already saw prices and suggestions before arriving. If a test is mid-flight in October, timestamp it or close it before the 28th.
Your consent banner fires late. If the first screen is Meta's, your site's cookies and consent trigger after the interaction, not before it.

We already lived a version of this in March 2026, when Meta redefined click-through attribution. Across the accounts we manage we saw 18% to 25% drops in reported conversions while CRM revenue stayed flat: the number moved, the sales did not. The lesson is not that Meta lies — it is that when the measurement definition changes, the first thing the report tells you is wrong. Product browsing is the same family of change; this time it affects the traffic that arrives rather than the column that counts it.

There is a second front you cannot audit today: a media buyer documented on LinkedIn that he discovered the traffic on an ad he thought was working was in fact a collection unit on IG Reels, not the creative, and that Meta did not separate the two in reporting. His request to Meta was reasonable: a column stating how much product browsing contributes to that ad. Until that exists, a blind November test leaves you unable to separate creative from catalog.

The decision, by account type

There is no universal verdict, because what changes depends on where your sale is carried. This is the table we are applying with clients this week:

Account type	Our call	Why
Landing page carries the sale: single product, quiz, advertorial, pre-sell, bundle builder	Opt out	An extra step in front of the converting page dilutes what the quality of that landing page already buys you in the auction
Landing page tests in flight, or time-boxed promotions	Opt out until the test closes	You lose sample and change the visitor mid-experiment; the result stops being readable
Lead gen, services, WhatsApp funnels, own form	Opt out by default and check ad by ad	Meta's experiment verticals were e-commerce, CPG and retail; your case is not inside the average — and if the business runs on leads, the real cost of a qualified lead is where the decision should start
Large catalog, healthy feed, cold discovery traffic, slow mobile site	Test on purpose, in an isolated campaign	Here the product layer can genuinely help: less loading friction and more assortment visible
Accounts with pixel + CAPI reconciled against CRM	Test, with margin as the judge	You can see whether the extra purchases arrive with a healthy ticket or only inflate the count
Retargeting and brand campaigns with a strong LP	Opt out	Retargeting already has the signal; the layer only adds a competing explanation between your ad and your page

The question that summarizes the table: does your landing page need to be the first screen after the tap? If yes, opt out. If no, and you have a catalog, test it with method.

What we are doing this week across client accounts

Audit the notices, account by account. Open the notice, confirm which account it covers, and hit «Opt out» before October 28. Because Meta does not declare the button's scope, the per-ad toggle is the only certainty: walk the live ads, starting with the ones producing sales or leads.
Baseline today, not in November. Export GA4 sessions by landing page and Meta landing page views for the last four weeks. Without that number, the expected drop on October 28 will be read as a performance problem when it is a measurement problem.
Decide per campaign, not per account. Apply the table above campaign by campaign; a single account usually contains ads that should opt out and ads that should be tested.
Post-change review. Reopen ads on October 29 to confirm the real state, and repeat that for any ad created, duplicated or pushed through the API after the date.
If you test, test on margin. One isolated campaign, 7 to 10 days, a healthy-margin product set locked in, and the decision made on contribution after ad spend, not on purchases or ROAS.

Why the answer is easier in Latin America

We run accounts in LatAm where the real destination of the sale is not a product page: it is a WhatsApp conversation, a catalog by DM, a short form. In that model the landing — or the chat — is the ad, and if Meta puts a storefront in front of it, you are paying for a step that does not merely fail to convert but moves the interested user away from the place where your team answers. For those ads the call is opt out, with no intermediate test.

Second nuance: Meta's experiment publishes no results by country, so do not assume a global average describes your market. The decision is taken on your account numbers, not on the press-release figure.

The position in one line

Product browsing is neither a trap nor a gift: it is Meta betting that its storefront converts better than your page, and that bet is paid for with your measurement and your margin. Opt out on October 28 if your landing page carries the sale; let it run only where the catalog and the P&L can defend it with your own data. And decide before the 28th: anything that runs between the default flip and your adjustment has already been decided for you.