Observed
A captured storefront or official public response supports the statement.
PageLens uses an evidence-first method. It distinguishes a fact visible in a selected storefront from a strategic hypothesis and from data supplied by an authenticated app owner.
A captured storefront or official public response supports the statement.
A connected store account returned the field for the app owner.
A recommendation or pattern to test; no outcome is claimed.
A public storefront capture can verify only part of a listing. Basic checks may all pass while creative effectiveness, localization, reputation and real conversion remain unassessed. Presenting that result as a perfect score would overstate what the evidence establishes, so current reports lead with the percentage of the rubric assessed and the source-linked improvements.
We retain versioned automated-check values behind the report for historical compatibility and consistent rechecks, but they are not presented as an overall grade. “Not assessed” never counts as a failure. We also do not penalize a short name, concise description or fewer screenshots by themselves.
Not scored from screenshot inventory or AI agreement. Visual notes and proposed sequences are hypotheses, not conversion evidence.
Published upper limits for captured subtitle/short description, full description and Apple promotional text. Each available field across the captured markets receives an equal share of this category's points. Missing fields are excluded, not failed.
Published 30-character upper limit, checked for each captured market. Unused space does not lose points. This does not assess search relevance or ranking.
Not scored from market count or matching titles. A brand can correctly retain the same name across markets; translation quality and market fit need additional evidence.
Not scored from aggregate ratings alone. A low rating can prompt further review, but does not establish recurring themes or their causes.
A usable owner-connected baseline for the selected market, or a clearly labelled all-market baseline. This checks data availability, not campaign success. Unavailable data is not scored.
Earlier v1 values remain historical and are not recalculated. Retired count-based findings stay in the evidence record, but are removed from next actions. Rechecks should be compared through matching scope and underlying evidence, not a headline number.
A free check inspects one market. A paid public report inspects the primary market and up to seven selected additional markets, subject to each storefront returning verifiable data. Captured source URLs, markets, timestamps and unavailable fields are retained in the report receipt. Storefront pages can change, and an unavailable parser field is not proof that the app lacks it.
Private keywords, impressions, conversion, rankings, downloads and revenue are never guessed. A read-only owner connection can add authorized metadata, reviews, localization inventory and performance reports where the connected API makes them available. Apple requires an Admin key to request a new Analytics Report type for the first time, and says the first ongoing report can take 24–48 hours. Google connections require a service account with the minimum Play Console permissions needed for the selected reports. Draft copy and screenshot concepts require owner review against the real app and platform rules before publication. A report is private and excluded from search indexing by default.
PageLens does not predict lift. Its experiments specify a treatment, a metric and guardrail; the store’s experiment tool and actual traffic determine whether a change wins. No review incentives or artificial ratings are recommended.