Case Study 23-2: How Spotify for Podcasters Changed the Podcast Analytics Game

Background

For most of the podcast industry's history, creators operated nearly blind. The primary metric — "downloads" — was a crude proxy for listenership that overcounted significantly and told creators almost nothing about how their content was actually performing.

The industry ran on download benchmarks: if you got 3,500 downloads per episode, you were in the top 10% of podcasts globally. If you got 25,000, you were in the top 1%. These numbers were relative comparisons to other download counts — not actual measures of listener behavior, retention, or engagement.

Podcast advertisers priced deals on CPM (cost per mille downloads) — meaning advertisers were paying for download requests, not verified listens. The entire economics of podcast monetization was built on an imprecise metric that everyone knew was imprecise but nobody could replace.

This changed when Spotify acquired Anchor in 2019 and began building Spotify for Podcasters — a free hosting and analytics platform that, because Spotify controls the streaming environment, could for the first time provide genuine listen-through data for Spotify streams.

What Spotify for Podcasters Changed

The core innovation: Spotify could track exactly how long each listener streamed each episode, on a second-by-second basis, across all Spotify listeners.

This produced metrics the podcast industry had never had at scale:

Completion Rate: What percentage of Spotify listeners completed 80%+ of the episode. (Spotify defines "completed" at 80%.)

Listen-Through Rate: Available in detailed episode analytics, this is a curve showing at what points in an episode Spotify listeners tended to stop listening — the podcast equivalent of YouTube's audience retention graph.

Real Listener Count: Distinct Spotify accounts that played an episode (not download requests — actual plays).

Follower Demographics: Age, gender, and country data for followers on Spotify.

Listen Time by Day and Hour: When Spotify listeners are engaging with episodes.

The Analytics Gap Between Platforms

Consider what this meant practically. A podcast hosted on a traditional platform (Libsyn, Buzzsprout, Podbean) without Spotify for Podcasters integration sees primarily: - Total download count (inflated) - Downloads by episode (relative comparison) - Basic geographic data - Listening apps used

A podcast using Spotify for Podcasters sees all of the above (via third-party stats from hosting platforms) PLUS: - Completion rate by episode - Listen-through curve (where Spotify listeners drop off) - Spotify stream count vs. download count (allowing calculation of the overcount gap) - Follower vs. non-follower listener split

This creates an information asymmetry between creators who have adopted Spotify for Podcasters and those who haven't. Creators with completion rate data can identify episodes that underperformed despite high download numbers (low completion = content didn't hold), and episodes that overperformed in engagement despite lower downloads.

A Practical Example: Two Episodes, Same Download Count

To illustrate the value of completion rate data, consider a hypothetical example that reflects real patterns in podcast analytics:

Episode A: "My Story — How I Started This Podcast" - Downloads: 3,200 - Spotify completion rate: 34% - Listen-through: major drops at 8 minutes and 14 minutes

Episode B: "The 5 Mistakes First-Time Investors Make" - Downloads: 3,100 - Spotify completion rate: 71% - Listen-through: gradual decline, no major drop-off

By the download metric, these episodes are nearly identical. By the Spotify engagement metric, Episode B is performing twice as well — its audience is staying to the end, engaging with the full content, and presumably finding it more valuable.

For a creator making strategic decisions about content type: this data tells them their audience engages much more deeply with practical, numbered-list educational content than with personal narrative content. Without completion rate data, they'd see roughly equal download numbers and have no signal.

Apple Podcasts Connect: The Complementary View

Spotify for Podcasters covers Spotify listeners — which represent a very large but not complete portion of podcast listening (Spotify's podcast listening market share was approximately 31–35% globally as of 2024, with Apple Podcasts at approximately 18–22%).

Apple Podcasts Connect provides complementary analytics for Apple Podcasts listeners, including: - "Engaged listeners" — listeners who completed 40%+ of an episode - Listener graph showing follow/unfollow over time - Performance by episode

The limitation: Apple Podcasts' analytics are less granular than Spotify's, partly because Apple's ecosystem is more privacy-protective by design.

Using both platforms together gives creators their two largest listener segments with engagement data — covering roughly 50–55% of all podcast listening with behavioral metrics.

Industry Response: The IAB Standards

The podcast advertising industry recognized the measurement problem and has worked to address it through the Interactive Advertising Bureau (IAB)'s Podcast Technical Working Group, which publishes technical standards for podcast measurement.

IAB Podcast Measurement Technical Guidelines (Version 2.0 and later) set standards including: - A "download" should only count file requests of sufficient length to indicate meaningful intent to listen - Filtered bots and automated crawlers from download counts - Minimum byte thresholds for counting a download

These standards improved download count accuracy without solving the fundamental problem: they still measure file requests, not completed listens. Spotify's completion rate data represents a different paradigm — measuring actual listening behavior rather than file delivery.

What This Means for Creator Strategy

The Spotify analytics development has a broader lesson for creators across all platforms: the quality of your analytics tool affects the quality of your strategic decisions.

Creators who ran podcasts before Spotify for Podcasters existed were making content decisions based on relative download comparisons. "This episode got more downloads than that one, so the topic worked better." That's a reasonable but weak signal — it conflates topic appeal (what gets clicked/downloaded) with content quality (what holds listeners to the end).

With completion rate data, the question becomes more precise: "This episode got similar downloads but twice the completion rate — meaning this content type is twice as engaging for my audience, which likely means twice as many people completed my CTA at the end."

For monetization: podcast advertisers who buy by CPM-downloads are paying for the reach metric. A creator who can demonstrate 70%+ completion rates can argue — correctly — that their mid-roll ad at the 20-minute mark is actually being heard by a much higher percentage of listeners than a competitor with the same download count and 35% completion. The data supports a higher rate.

The Limits of Better Analytics

One important counterpoint: better analytics tools can create their own form of analytics paralysis. Several podcast creators, upon gaining access to listen-through curve data, began making anxious micro-adjustments to their episode format — shortening segments, cutting music intros, restructuring mid-rolls — based on a few weeks of Spotify data.

The sensible response is the same as it is across all platform analytics: use the data to identify significant, persistent patterns (completion rate consistently below 50% across multiple episodes is a real signal) rather than reacting to individual episode variations (one episode's completion rate dropped because it aired during a major news event).

Analysis Questions

  1. The podcast industry was built on "downloads" as its primary metric for years, despite everyone knowing it was an overcounting metric. What does this tell us about how metrics standards get established in industries, and why do imprecise metrics persist even when better options exist?

  2. Spotify for Podcasters provides completion rate data only for Spotify streams, which represents roughly 31–35% of podcast listening. A creator sees a 68% completion rate on Spotify — but has no data on how Apple Podcasts, Overcast, or other app users listen. How should a creator interpret partial analytics data like this? What assumptions is it safe to make?

  3. The chapter argues that "better analytics tools affect the quality of strategic decisions." Can you think of a situation where having more detailed analytics would lead to worse creative decisions — where the data would constrain or mislead a creator rather than helping them?

  4. Podcast advertisers pay on CPM-downloads. If you were a podcast creator with completion rate data showing 72% listen-through and you were negotiating an ad deal with a brand paying industry-standard CPM, how would you use your completion rate data to negotiate a better rate?

  5. Spotify's analytics advantage comes from controlling the streaming environment — they can track exactly what happens during playback. What are the privacy implications of this level of listener behavior tracking, and how should podcast creators think about their ethical obligations to their audience when using platforms that collect this data?