In-Page Push Campaign Management for iOS-Heavy Markets

In-page push reaches the audience classic push cannot: iPhone users, and everyone on Chrome who dismissed the permission prompt. In Indonesia, Vietnam and the Philippines that is a large share of the mobile audience, and in travel it is a disproportionate share of booked value. Expect the reported click rate to land two to four times above classic push, and expect to discount most of that difference before you plan anything around it. Why the click rate is a trap: The format is an element drawn onto a publisher page, not a notification anyone subscribed to, so a portion of every click is a thumb meeting a widget that appeared mid-scroll. We treat the network's click column as a billing figure, not a quality signal, and judge sources on what happens next: share of sessions under three seconds, scroll depth, and conversions per thousand clicks. A source whose sessions are seventy percent sub-three-second comes off the buy no matter how good its rate looks in the dashboard. Reconciliation, done daily: We compare clicks the network bills against clicks the tracker records, source by source. A gap above ten percent goes back as a credit request with the log attached rather than being absorbed into blended cost. Some networks in this category settle those without argument. One does not, and we stopped buying from it, which is the kind of thing you will only learn from a desk that has been billed for it. Format, and where the interstitial does not go: The full-screen interstitial converts better than the corner widget and costs more in goodwill. We run it on arbitrage and entertainment inventory and keep it away from anything adjacent to your own audience, at one impression per user per site per day. The corner widget carries the always-on buying. Both are capped at the source level rather than the campaign level, because campaign caps do nothing when one publisher supplies most of the volume. iOS is bought as a separate campaign: Different bid, different landing page, different creative. Safari handles redirects differently, the order values in travel and in donations run higher, and blending iOS into an Android-dominated campaign hides both. In this region that split is usually the difference between a channel that looks marginal and one that pays. What we refuse: Decoy close buttons, back-button hijack, scroll-triggered redirects, and anything that fires before the page has painted. We do not take run-of-network buys where source identifiers are hidden, because a source list you cannot read is a source list you cannot cut. We do not place your brand alongside adult or download-bundle inventory to reach a cost target. Not for: App install campaigns hoping for cheap attributed volume. In-page push installs look reasonable in the network dashboard and thin out badly in retention cohorts once your MMP has thirty days of data, and we would rather say that now than defend it later. Also not for ecommerce with an average order below thirty dollars, where the accidental-click share eats the entire contribution margin. What arrives: A source list you keep, split by format and by operating system, with the post-click quality metric that justified each inclusion. Daily reconciliation logs. Weekly cost per conversion by source group, format, geo and creative. Landing page notes where a source is failing for reasons that are yours rather than the network's, stated plainly.