- One vendor’s numbers are public: at Aditude, eCPMs fell 10.5% among readers on the update, Digiday reports. [1]
- Safari on iPhone and iPad was 11.4% of that network’s ad revenue, and about 8% of users had updated. [1]
- Multiplied, those three give about 0.1% of total ad revenue today. Aditude’s own estimate of about 1% matches full adoption better. That arithmetic is ours.
- The WebKit bug filed by The Trade Desk was marked resolved on October 9, after a fix appeared in iOS 27.2 beta 3. [2][3]
- People on the released iOS are still blocked, and the fix is not expected before the end of October. [1]
- A test of more than 70 ad tech domains seen by Digiday found only the first handful blocked. [4]
- Apple has published no iOS 27 usage yet. Its last count has iOS 26 on 79% of all iPhones. [5]
- Publishers are testing server-side routes such as IAB Tech Lab’s Trusted Server. [6]
A week ago we reported that Safari on iOS 27 was refusing requests to a list of nine domains and that nobody had published what it cost. Both halves of that sentence have moved.
The bug The Trade Desk filed with WebKit is closed. On October 5 a WebKit engineer wrote, “A new iOS 27.2 beta went out today. Please test with it.” The Trade Desk’s Ian Meyers confirmed seeing the changes in build 24B5099f, and on October 9 the ticket was marked resolved. [2] Apple’s developer site lists that build as iOS 27.2 beta 3, released October 5. [3] And a first number has appeared, from a company that sells ad space for publishers. [1]
Neither development tells you what your own site is losing. That takes three numbers, and you already hold two of them.
What is fixed, and what is not
The fix is narrow. The ticket concerned one domain, adsrvr.org, which The Trade Desk uses to bid on and deliver ads. [2] It says nothing about the identity domains that were on the same list. And it lives in a beta. Digiday reports that anyone on the released version of iOS is still blocked and that the fix is not expected to reach them until the end of the month at the earliest. [1]
The wider fear has not materialized yet. Digiday says it has seen results from a test of more than 70 ad tech domains that found only the first handful blocked, with rival identity products from LiveIntent, Lotame, and Yahoo getting through, and that the code letting Apple change the list remotely has not shipped. [4] That is one test, relayed by one outlet. It is also the only evidence anyone has put on the table about the size of the list.
The reach is wider than Safari in one respect: every browser on iPhone and iPad runs on Apple’s engine, so Chrome and Firefox on those devices behave the same way. [4]
The one measured number
Digiday’s October 12 report carries figures from Aditude. [1]
Treat them as one company’s traffic. The report gives no publisher count, no dates, and no description of how the 10.5% was measured, and we could not find the figures published by Aditude itself.
It also contains a sentence that does not add up as written. Aditude, Digiday says, “estimates the overall revenue hit at about 1% at this point.” [1] Multiply the three published factors and you get something else: 11.4% of revenue, times 8% updated, times a 10.5% fall, is 0.096% of total ad revenue. About a tenth of one percent. The 1% only appears if you drop the adoption factor (Aditude’s 11.4% times its 10.5% is 1.2%, the loss once everyone has updated) or if Aditude’s 8% is a share of all readers, not of Safari readers (8% times 10.5% is 0.84%). The report does not say which. The arithmetic here is ours, and it matters, because the two readings are ten times apart.
Three numbers, multiplied
Share of total ad revenue lost to the block, today
The defaults are the Aditude figures Digiday reported. Replace them with your own. The middle field is the one to update every week.
Model: Safari revenue share × updated share × eCPM gap. It assumes the gap is the same across your inventory and that nothing else differs between the two groups, which is why step three below compares them in the same week.
Safari’s share of your ad revenue
Pull revenue by browser and device from your ad server or wrapper analytics for the last full week. Count every browser on iPhone and iPad, not only Safari, since they share the engine. [4]
The share that has updated
Split those sessions by browser major version, 27 and up against everything older. If your ad reporting cannot do it, pass the version to the ad server as a key-value from the page.
Changes weeklyThe eCPM gap
Compare eCPM for the two groups over the same days, on the same page types. Do not compare this week with a week in August. The fourth quarter lifts prices on its own and would hide the gap.
Look at who is missing
Break the updated group down by bidder. A buyer whose win rate collapses only there is one that cannot load. That tells you whether the gap is one vendor or several.
What moves each number
The first number is yours and changes slowly. The third depends on your demand. A site that sells mostly direct, or through an exchange that does not rely on the blocked domains, will see a smaller gap. Index Exchange’s chief executive told Digiday the company has been largely unaffected because it is not in the identity business, and two news publishers with subscription models told it they had seen no impact. [1] If The Trade Desk is a large buyer of your Safari inventory, part of your gap should close when the fix ships.
The second number is the one that grows without you. Apple has not published usage figures for iOS 27. Its most recent count, taken on June 7, 2026, shows last year’s release on 79% of all iPhones and 86% of those introduced in the last four years. [5] Using Aditude’s other two figures, here is what that curve does.
| Share of Safari sessions updated | Share of total ad revenue lost |
|---|---|
| 8%, as reported | 0.10% |
| 25% | 0.30% |
| 50% | 0.60% |
| 79%, where iOS 26 stood in June | 0.95% |
| 100% | 1.20% |
That table is our projection from Aditude’s reported numbers, with the eCPM gap held at 10.5%: 0.95% at 79% updated, 1.20% at 100%. [1] The gap will not hold still. It narrows if blocked vendors are restored or rerouted, and widens if the list grows. What the table shows is the shape: the loss today is small because few people have updated, not because the effect is small.
What to do with the answer
If your number is small, say under half a percent, the right response this month is measurement. Set up the weekly split, write the three figures down, and watch the middle one.
If it is larger, or heading there, the option publishers are testing is to stop making the calls from the reader’s device. Digiday reports a rush of interest in IAB Tech Lab’s Trusted Server, which moves the ad process onto servers the publisher controls, and that Paradium, the owner of Men’s Journal and Parade, is already running one blocked vendor through it. [6] Its head of revenue and data was candid about the cost: “Viewability, verification, and getting buyers to accept our numbers are where most of the work remains.” [6] That is the same trade we described for server-side tagging: you recover signal and take on the job of proving it.
AdExchanger’s editors put the difference from Apple’s earlier privacy changes in one line: this one “effectively blocks some companies from conducting parts of their business entirely.” [7] For a publisher that turns a strategic argument into an operational one. You do not need a view on Apple’s motives to run the calculation. You need three numbers, checked every week, and a threshold above which you act.
Sources
- Digiday · Ad industry tallies early losses from Apple’s Safari tracking blockTrade press, October 12, 2026. Source for Aditude’s figures, which we could not find published by Aditude. No method or sample given.
- WebKit Bugzilla · Bug 324771: Inclusion of adsrvr.org in IS_REQUEST_UNCONDITIONALLY_BLOCKABLE domain listPrimary. Marked resolved October 9, 2026; read October 13.
- Apple Developer · ReleasesPrimary. Lists iOS 27.2 beta 3 (24B5099f), October 5, 2026.
- Digiday · Apple’s secret blocklist is the talk of Advertising Week New York (mostly off the record)Trade press, October 7, 2026. The domain test it cites is not published.
- Apple Developer · App Store support: iOS and iPadOS usagePrimary. Measured June 7, 2026; no iOS 27 figure published when read on October 13.
- Digiday · Apple’s Safari ad tech blocklist sends publishers back to IAB Tech Lab’s Trusted ServerTrade press, October 8, 2026.
- AdExchanger · The Apple Bites BackTrade press, October 9, 2026.
Frequently asked questions
How much revenue has Safari’s ad tech block cost publishers?
There is one published figure so far. Digiday reported on October 12, 2026 that Aditude, which sells ad space for publishers, saw eCPMs fall 10.5% among readers who had installed the update. Safari on iPhones and iPads was 11.7% of the ads its customers show and 11.4% of their ad revenue, and about 8% of users had updated. Aditude estimated the overall hit at about 1%. The report gives no sample size, dates, or method.
Is The Trade Desk still blocked in Safari?
On the released version of iOS 27, yes, according to Digiday. The WebKit bug The Trade Desk filed was marked resolved on October 9, 2026, after its engineer confirmed seeing the changes in iOS 27.2 beta 3, which Apple released to developers on October 5. Digiday reports the fix is not expected to reach users before the end of October at the earliest.
Were the identity vendors unblocked too?
Nothing public says so. The resolved bug concerned adsrvr.org, The Trade Desk’s ad request and delivery domain. It does not mention the other domains on the list, which belong to Unified ID 2.0, ID5, LiveRamp, Permutive, and Audigent.
How do I calculate my own exposure?
Multiply three numbers: the share of your ad revenue that comes from Safari on iPhone and iPad, the share of those sessions that are on Safari 27 or later, and the eCPM gap between updated and not-yet-updated Safari sessions in the same week. The result is the share of total ad revenue currently lost. Recompute it weekly, because the middle number rises as people update.
How fast will iOS 27 adoption grow?
Apple has not published usage figures for iOS 27. Its most recent count, taken on June 7, 2026, shows the previous version, iOS 26, on 79% of all iPhones and 86% of iPhones introduced in the last four years. That is a reasonable ceiling to plan against, reached over months, not weeks.
Does the block affect Chrome on an iPhone?
Yes, according to Digiday. Apple requires every browser on iPhone and iPad to use its WebKit engine, so the same blocking applies in Chrome, Firefox, and other browsers on those devices.
Is a server-side setup a way around it?
It is what publishers are testing. Digiday reports that Paradium, which owns Men’s Journal and Parade, is running Permutive through IAB Tech Lab’s Trusted Server, which moves the ad process from the reader’s device to servers the publisher controls. Its head of revenue and data says viewability, verification, and getting buyers to accept the publisher’s numbers are where most of the work remains.
Mikołaj Salecki
Writes about media, tech, and AI business for people who actually run digital. Former agency lead. Skeptic of frameworks that read better than they perform.
