Rewarded Placement Design for Apps With a Paid Tier

A rewarded placement earns when the reward is the thing the user was already trying to get and could otherwise only pay for. Put the same rewarded button on a settings screen and opt-in runs five to ten percent of daily users. Put it at the moment a build fails on quota and it runs three to four times that, because the user is already standing at the wall you are offering to move. Where the placement goes: We map friction points from your own event data before designing anything: where sessions end, where a quota is hit, where a paid feature is touched by somebody who has not bought it. The placement goes at the wall, not near it. In the developer tools we work on, that has meant extra build minutes, API calls above a plan's quota, and one-off exports of a report the free tier will render but not let you take away. In field and industrial apps it has more often meant no rewarded placement at all, and we have said so and stopped there. The reward economy: A reward with no ceiling stops being worth a minute of anyone's attention. We design the daily cap, the sink, and what happens at the cap, where an offer of the paid tier converts better than any other moment in the app. The reward has to be consumable and it has to expire. Permanent grants bought with attention erode the price of the paid tier within a quarter, and that damage does not show up in any ad report. The cannibalisation test: This is the part most rewarded work skips. Before the design ships to everyone it runs against a user-level holdout for twenty-eight days, and we read trial starts, paid conversion and subscription revenue in both arms, not only ad revenue. A rewarded loop that adds four euro cents per daily user and costs two points of trial conversion is a loss, and it will look like a win in every ad dashboard you own. Where the test says the design costs more than it earns, the recommendation is removal, and we have delivered that recommendation and been paid the same fee for it. Validation: Server-side verification on every reward, callback signed and verified on your backend, reward granted from the server and never from the client. Without it the reward is a client-side boolean and it will be found and farmed. We specify the endpoint and the verification, your engineers implement it, and we test it by trying to cheat it. Copy: Opt-in copy written by us in English and Italian. It states what the user gets and what it costs in seconds, and it does not use the word free where the user is paying with a minute of their day. The Italian is written rather than translated, because the register that works on a technical Italian audience is not the one a localisation vendor will hand you. What is not included: Mediation setup, floors, network configuration, and the rewarded creative itself. We design the placement, the economy and the test. Your team ships it and we read the result. Who this is not for: Apps on a company licence where the user is not the payer. A technician watching an advert to get at a feature their employer has already bought is a design with no economics under it, and we decline that work. Apps aimed at children. And products whose paid tier sells precisely the thing the reward would give away, unless you are willing to hear that the right answer is no rewarded placement at all. What you get: The placement map against your own funnel events, the reward economy with caps and sinks specified, the opt-in copy in both languages, the server-side verification spec, and the holdout design with its read.