Google shipped an age signal your marketing team may not touch

Google opened the gate, then fenced app marketing out#

A fox at a gate does not pause from fear. It pauses because the scent on the latch is unfamiliar, and a careless crossing costs more than a patient prowl. Google Play hung a new gate on 29 July 2026, and app marketing is named on the notice that says keep out.

Paul Feng, VP of Product Management at Google Play, used the Android Developers Blog to announce the expansion of the Google Play Age Signals API to all Play developers globally. Google calls it a privacy-preserving tool that puts parents in the driver’s seat, letting them share a child’s age range, 16-17 for instance, directly with apps, and letting adults share their age when prompted. Controls sit inside Google Family Link, where age ranges are never shared by default and parents can update or turn them off at any time.

Two clocks are running, and confusing them wrecks a quarter of planning. Developer availability is global today. User availability is staged, and staged is not settled: building on current availability in Brazil, the Android Developers Blog says the experience reaches users in Australia and Canada by mid-August, with a full global rollout later this year. An app marketing forecast that assumes every user can hand over an age range this month is forecasting for a forest that has not grown.

The sentence that changes the work#

The Play Age Signals documentation states that you may only use information from the Play Age Signals API to provide age-appropriate content and experiences in compliance with laws, and that you may not use it for any other purpose including, but not limited to, advertising, marketing, user profiling, or analytics. The sanction is no gentle slap: misuse may result in termination of your API access and suspension or takedown of your apps from Google Play. The launch post sets no deadline and no mandate, and policy timings, when they land, are posted in Google Play policy announcements.

A rich new attribute lands on every app in the store, and the app marketing team is fenced out by name. Not an oversight, not a loophole to be lawyered open. It is the design, drawn deliberately.

Age is a safety input, never an audience segment#

The reflex in most growth teams is to reach straight for the segment builder. Resist it. Age here is a safety primitive, not an audience primitive, and that single distinction shapes the whole app marketing response.

Google is explicit about proportionality. Rather than enforcing one-size-fits-all rules, the Android Developers Blog says, developers get the flexibility to choose how they integrate safety signals, and it offers a homely example: a weather app should not need the same safety settings as entertainment or media apps. With this reliable signal, the post adds, you retain complete agency to tailor content, features and settings to match your audience. Read blog and terms together and the boundary is obvious: shape the experience, never the targeting.

The permitted moves are product moves: set a safer default, gate a feature, trim what a younger account can discover. The prohibited ones read like a media plan. No lookalike, no retargeting pool, no cohort split in reporting, no slide dressed up as an app marketing strategy insight. Careful teams still trip on analytics: pushing an age band into a product analytics tool is profiling, and profiling is named in the prohibition.

Two details deserve a place in the den. The Play Age Signals documentation records that the API sits in beta and that its terms of service were last modified in October 2025, so the ground can shift under you. It also states that no data is collected by the Play Age Signals client library, and that developers are responsible for ensuring their use complies with all applicable regulations and Google Play policies. A signal that is not yours to burrow away.

Teams who have watched their audience data thin will recognise the shape from our note on first-party data decay: attributes keep arriving, permissions keep narrowing. Move the value up into brand strategy and product craft, where no terms of service can fence you out.

Three answers at the door, and what each one costs#

The engineering shape decides what your app marketing can honestly promise. SDK 0.0.4 introduces a two-function architecture, per Google’s request age signals guide. Your app calls requestAgeSignalsAccess with an Activity and reads the ageSignalsStatus handed back, which arrives as one of exactly three values.

That third status is the quiet correction to the user-choice framing. Google’s request age signals guide says that for eligible users in US states whose laws require app stores to provide verified age information, the in-app prompt is not triggered; those users verify or set up supervision when they visit the Play Store app. In other regions the user picks from Ask before sharing, Always Share or Never Share, so Never Share is an answer the API is built to return.

The dates stamped on the Play Age Signals documentation show the direction of travel. From 17 March 2026 the API started returning age signals for users in Brazil, for requirements under Digital ECA, and it has started returning signals for eligible Texas users who created accounts after 28 May 2026, as compliance work for Texas SB2420. Age verification law is arriving in patches, and each patch changes what your app must do at its own front door. Play Console already ships Restrict Minor Access, which the Android Developers Blog describes as helping developers manage who can discover their apps.

Design for the NOT_SHARED path first. A mobile app marketing plan that only works once the signal arrives has a hole through the middle, and it gapes widest wherever the user-side rollout has not landed.

What age assurance looks like at full scale#

To see age assurance at real scale, follow Roblox, which has run this in public and published the counts itself. It is the best-lit trail an app marketing team has, with one caveat running through every figure below: these are the company’s own uncontrolled claims, not independent measurements.

The Roblox Q2 2026 shareholder letter, filed on 30 July 2026, reports that by the end of Q2, 57% of global daily active users have age-checked, led by strong execution in developed countries. The United States and Australia surpassed 70% penetration, against a stated long-term goal of 90%. Among under-18 users, penetration runs at 60% globally and 75% in the United States. Proactive age checks for communications went global in January, with Roblox Kids and Select age-based accounts in June. In APAC, modified Verified Parental Consent requirements moved the numbers, and Indonesia saw a double digit percentage increase in Q2.

The part that gets misquoted#

Roblox reported Q2 revenue of $1,469 million, up 36% year over year, and bookings of $1,557 million, up 8%, growth the company said landed at the low end of its guidance range. Roblox did not blame age checks for that shortfall. It pointed to a decline in per hour monetisation, most notably with younger cohorts in the United States and Canada, caused by a greater than expected shift of engagement from high monetising 2025-vintage viral games to both new and evergreen games with lower hourly monetisation, compounded by changes in its Recommended for You algorithm and, to a lesser degree, by disabling the sale of cross-experience game passes.

That is the quarry worth tracking. A platform pushed age assurance across the majority of its users, watched a monetisation wobble arrive in the same quarter, and pinned that wobble on engagement mix and recommendations rather than on the age gate. Read it as one large public dataset declining to blame age checks, not as proof they are free. Our app retention playbook holds under these conditions, and so does the arithmetic in our piece on the store fee squeeze.

Self-declared ratings and the app marketing exposure#

Tempted to treat store age ratings as a proxy for the signal you may not touch? Look at what a rating is worth.

A study by Deepsee.io, dated 15 July 2026 and covered by PPC Land on 3 August 2026, analysed 4,346 AI companion apps and found 2,589, or 60%, carrying ratings that leave them accessible to minors. Of the 306 apps listed on both stores, 106 are rated for minors on Android but 17+ on iOS.

Two caveats belong in the same breath. Ratings are self-declared, so the count measures declarations rather than content, and Deepsee sells brand-suitability screening, so the finding serves the finder. Take the direction, not the decimal, and carry both caveats into any app marketing deck reusing these numbers.

Distribution can vanish while you sleep#

The second exposure is the store. MacRumors reported that Apple removed Telegram from the App Store on the night of 3 to 4 August 2026 for roughly 40 minutes, reinstating it once the offending content was removed. Telegram’s founder says a takedown extortionist planted illegal material in public groups and reported it to Apple. That account is Telegram’s own and Apple has made no public statement, so treat the mechanism as an allegation, not a finding. The lesson survives either version: a channel you do not own can close at twilight on another’s judgement.

The stakes keep climbing. In its own results announcement, Apple Newsroom reported revenue of $109.4 billion for the quarter ended 27 June 2026, up 16% year over year, with Services setting a June-quarter record at $30.7 billion; Apple does not break out App Store or advertising revenue, so Services is a blended line, not store economics. Meanwhile the European Commission confirmed that AI Act Article 50 transparency obligations apply from 2 August 2026, with penalties up to €15 million or 3% of global annual turnover. The regulatory floor is rising everywhere.

Regulated categories feel this first and hardest. Anyone stalking growth in healthcare marketing already works inside the constraints we set out on app privacy. With the age attribute off the table, app marketing growth comes from craft: sharper paid social creative, search and AI visibility that earns the click before the install, and audiences defined by behaviour rather than birth year, the argument we made about the older cohorts most marketers ignore.

Frequently asked questions#