Loading content…
The number of new reviews an app received over a rolling window — typically 7, 30, and 90 days. Velocity is more diagnostic than total review count for established apps: a 5-year-old listing with 10,000 reviews and a v7d of 0 is signaling stagnation, while one with v7d of 50 is signaling healthy momentum.
Also known as: review frequency, review rate, review momentum, review cadence, new review rate
AppRanks calculates velocity by diffing successive review-count snapshots, so it captures real new reviews rather than estimating from rating-change patterns. The diff approach also means we capture review removals (when a marketplace deletes spam) — these surface as negative-velocity blips. For a Shopify app with 100 reviews and a v30d of 12, the rule of thumb is healthy: roughly one new review every 2-3 days, putting it on the same growth curve as the category median. We display velocity as a raw count rather than a rate so it stays comparable across apps with very different rating bases.
Review velocity is the earliest visible health signal on a marketplace listing. A static rating with a v30d of 0 looks identical to a rating with v30d of 50 in headline metrics — but the second app is alive and likely growing, while the first may be a stalled listing whose merchant base has churned. For developers planning a campaign or release, velocity tracks whether new releases are converting customers into reviewers (which feeds future install velocity through marketplace ranking signals). For competitive intelligence, comparing a competitor's v90d trajectory against your own is more diagnostic than comparing total review counts, because it normalizes for time-since-launch differences.
AppRanks computes velocity by diffing the total review count between successive snapshot timestamps, then aggregating into 7-day, 30-day, and 90-day rolling windows. The diff approach captures real new reviews rather than estimating from rating-change patterns — and it correctly captures review removals when the marketplace deletes spam or fraudulent reviews (these surface as negative-velocity blips on the timeline). For apps newer than 90 days, the v90d window is a partial measurement; we display the actual count rather than annualizing.