At a glance
D7 Retention % is the share of a cohort that returns and performs the retention event on the seventh day after first activity in Amplitude. It is the canonical product-stickiness gauge: far enough out to filter first-day curiosity, close enough to react to changes, and the number most teams anchor their retention reporting on. The card pairs with sibling metrics in the Analytics category to build a complete diagnostic picture; cross-reference the related cards listed below for context.
| What it counts | The percentage of users in a cohort who return and perform the retention event on day 7 after their first activity, using Amplitude’s N-day retention method. |
| Sample type | Backend API data from Amplitude, refreshed on the standard data refresh. |
| Why it matters | D7 is the industry-standard read on whether a product is forming a habit. It is stable enough to trust week to week yet responsive enough to flag real changes in engagement quality. |
| Reading the value | Compare the current 30-day window against the prior period and against an 8-week baseline. A sustained drop signals weakening stickiness rather than noise. |
| Currency | percent |
| Time window | 30D vsP |
| Alert trigger | drop >5pp vs 8wk baseline |
| Sentiment key | amp_retention_d7 |
| Roles | owner, marketing |
Calculation
Calculated automatically from your Amplitude data. Amplitude groups users into cohorts by their first event, then measures the percentage who returned and performed the retention event on day 7. This card reports that day-7 percentage over the 30-day window and compares it both against the prior period and against an 8-week rolling baseline, so a genuine shift in stickiness stands apart from week-to-week noise. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A representative reading of D7 Retention % for a typical merchant on Amplitude. Suppose the 8-week baseline sits at 15 percent and recent cohorts now return at 9 percent on day 7. That 6-point drop clears the alert threshold and flags weakening stickiness, not the random wobble a single cohort might show. Because D7 is the habit-formation checkpoint, a sustained fall here is a serious signal that the experience is no longer pulling users back the way it did. Use Vortex Mind to trace whether the decline traces to a release, a campaign mix shift, or a content change; for natural-language exploration, ask Ask Viq to compare D7 between first-time and repeat buyers.Sibling cards merchants should reference together
| Card | Why merchants reach for it |
|---|---|
amp_retention_d1 | The early-onboarding signal that usually leads D7. |
amp_retention_d30 | The long-run tail that confirms whether D7 stickiness endures. |
amp_retention_curve | The full curve that puts the D7 point in context. |
amp_stickiness_dau_mau | A complementary frequency view of how often active users return. |
amp_returning_users | The absolute repeat-user count behind the D7 percentage. |
Reconciling against Amplitude
Where to look in Amplitude’s own dashboard: Open the Retention chart, set the method to N-day, and read the day-7 column for your start and return events. Confirm the cohort range and segment filters match the Vortex IQ profile. Event Segmentation can verify the return-event volume, and the Dashboards section may pin a saved retention chart for team reference. Why the Vortex IQ value may legitimately differ:| Reason | Direction | What to do |
|---|---|---|
| Period boundary. Vortex IQ uses a 30-day rolling cohort window with an 8-week baseline; an Amplitude chart may use a different range. | Variable | Match the cohort date range in the Amplitude retention chart. |
| Time zone. Amplitude defines day boundaries by the project time zone; Vortex IQ aligns to the merchant reporting time zone. | Marginal | Confirm the project time zone matches your reporting time zone. |
| Filter or segment scope. Profile-level filters (platform, B2B, test users) change which cohorts feed the rate. | Variable | Apply the same segment to the Amplitude retention chart. |