Web Push Base Monetization for Publishers and Media Owners

A web push base earns most of what it will ever earn in the first thirty days of each subscriber's life and close to nothing past a hundred and eighty. That curve decides the value of a base far more than which network buys it, which is why the first thing we do is not connect demand. We pull your cohort data and price the base as it actually ages. The permission prompt is the asset: Everything downstream is set at the moment the browser asks. Sites with very low acceptance rates get enrolled by Chrome into its quieter permission interface, at which point the opt-in rate collapses and does not recover on its own. We move the prompt behind an intent signal, a second article, a fare search, a finished read, put a plain-language pre-prompt in front of the native one, and accept a smaller base that lives considerably longer. What we will not connect: A base collected behind a fake captcha, an age gate, a click-Allow-to-continue overlay, or a download button. Chrome's abusive notification enforcement exists for exactly these and can remove a site's ability to prompt at all. Separately from the enforcement risk, it is not work we do. If your prompt is one of these, the audit says so and the engagement ends there at the audit fee. Send policy and the number that governs it: Two to three notifications per subscriber per day, with the unsubscribe rate per send as the constraint rather than revenue per send. When unsubscribes cross the threshold we agree at the start, frequency comes down that week. Publishers who reverse that priority have one good quarter and a base worth very little a year later. We have taken over three of those and the repair is slow and partly impossible. Demand sorted by cohort: The freshest cohort goes to direct advertisers at a floor negotiated per geo, because that is the inventory an advertiser will pay a real price for and it is where your reputation is decided. Aging cohorts go to network demand. Dormant cohorts are retired rather than sold into the bottom of the market, where the bidders are the offers that generate the unsubscribes in the first place. Regional specifics we will tell you before signing: Indonesian and Vietnamese bases price low per thousand and hold engagement longer than European ones. Singapore prices higher and churns faster. iOS is largely absent from any base, since Safari has served web push only to sites added to the Home Screen since iOS 16.4, so an Apple-heavy readership will not build a base worth monetising and should be earning elsewhere. Not for: Operators who bought traffic to build the base and need it repaid this quarter. The send frequency that does that is the send frequency that ends it, and we will not be the ones holding the schedule. Also not for anyone wanting a valuation before we have seen the opt-in cohort curve. Every such number is a guess, including ours. What arrives: Revenue by weeks since opt-in, plotted as a curve rather than an average. Revenue per thousand by cohort, geo and operating system. Unsubscribe per send by cohort. The rebuilt prompt with its own before-and-after acceptance rate. A ninety-day projection with its assumptions listed separately so your finance team can argue with them line by line.