Guides
Podcast audience analytics: what each dashboard counts
Podcast audience analytics: learn what Spotify, Apple and YouTube count, which metrics to track monthly, and where host exports and third-party tools fit.
What to take away
- Spotify for Podcasters, Apple Podcasts Connect and YouTube Studio each measure a different thing; none of them measures a listener.
- Downloads are server requests. A download is not a play, and a play is not a finished episode. Pick three to five metrics you will read every month.
- Set your own baseline before you compare yourself to any published median.
- Keep the raw exports. Platforms change definitions, and a renamed metric breaks a year of trend lines.
What each dashboard actually counts
Spotify for Podcasters reports starts, streams and listeners, and it splits consumption by episode and by platform. Its numbers cover Spotify only, so an episode can look flat there while Apple carries it.
Apple Podcasts Connect reports plays and unique devices, plus follower counts and completion where the listener has not opted out of sharing. Apple's privacy rules mean small shows see sampled or delayed figures rather than a full count.
YouTube Studio reports views, watch time and average view duration, and those are the numbers a video-first show should read first. A YouTube view is not comparable to an Apple play, so never add the two together.
Metrics worth a monthly slot
| Metric | What it answers | What it cannot tell you |
|---|---|---|
| Starts per episode | Did the title and thumbnail earn a click | Whether anyone stayed |
| Completion rate | Did the episode hold attention | Why they left |
| Followers added per week | Is the show compounding | Which episode caused it |
| Unique devices per month | How wide is the reach | Who those people are |
| Watch time on video episodes | Does the format work on YouTube | Whether audio listeners agree |
Read the table as a set, not as a scoreboard. A show with high starts and low completion has a packaging problem or an opening problem, and the fix is different in each case.
Completion needs one worked sum. A typical band for a mid-length episode is about 60 to 80 percent, and it falls as runtime grows. On your own show, 1,000 starts and 620 completions is 62 percent.
What each metric answers
Metric
- Starts per episode
- Title earned click
- Completion rate
- Held attention
- Followers added weekly
- Show compounding
- Unique devices monthly
- How wide reach
- Watch time on video
- Format works YouTube
Answers
- Starts per episode
- Whether anyone stayed
- Completion rate
- Why they left
- Followers added weekly
- Which episode caused it
- Unique devices monthly
- Who those people are
- Watch time on video
- Audio listeners agree
Cannot tell
- Starts per episode
- Completion rate
- Followers added weekly
- Unique devices monthly
- Watch time on video
Tools that do the counting
Spotify for Podcasters and Apple Podcasts Connect are free and come with hosting on those platforms. YouTube Studio is free with a channel.
Podtrac and Chartable measure across hosting platforms and give you a single download figure plus industry rankings, which is what most sponsors ask for. Both are paid, and pricing changes, so check their current published rates.
Your host matters more than any dashboard. Buzzsprout, Transistor, Libsyn and Captivate each publish their own analytics documentation, and the numbers they report are the ones a sponsor will accept. Read your host's definition of a download before you quote it.
Where audience analytics come from
| Origin of the number | What it is for | How often to read it |
|---|---|---|
| Your host's file server | The raw download figure you export and keep | Monthly, and before a sponsor call |
| The listening app | Detail on that app's own share of your listeners | Monthly, allowing for the app's lag |
| The video platform | Watch time for episodes you also publish as video | Weekly if video drives the show |
| A third-party measurement firm | One cross-platform figure a sponsor accepts | When a sponsor asks for one |
| Listener replies and surveys | Why people stayed or stopped | Quarterly |
Treat the host export as your baseline and the app dashboards as the split.
What is changing in podcast analytics
Apple's privacy rules and Google's consent requirements both limit how much listener-level detail reaches a publisher. Apple reports plays only from listeners who leave data sharing on, and it samples small shows, so the same episode can show a different count on the next refresh. Expect more sampled data and longer reporting delays, not less.
Reporting definitions move too. The IAB Podcast Measurement Technical Guidelines set what counts as a valid download, including filtering for bot traffic and duplicate requests.
Video podcasts keep pulling share toward YouTube, which means watch time and average view duration become the numbers a video sponsor asks about. Audio-only shows should keep a separate audio baseline rather than chasing a video metric.
Platform payout terms change without much notice. Read the published terms on each platform before you build a revenue projection on them, and label every projection as an estimate.
Checks to make before you quote a number
On disclosure, U.S. Federal Trade Commission: Endorsements, Influencers, and Reviews is worth reading first. The rules cover what you say about a product, and they cover the audience figures you quote beside it. Say where each number came from and which month it covers.
U.S. Copyright Office: Copyright and the Music Marketplace covers music rights inside an episode, which is a separate question from who listened. Keep clearance notes filed apart from your analytics exports.
Internal Revenue Service: What kind of records should I keep? allows any recordkeeping system that clearly shows income and expenses, as long as the documents support purchases, sales, payroll and assets. Clean records also feed the podcast revenue metrics you report each month.
Keep the raw analytics exports with those records, so a renamed metric stays easy to explain.
If you want the audience to compound rather than spike, the work is in growing a loyal podcast audience, not in adding another dashboard.
Common questions
Why do my download numbers differ between my host and Spotify?
Each counts a different event. Your host counts server requests for the audio file, while Spotify counts streams inside its own app. Say your host logs 3,000 downloads in a week and Spotify reports 900 streams. The host counts requests from every app; Spotify counts its own share. Neither figure is wrong, and they will never match.
What should a small show track first?
Starts per episode, completion rate and followers added per week. Those three tell you whether the packaging works, whether the episode holds, and whether the show is compounding.
How often should I compare myself to an industry median?
Rarely, and never as a target. Published medians come from different show sizes and genres. Set your own baseline first, then use a median only to sanity-check an order of magnitude.
Do I need paid analytics tools?
Not at first. Your host's dashboard plus Spotify and Apple will cover a show under a few thousand downloads per episode. Add Podtrac or Chartable when a sponsor asks for a third-party figure.
