Meta Starts Measuring Conversions on the Customer's Device on November 12, 2026, and the First Step Is Checking Whether Your Site Qualifies

Meta is adding a Pixel setting that matches an ad view to a purchase on the buyer's phone or computer, and it ships turned on by default. It's called "Measure conversions on customer devices", it lives in Events Manager (the Meta panel where you manage your Pixel and other data sources) and, according to the panel text as reproduced by ppc.land, measurement starts on November 12, 2026, 15 days before Black Friday on November 27. The detail that matters most is in Meta's help page: the setting is on by default for eligible pixels, and to be eligible your site has to require a login.
What changes between the ad and the purchase
What's new is where the match happens. Until now the Pixel, which is Meta's code running in the buyer's browser, and the Conversions API (CAPI), which sends the same events from your store's server, deliver purchases to Meta's servers, and Meta then tries to connect them to someone who saw or clicked an ad, which it can't always do. The example on Meta's page about the setting is a lamp: someone sees the ad on Instagram, buys on the site two days later, and Meta may have a record of the view and receive the purchase without being able to connect the two.
With the setting on, that match happens on the customer's device, using information the person already shared with the site (Meta mentions an email) and temporary device data that doesn't identify them. According to Meta, no identifying customer information is shared: Meta only receives confirmation that a purchase happened and which campaign to attribute it to. The setting runs alongside the Pixel and CAPI without replacing them, so if you already use CAPI, for example the one-click version we covered in this note, you keep it. Meta also states that the attribution window and model don't change: the setting "affects which conversions can be measured, not how they're credited to your ads".
How to tell if your site qualifies
Eligibility comes down to two conditions Meta publishes on its page about how to turn it on. The first is that the Pixel is installed directly on the page, not inside an iframe (another site's page embedded in yours) and not as an image-only pixel. The second is that the site requires people to log in to an account with an email or a mobile phone number.
Checking takes three questions. The first is whether a customer can finish a purchase without logging in. If they can, then by Meta's wording the site doesn't meet the login condition, although Meta doesn't explain how it applies that to a store that offers both options (for example, one with both account login and guest checkout). The second is for whoever installed the tag, especially if a third party hosts the checkout: does the Pixel load directly on the page, or through an iframe or an image pixel? The third is answered in the account: Events Manager, Data sources, your pixel, Settings, where the toggle "Measure conversions on customer devices" appears below "Meta Pixel settings". Meta says you can turn it off at any time.
It's worth looking at the toggle even if your site seems not to qualify, because according to ppc.land the requirements don't appear in the panel or in the notice Meta sent advertisers in mid-October, as reproduced by the sites Wortfilter and ASTIG on October 18, 2026.
The baseline worth having before November 12
For a site that qualifies, the risk is reading a change in what gets measured as a performance improvement. A recovered conversion is a purchase that already happened and that the report now attributes to an ad; whether the ad caused it is a separate question, as ppc.land puts it. Meta hasn't published any figure for how many conversions the setting adds.
What you can build is your own comparison. We recommend a sheet with two columns per day, purchases attributed by Meta and paid orders from your backend or ecommerce platform, covering the weeks before November 12 and continuing after it. If the ratio between the columns jumps on that date with no change in the store, the jump comes from measurement. The backend has its own noise, with cancellations, returns and late-confirmed payments, so it's best to use paid orders and the same criteria before and after.
The idea comes from two practitioners. Yiqi Wu, CEO of Aimerce, a company that sells Pixel tracking for Shopify stores and so has a commercial interest in how advertisers react, proposed on X on October 26, 2026 the same day-by-day comparison over the last month, and keeping CAPI on. MetricMaven, which publishes measurement guides, also suggested running a holdout before crediting Meta with a lift. A holdout is a test in which part of the audience doesn't see the ads, so you can compare how much each group buys.
Leave it on or turn it off, and under what condition
Whether to leave it on or turn it off depends on whether the numbers keep adding up against your orders. Meta gives two reasons to keep it on: seeing more of the conversions your campaigns drive in reporting, and "improve campaign performance with a stronger signal", a phrase that doesn't say whether recovered conversions feed bidding or only the report. Wu wrote that he wouldn't opt out by default, and set the condition for doing so: the numbers stop matching orders and you can't explain why. That's a criterion you can apply with the two-column sheet, even though Meta doesn't clarify the rest.
The decision also has a legal side. Meta writes that it's the advertiser's responsibility to make sure the integration doesn't share prohibited information with Meta. ppc.land adds that on-device processing doesn't settle consent questions by itself, and picks up the ones raised by the firm Captain Compliance, including whether people who declined consent are excluded and what the code reads on the device.
What Meta hasn't said yet
Several questions are still open: whether recovered conversions feed delivery or bidding, how they're deduplicated against events already arriving through the Pixel and CAPI, whether purchase value travels with the confirmation, and whether there will be a trial period (the notice, as reported, describes none). It's also unclear what happens in Safari; as context, and not as a claim about this setting, WebKit announced on October 19, 2026 that Safari 27 will block providers that turn emails into tracking identifiers. The date itself comes from the panel and from reports about the notice, not from the help pages, and Wortfilter describes a gradual rollout starting November 12.
On the market, Meta declares only one exclusion: the setting is not active for people located in the European Union and the United Kingdom. It doesn't publish a country list. That means US users are not excluded, so Black Friday and the rest of the holiday season matter here too, although it doesn't confirm the setting is available in every account. In any specific account, the most direct signal is the toggle in Events Manager.
This content was developed with AI assistance and reviewed by the Zenda team. Any bad ideas are entirely our own.