How beehiiv measures podcast downloads (IAB v2.2 compliant)

Podcast analytics on beehiiv are measured in accordance with IAB Tech Lab Podcast Measurement Guidelines v2.2. This article explains how downloads are counted, what traffic is filtered out, and how the metrics in your analytics dashboard are calculated.

A download is counted when a listener's app or player requests your episode file from beehiiv's servers and transfers enough data to cover at least one minute of audio. For episodes shorter than one minute, the entire file must be downloaded for it to count.

Raw download data includes requests from bots, crawlers, and automated systems that aren't real listeners, beehiiv removes all such requests before reporting any numbers.

• Bots and crawlers: Every request is checked against the IAB/ABC International Spiders & Bots List and the OPAWG List. Any known non-human traffic is excluded.

• Data center traffic: Requests from known data centers and cloud providers (e.g., AWS, Google Cloud) are excluded using the IPHub list, which is updated regularly.

• Pre-loads: Some apps automatically load episode files before a listener presses play. We exclude any request that doesn't transfer enough data to cover 1 minute of audio and any embedded metadata, such as title, artist, or artwork, so pre-loads don't inflate your numbers. The player present on beehiiv-hosted websites is set to preload=none by default.

• Apple Watch downloads: When someone listens on an Apple Watch, their iPhone automatically downloads the same episode in the background. Per IAB 2.2 guidelines, that duplicate download is excluded.

• Header-only requests: Requests that only check whether a file exists or has changed, without actually downloading content, are excluded.

Unique listeners are counted using a combination of IP address and user agent (the app or podcast player making the request).

Deduplication window: A fixed 24-hour window is used, split by day in Coordinated Universal Time (UTC). If the same listener downloads the same episode more than once within the same day, it counts as one download and one listener. 

Play-pause-play: If a listener pauses an episode and resumes later, the separate requests are combined and counted as a single download.

IP addresses are used to identify unique listeners. beehiiv handles IPv4 and IPv6 addresses differently during deduplication.

Download (‘Downloads’): A unique, validated request for an episode that passed all validation filters.

Listener (‘Unique Listeners’): A unique IP and user agent combination within a given time period (day, week, month). One listener can account for downloads across multiple episodes.

Ad Delivered (coming soon): Will count an ad as delivered when logs confirm the portion of the episode containing the ad was downloaded as part of a valid download.

In accordance with IAB 2.2 requirements, beehiiv:

• Conducts a self-audit at least twice per year to verify our filtering logic matches IAB 2.2 specifications.
• Re-validates the IP exclusion list at least every 90 days.
• Runs automated daily monitoring to catch anomalous traffic spikes.
• Maintains records of all self-audit findings, available to relevant parties on request.