- Meta announced on June 9, 2026 that it is retiring “Your activity off Meta technologies,” the setting that let a person disconnect off-site activity from their account.
- The replacement, “Activity from other businesses,” governs only how Meta uses that data to personalize. Meta still receives it.
- Users who had disconnected are matchable again, so expect Event Match Quality and retargeting pool sizes to rise.
- Nothing in the change hands you a consent record. What you can produce is still whatever your own form captured or your vendor attached to the lead.
- Re-baseline your match rate before you credit any cost-per-lead improvement to your own work.
Meta announced on June 9, 2026 that it is retiring the setting that let people disconnect their off-site activity from their Meta account. The replacement does not do the same job. It governs personalization, not collection. If you buy leads or calls, that one change moved two numbers in opposite directions at the same moment.
Quick answers:
- What exactly did Meta remove in 2026?
- Does Meta still receive my off-site activity?
- Will my Conversions API match rate go up?
- Does this change what consent record I need?
- Is this different for health and insurance advertisers?
- What should I check first?
What Meta changed on June 9, 2026, and what rolls out from July
Meta is discontinuing “Your activity off Meta technologies,” the control that let a person disconnect off-site activity from their account. Its replacement, “Activity from other businesses,” governs only whether that activity personalizes your content and ads. Meta still receives the data. The change takes effect in the US and other countries from July 2026.
Meta’s June 9 newsroom post says the company is “expanding the ‘Activity from other businesses’ setting, which lets you control how we use this data to personalize your experience.” The changes “will go into effect in the US and a number of other countries next month with more countries to follow.”
Note what the post does not claim. Meta is explicit that “we aren’t collecting any new data as part of this update.” True, and it is the sentence most coverage stopped at. Nothing new is being collected. Something old is being un-hidden.
Two details from the independent coverage matter to a buyer, and Meta’s own post buries both. Writing the same day, The Hacker News quoted Meta widening where the data lands: “In the future, we’ll use this information to personalize other parts of your experience, including the content you see in your Feed and AI responses.” The same report named the first markets: “The change is expected to go into effect in the U.S. and a number of other countries, including the U.K., Brazil, Thailand, South Africa, Turkey, South Korea, Ecuador, Nigeria, and Kenya, starting next month.” So the event your Pixel fires now shapes a feed and an assistant answer, not just an ad auction, and it does so on a market-by-market schedule you should check before you compare two countries’ numbers.
Meta published no calendar day for when the old control leaves accounts. It said “next month.” Treat July 2026 as the window and confirm from your own account, not from a blog post that invented a date.
The old control disconnected you. The new one only changes what you see.
These two settings are not versions of each other. They act at different points in the pipeline, and that difference is the whole story.
Meta described the old mechanism plainly when it launched off-Facebook activity controls in 2019: “If you clear your off-Facebook activity, we’ll remove your identifying information from the data that apps and websites choose to send us.” It added that “we won’t use any of the data you disconnect to target ads to you on Facebook, Instagram or Messenger.” That is a severed link. The event still fired. It just stopped being about a person.
The replacement acts later. Meta’s help page for the updated setting is direct about it: “Regardless of your choice, businesses and organizations may send us your activity.” Turning it off changes what you are shown. It does not stop the Pixel event, the Conversions API payload, or the offline upload from arriving and resolving to an account.
So the control moved downstream. It used to sit between your event and Meta’s graph. Now it sits between Meta’s graph and the user’s feed. Everything upstream of the feed is unchanged. That includes every part of the pipeline an advertiser actually touches.
Your match rate and your retargeting pool just got bigger
Here is the advertiser side, without the cheerleading.
People who had disconnected were functionally missing from your matchable pool. They browsed and they converted. The event landed with the identifying information stripped out. Retire that control and those users return to the pool. That is not a forecast. It is the mechanical result of removing the thing that was removing them.
You will see it first in Event Match Quality. Meta defines EMQ as “a score (out of 10) that indicates how effective the customer information sent from your server may be at matching event instances to a Meta account.” The score is calculated partly from “the percent of event instances that are matched to a Meta account.” That denominator just changed underneath you. Meta’s docs add that “high quality event matching may improve ads attribution and performance.”
Website custom audiences should follow. A retargeting pool is a count of matched people. A pool that had been quietly shrinking for six years now has room to refill.
For a sense of scale, use the source with no reason to inflate it. Announcing the disconnect control in 2019, Meta wrote: “We expect this could have some impact on our business, but we believe giving people control over their data is more important.” A platform does not warn investors about a feature that costs it nothing. Retiring that feature reverses whatever the cost was.
A better-matched lead is not a better-consented lead
Now the part nobody else on this search result page has written, because nobody on it buys the leads.
Match quality is a statement about identity resolution. It says Meta is more confident this event belongs to this account. It says nothing about whether the person agreed to be contacted, on what terms, or with a record you can hand to somebody else. Those were always separate systems. One of them just improved and the other did not, which makes it very easy to feel like both did.
Price the difference, because it has one. A bigger matched pool pulls your reported cost per lead down. It does not pull down the cost of the records you later write off, and those two figures live in different columns. If your July cost per lead falls and your contact rate falls with it, you did not buy cheaper leads. You bought more of the same leads at a better exchange rate.
The record is the half that did not move. Meta never handed you a consent artifact and was never going to. What you can produce is whatever your own form captured, or whatever your vendor attached to the record when it sold it to you. That is a commercial question, settled in a purchase agreement, and it is now the only side of this you still control.
So put it in the contract, which is what serious buyers already do. The trade publication Lead Gen Economy describes the practice in one line: “Many buyers require TrustedForm certificates as a condition of purchase.” It states the consequence just as plainly: “Leads lacking consent verification should be returned or rejected.” That is a delivery condition with a return right attached, not a policy statement. It is the same clause structure you would write for any input you buy and cannot inspect on arrival.
Then read the shelf life, because the artifact expires and almost nobody checks. The same source notes that TrustedForm “certificates are deleted after 90 days unless retained,” with longer storage sold as a separate product. A record you paid for and did not retain is a record you do not have. Ninety days is shorter than most lead-buying relationships and shorter than the lag on most disputes, so retention is a line item somebody has to own by name. Our comparison of the two certificate tokens walks through which one fits which buying model.
Three clauses to write into the next vendor agreement, in the order they save money:
- Every delivered lead arrives with its consent artifact, and a lead that arrives without one is returnable at no cost to you.
- The artifact is retained for a stated period, and the agreement names who pays for that retention.
- You can pull and inspect a sample of certificates on your own schedule, rather than requesting them the week you need them.
Run the two ledgers side by side. Your Meta ledger this quarter: more matched events, a fuller audience, probably a better reported cost per lead. Your record ledger this quarter: exactly what it was in May. The first is graded by the platform that benefits when the number goes up. The second is graded by whoever buys the call from you, on the day they ask to see it.
Health, insurance, and finance are where this bites
Restricted categories are where a thin paper trail turns into a returned batch. It is worth being precise about which ones.
In health insurance the split matters. U65 is off-exchange private coverage. ACA is on-exchange marketplace coverage. Different products, different enrollment rules, different buyers. Lumping them together is how operators end up dialing a record that could not enroll that week. These are also the verticals where the buyer at the end of the chain asks to see the artifact before paying, so a missing one stops being paperwork and becomes a credit note.
Watch for a good quarter that arrives without a reason. A campaign whose cost per lead improves in August 2026, in a restricted vertical, with no creative change and no bid change, is a campaign whose input pool changed. That is fine. It is not something to scale into until you know it was the pool and not your work.
What to do in the next 30 days
Three things, in order.
Re-baseline before you credit anything. Pull your Event Match Quality and matched audience sizes for June 2026 and freeze them as a reference. If those numbers rise from July onward, the rise is Meta’s, not yours. Compare September against the July baseline. Read it against June and you will mistake a platform change for a campaign win, then scale into it.
Re-read your own consent language against what now actually happens. If your privacy policy or form disclosure describes a user’s ability to disconnect off-site activity from their Meta account, that description is now wrong. It is the cheapest thing on this list to fix and the most annoying to explain later.
Move the check to the conversion event, because that is the only place left where you control it. Platform match quality never verified anything about the person’s intent and was never going to. Verification at the point of conversion does: one-time-password confirmation that the number belongs to the person, bot and form filtering before the record enters your system, and a timestamped artifact that is still there on the day somebody asks for it. Buying from third parties, the same check belongs before the bid, not after the post.
There is an honest way to read this change. For six years, a user who opted out was quietly doing a piece of filtering on your behalf. You never chose it, never paid for it, never saw it on a dashboard. It filtered anyway. That filter is gone now, and Meta replaced it with nothing, because filtering for you was never Meta’s job. You did not lose a control. You inherited a cost line. The only question is whether you find it on this quarter’s return rate or next year’s.
Worth reading next: how Meta’s broad targeting and optimization signals behave for lead gen, and what Global Privacy Control means for a lead buyer.
Frequently Asked Questions
What exactly did Meta remove in 2026?
Meta is discontinuing “Your activity off Meta technologies,” the setting that let a person disconnect activity businesses sent to Meta from their account. Meta announced the change on June 9, 2026 and said it takes effect in the US and other countries the following month.
Does Meta still receive my off-site activity?
Yes. Meta’s help page for the replacement setting says that “regardless of your choice, businesses and organizations may send us your activity.” The new control governs only whether that activity personalizes the ads and content you see. It does not stop the data arriving.
Will my Conversions API match rate go up?
Probably, in markets where the change has rolled out. Event Match Quality is scored partly on the percentage of event instances matched to a Meta account, and users who had disconnected were unmatchable. Removing the control returns them to the pool. Measure it against a June 2026 baseline rather than assuming it.
Does this change what consent record I need?
No. Nothing here produces a record for you. Meta improved how confidently it matches an event to an account. It did not start handing advertisers an artifact showing that a person asked to be contacted. That still comes from your own form or from your lead vendor, which makes it a contract term rather than a platform setting.
Is this different for health and insurance advertisers?
The mechanics are the same. The consequences are not. Health, insurance and finance are the verticals where the buyer at the end of the chain actually asks to see the record before paying. A better-matched, worse-documented lead is likelier to come back as a return there than almost anywhere else.
What should I check first?
Your baselines. Record Event Match Quality and matched audience sizes as they stood in June 2026, before the rollout, so you can separate a platform effect from a campaign effect. Then check whether your consent disclosures still describe a control that no longer exists.
