Engineering posts on in-app purchases and subscriptions
This index collects hands-on engineering writing about subscriptions, paywalls, and store platforms. Use it to find a focused tutorial or technical discussion, then follow the linked article for the full detail.
Start with changes that affect your setup
Start with the most recent post when you need to know whether a store-platform change requires action on your side. The leading item, by Perttu Lähteenlahti on November 13, 2026, explains a move to newer App Store Connect APIs for product-editing tools and states that no SDK update or configuration change is needed. That distinction matters: a backend migration can be significant for the platform team while requiring nothing from app developers.
Choose by platform and problem
For subscription structure, compare the piece on Apple bundles and suites, by Perttu Lähteenlahti on November 12, 2026, before deciding how to package combined offers across apps. For Android interface work, the Compose paywall article, by Jaewoong Eum on November 11, 2026, traces a brief pause after a package tap and measures each attempted fix separately, which is a useful model for performance work: change one variable, observe the result, and avoid bundling several fixes into a single conclusion. The KSP article, by Jaewoong Eum on October 27, 2026, takes a similar approach, beginning with a processor you might write and then examining the input model and the conditions under which that tool fits.
Tooling, data, and cross-platform builds
Use the Google Play Billing handbook comparison, by Jaewoong Eum on October 22, 2026, to pick a reading path based on your situation, rather than reading both guides cover to cover. The AI Toolkit update, by Jaewoong Eum on October 16, 2026, covers a command-line interface, reviewable store-product plans, and paywalls designed with codebase context. The Charts v3 post, by Charlie Chapman and Rhys Kentish on October 6, 2026, describes a rebuilt data infrastructure aimed at fresher data for decisions. For cross-platform build questions, the KMP article, by Jaewoong Eum on September 30, 2026, explains wrapping a Swift package in a Gradle module so the iOS SDK can be built without CocoaPods.
Paywall design and custom flows
Foldable-device advice, by Austin Blake on September 28, 2026, covers how paywalls should adapt across multiple fold angles. Custom Android purchase flows, by Jaewoong Eum on September 18, 2026, show how to build a paywall against the Purchases SDK. The animation article, by Jaewoong Eum on September 7, 2026, compares Rive and Lottie bundles, including code changes and constraints to check before choosing either option. The paywall AI Editor piece, by Azat Valiev on November 10, 2026, offers lessons from rebuilding an agent after a single internal message prompted a new approach.
When an agent takes action
Rico can now make changes rather than only suggesting them. In the example given, by Austin Blake on September 14, 2026, it built a paywall experiment off a default paywall, creating the Offerings, the Packages, and a running test with no dashboard required. Use this as a concrete check when evaluating agent tooling: does it produce the actual store objects and a running test, or only a recommendation?
How to use this index
A practical reading checklist: first identify whether your question concerns Apple, Google Play, or Android implementation; second, check whether the article is a migration notice, a measured performance investigation, or a build walkthrough; third, note any stated constraints before adopting the approach in your own app. Pagination is available for older posts beyond the first page.
Frequently asked questions
Apple's Bundles and Suites: which is right for your app?
This post, by Perttu Lähteenlahti on November 12, 2026, prepares you to offer combined subscriptions with your apps, comparing Apple's Bundles and Suites options so you can decide which structure fits your app.
When Is KSP the Right Tool?
This post, by Jaewoong Eum on October 27, 2026, starts from a KSP processor you might write for the job, then traces inward through KSP's input model and the conditions where it fits.