| Patent number | US12561221B2 |
| Official title | Techniques for managing performance metrics associated with software applications |
| Assignee | Apple Inc. |
| Inventors | Nicholas Kistner, Vivek Krishnan, Andrew T. Maher, Fabio Benedetti, Raymond T. Powers, Ari K. Campbell-Marder, Matthew Tan Teik Hoe, Gregory D. Eldridge, Arseniy Kaplun, Daniela S. Antonova, Peter J. Gremett |
| Priority date | February 26, 2023 |
| Filed | May 23, 2023 |
| Granted | February 24, 2026 |
| Status | Active. Anticipated expiry October 24, 2043 |
| Family | European counterpart EP24157407.8A, continuation US20260220015A1 (pending) |
| Source | patents.google.com/patent/US12561221B2 |
What This Patent Covers and Why It Matters for ASO
Feature it describes: the way App Analytics presents your performance against comparable apps.
Every developer who has opened the benchmarks view has wondered why Apple shows a percentile band rather than a position. This patent answers it, and the answer is in claim 1 rather than in a design rationale.
The claimed method obtains your app’s metric, generates percentile rankings based on the metrics of similar applications, identifies which percentile range your app falls within, and then displays two things: your actual metric value, and the percentile range. The claim then specifies that the display deliberately obscures where inside that range your metric sits.
That obscuring is the invention. It is not a simplification for readability. It exists so that competing developers cannot reconstruct each other’s exact standing by watching band boundaries move.
Consequences for how you use the tool:
- Movement inside a band is invisible to you. You can improve substantially and see no change in the display, and you can decline substantially and see no change either. Only band crossings are observable.
- Band crossings are therefore large events, not small ones. Treat a crossing as a real signal and treat a stable band as no information rather than as evidence of no change.
- Your own absolute value is shown and is exact. That is the number to track over time. The band is context, not measurement.
The metrics the patent names are downloads, retention, stability, monetization and conversion rate. Stability appearing alongside the commercial metrics is worth noting. Crash behavior is being presented to developers in the same frame as conversion and retention, which is consistent with Apple treating technical quality as a performance dimension rather than a support issue.
Patent Summary
A computing device obtains a performance metric for an application. It generates percentile rankings for that metric based on the corresponding metrics of similar applications, then determines which percentile range the application’s own metric falls into.
The interface then displays the application’s actual metric value together with the percentile range it occupies, while withholding the application’s precise position within that range. The result is that the developer learns their own exact figure and their approximate standing, and learns nothing about any individual peer.
The metrics covered include downloads, retention, stability, monetization and conversion from free to paid. A European counterpart was filed and a continuation remains pending, so the family is still being prosecuted.
Read alongside the peer group patent, the two describe one system in two halves: how the comparison set is assembled and privatized, and how the result is shown without leaking it back.
A detailed breakdown of the interface, the figures and the full claim language will follow in a later update to this article.
Related reading: US12430660B2 on how the peer group itself is built.
Bir yanıt yazın