What is the difference between a download and a play?

A download is an RSS-side event. Under the IAB Tech Lab guidelines, roughly 60 seconds of the audio file must be requested, and repeat requests from the same IP address and user agent are deduplicated within a 24-hour window. A play is a platform-side event with its own threshold. Spotify counts a play at 30 seconds of listening or watching as of June 2026, while Apple counts a play the moment someone presses play. Three different systems, three different numbers, all correct.

Why do my hosting numbers not match Apple and Spotify?

Because they measure different things at different points in the chain. Your host counts file requests. Apple and Spotify count in-app activity for their own users only. Your host's number should be the largest because it includes every app. If Apple or Spotify exceeds it, check whether your feed redirect or prefix is configured correctly.

How many downloads is a good podcast?

There is no useful absolute answer, and any article giving you one is guessing. The number that matters is your own trend. A show growing 10% month over month with 400 downloads per episode is in better shape than a flat show at 4,000. For sponsorship, audience specificity beats volume until you reach the scale where marketplaces sell on reach alone.

What is a good podcast completion rate?

Track it against your own baseline rather than an industry figure, because completion varies enormously by episode length and format. Apple reports Average Consumption, the percent of an episode played per device, and it is the most useful single number Apple gives you. Watch the direction it moves after you change your intro, your episode length, or your ad placement.

Do I need a podcast analytics tool beyond my host?

You need one if you publish across multiple platforms, want consistent numbers across a back catalogue, run sponsorships that require a single agreed delivery figure, or manage more than one show. A measurement prefix sits in front of your existing host and gives you one dataset that does not change when you switch hosting providers.

Podcast analytics: the metrics that actually matter

Podcast analytics is the practice of measuring what happens after you publish, using numbers whose definitions you can state out loud. That second half is where most shows come unstuck. Downloads, plays, listeners, streams, and followers are counted by different systems, at different moments, under different rules, and comparing them without knowing the rules produces confident wrong conclusions.

This guide covers what each number counts, which ones deserve your attention, and how to build a reporting routine that survives a sponsor asking hard questions.

What each podcast metric actually counts

Start with the definitions, because everything downstream depends on them.

Downloads are counted from your RSS feed by your host or your measurement prefix. The IAB Tech Lab Podcast Measurement Technical Guidelines version 2.1 is the standard buyers expect: at least roughly 60 seconds of the audio file has to be requested, and repeat requests from the same IP address and user agent are deduplicated inside a 24-hour window. That filters out partial buffering and most automated traffic. What counts as a podcast download covers the edge cases in detail.

Plays are counted inside a listening app, and each app sets its own bar. Spotify redefined a play in June 2026 as an episode watched or listened to for at least 30 seconds, developed alongside the Alliance for Measurement in Podcasting. Apple counts a play as the moment someone presses play, with no duration threshold.

Listeners are unique people or devices rather than events. Apple reports Listeners as the total number of people who listened to or watched your show, and separately reports Engaged Listeners, defined as people who played at least 20 minutes or 40% of an episode.

Followers and subscribers are intent signals. Someone tapping follow tells you they want the next episode. It does not tell you they listened to this one.

Consumption is how much of an episode gets played. Apple's Average Consumption is a percent average of how much of the episode played per device.

| Metric | Counted by | Threshold | Best used for | |---|---|---|---| | Download | Host or prefix | ~60 seconds requested, 24-hour dedup | Sponsor delivery, trend tracking | | Play (Spotify) | Spotify | 30 seconds played or watched | Platform-level engagement | | Play (Apple) | Apple | Press of the play button | Directional interest | | Engaged Listener (Apple) | Apple | 20 minutes or 40% of episode | Content quality signal | | Follower | Platform | Tap of follow | Audience intent | | Average Consumption | Apple | Percent played per device | Episode structure and length |

The three questions your analytics should answer

Numbers without a question attached become a dashboard you check and forget. Every reporting session should answer three things.

Is the audience growing? Look at rolling 30-day downloads per episode, using the median across your last 8 to 12 episodes. The median protects you from one outlier episode distorting the picture. Month-over-month direction matters far more than any single week.

Are people finishing? Average Consumption on Apple and completion on Spotify tell you whether the episode delivers on its opening. A show that grows downloads while consumption falls is buying attention it cannot keep, and that shows up as a plateau three months later.

Where are new listeners coming from? New versus returning audience splits, platform breakdowns, and geography let you tell discovery from loyalty. Spotify's 2026 analytics update added a first-time versus returning listener breakdown specifically for this.

Answer those three, and most other numbers become optional. The case for podcast analytics covers why the decision has to come before the dashboard.

Why your numbers never match across platforms

Podcasters lose hours to this every month, so it is worth stating plainly.

Your host counts file requests across every app that touches your feed. Apple counts activity from Apple Podcasts users only. Spotify counts activity from Spotify users only. YouTube counts views under its own rules. These are four measurement systems observing four overlapping but different populations.

The practical consequences:

Your host or prefix number should be the largest. It sees everything. If a platform figure exceeds it, your feed redirect or prefix is misconfigured.
Never add platform numbers together. Apple plays plus Spotify plays plus your download total triple-counts a large share of the same people.
Pick one number for sponsors and keep it. Downloads under the IAB standard, in a stated 30-day window, is what buyers understand.
Compare each platform only against itself over time. Cross-platform level comparisons are noise. Cross-platform trend comparisons are signal.

Build a reporting routine you will actually keep

Set your window. Thirty days after publication is the standard. Use it in every report so numbers stay comparable across months.
Install a measurement prefix. A prefix sits in front of your existing host and counts every request under one consistent rule, so your data survives a hosting switch. Installing Podder's prefix takes a few minutes across the major hosts.
Record five numbers per episode. Median 30-day downloads, Apple Average Consumption, Spotify plays, new followers, and one response signal such as a link click.
Review monthly, not daily. Podcast data is noisy at a daily resolution. Monthly review with a rolling three-month trend line is where real patterns show up.
Write one sentence of interpretation per month. What changed, what you think caused it, and what you will test next. This turns a spreadsheet into a decision log.
Check the data threshold. Apple needs at least five unique listeners in the selected timeframe before it shows data, and defaults to a 60-day view. Small shows should widen the window rather than assume the data is broken.

The 7 key podcast marketing metrics makes a good starting scorecard if you want a template.

Common podcast analytics mistakes

Chasing total lifetime downloads. Lifetime totals only go up and tell you nothing about current health. Per-episode medians tell you where you actually are.

Treating a chart position as an audience metric. Charts measure recent momentum relative to other shows. A drop from position 40 to 90 usually means someone else launched, not that you lost listeners. How podcast charts work explains the difference.

Switching measurement windows between reports. Reporting 30-day figures in a good month and 90-day figures in a bad one destroys sponsor trust permanently.

Reading one episode as a trend. Single-episode swings are usually publication timing, a holiday, or a platform anomaly.

Ignoring consumption entirely. Downloads measure whether your title and cover art worked. Consumption measures whether the episode worked. Growth without the second one does not compound.

Where to go next

Podcast analytics gets useful the moment your numbers stop moving depending on where you look them up. One consistent dataset, one agreed window, one monthly review, and three questions you keep asking. That is the whole system, and it beats a wall of dashboards you never open.

If you are migrating measurement after Chartable's shutdown, our comparison of Podder and Chartable covers what transfers and what does not, and the analytics and insights tools directory lists the wider field.

Ready to get one consistent set of numbers across every platform? Start with Podder Analytics.