Podcast users
Podcast users shows member counts for Spotify connections and month-to-date feed access through other apps. The counts can overlap.
ReferenceView as Markdown
Podcast users shows member counts for Spotify connections and month-to-date feed access through other apps. The counts can overlap.
Stay in the loop 🚀
Total adds the Spotify and Other apps counts, so it cannot measure unique listeners.
| Series | Stored measure | Date used for the graph |
|---|---|---|
| Spotify | Distinct members with a Spotify user record; separate counts for All and Active. | Summary createdAt. |
| Other apps | Distinct members with feed access from the start of the month through the recorded date. | Summary dataDate. |
The report does not measure Spotify listening or establish when a connection was last used.
Other-apps aggregation reads podcast_access_daily, already deduplicated by member, podcast and day. It deduplicates members again across all podcasts for the overall count. Summing individual podcast audiences would count people repeatedly.
When no date is supplied, the other-apps writer uses the previous day and its month start. Feed access records an app requesting the episode list, not a completed play. .
Active selects the stored active count; changing the filter does not recalculate old snapshots against today's subscriptions.
The other-apps writer checks active owned subscriptions or accessible shared seats in the same tenant. It also excludes deleted members from the active count. The Spotify writer uses EffectiveMemberPackageStatusView to check active owner or shared-subscription counts.
The two active checks use different implementations. For an unexplained Active discrepancy, inspect both writers and the view before assuming identical tenant filtering. All selects the unrestricted connection or feed-access count; it still requires the relevant source record. .
Each repository selects its latest available snapshot inside each requested period and date range. Spotify and other apps can therefore use different saved dates in the same column. , .
The selected granularity changes snapshot grouping, not the feed-access measurement window. Daily and weekly columns still show month-to-date access. Quarterly and yearly columns take the final available month-to-date snapshot, rather than summing monthly audiences.
The service always returns two fixed series, with no Others category. Details shows their separate counts; Total adds them. For example, 120 connections plus 80 feed-access members produces 200, even if 30 members belong to both groups. The summaries contain no overlap count from which to calculate unique combined reach. .
Hiding a Details series changes the tooltip total and percentage denominator. Switching to Total sums both underlying series regardless of hidden labels. Tooltip percentages describe the visible counts' shares, not growth. , .
| Symptom | Check and interpretation |
|---|---|
| Other apps falls near a month boundary | Compare dataDate values. A new month restarts the accumulation window. |
| Daily and weekly values look similar | Compare the latest snapshot selected by each view. Neither calculates a daily or weekly unique audience. |
| Podcast totals exceed the overall audience | Check whether the same members accessed several podcasts. The overall count deduplicates across podcasts. |
| Total differs after changing view | Restore hidden series in Details. Its tooltip sums visible parts; Total sums both. |
| Missing or empty graph | Check the API response and saved source rows. The frontend renders an empty graph on missing response data. |
The page checks Analytics access, while /podcast-users-summary requires READ for DASHBOARD. Check both when a role can open the page but cannot load data. .
Source checked at 6cd1850 on 28 September 2026; UI observations are from 27 September. No new role replay or numerical reconciliation was performed. Spotify record retention and live aggregation freshness remain unverified. The UI help's selected-period wording conflicts with the month-to-date writer.
For controls and graph interpretation, see .