Generating revenue from push notification traffic.
What named sources recommend, and how they say to do it or measure it, newest first.
Nic Newman of the Reuters Institute recommends that publishers give users ways to vary the number and type of alerts they receive.
From Nic Newman, Reuters Institute for the Study of Journalism, reutersinstitute.politics.ox.ac.uk
Keep in mind: Written about news alerts in general, not about sponsored push. It is the chapter's conclusion, framed as a way to avoid unsubscribes.
Observations and reported facts, newest first.
Apple's App Store Review Guideline 4.5.4 bars promotional push notifications in iOS apps unless users explicitly opted in through consent language in the app.
From Apple, App Store Review Guidelines, developer.apple.com
Keep in mind: Guideline also requires an in-app way to opt out, and says abuse can lead to revocation of push privileges. Covers in-app push on iOS only, not web push. Date is the page's 'Last Updated' date.
Google's Chrome team announced on 2025-10-10 that Chrome automatically removes notification permission from sites that send many notifications and get very low engagement.
From Archit Agarwal, Product Manager, Chrome, Google (Google Ad Manager), blog.google
Keep in mind: Applies to web push in Chrome on Android and desktop, not to installed web apps. Users can re-grant permission or turn the feature off. Google is an In View client.
Google's Chrome team reported on 2025-10-10 that less than 1% of all notifications in Chrome receive any user interaction.
From Archit Agarwal, Product Manager, Chrome, Google (Google Ad Manager), blog.google
Keep in mind: Figure covers all Chrome notifications, not only publisher or ad-bearing push. Google gives no method or sample size. Google is an In View client.
Google's Chrome team reported on 2025-10-10 that, in its tests of permission removal, websites that sent fewer notifications saw an increase in clicks.
From Archit Agarwal, Product Manager, Chrome, Google (Google Ad Manager), blog.google
Keep in mind: Google gives no size for the click increase and no definition of 'lower volume'. Google is an In View client.
The Reuters Institute Digital News Report 2025 found that 43% of people who do not get news alerts say they actively disabled them.
From Nic Newman, Reuters Institute for the Study of Journalism, reutersinstitute.politics.ox.ac.uk
Keep in mind: Survey across eight countries (USA, UK, Germany, Japan, Brazil, India, Kenya, South Africa). It covers news app alerts, not ad-bearing push, but it measures the opt-out behaviour that sponsored push puts at risk.
The Reuters Institute reported in June 2025 that The Times of London caps news alerts at four a day to avoid spikes in app uninstalls.
From Nic Newman, Reuters Institute for the Study of Journalism, reutersinstitute.politics.ox.ac.uk
Keep in mind: Editorial app alerts, not ad-bearing push. The same chapter reports that The Jerusalem Post and CNN Indonesia send up to 50 alerts a day.
Pushly reported on 2025-04-14 that impressions from push notifications at GRV Media, a UK sports publisher, rose by over 100% in 12 months.
From Pushly, pushly.com
Keep in mind: Vendor press release on a contract renewal; no baseline given. The release does not say whether these impressions carry ads.
Jeeng launched AdServe for Push in July 2021 to traffic ads from a publisher's Google Ad Manager inventory into browser push notifications.
From Jeeng (PowerInbox (Jeeng)), Martechvibe
Keep in mind: Launch announcement from 2021; Jeeng's own Jeeng Push page (copyright 2026, read 2026-11-02) still lists AdServe for Push. The release also cites unnamed data that 'nearly 60 per cent' of consumers accept ads in push; dropped as unsourced.
Add https://mcp.beeler.tech/mcp to Claude or ChatGPT as a custom connector, once. Then ask: “What do publishers report about monetize push notification inventory?” It answers from these statements, with every source named and linked.
Every statement links to its source and says when Beeler.Tech has a tie to it, such as an In View client. Vendors are named as vendors. Community observations are anonymous; here is how we use them. Corrections are made in public.
