Podcast measurement methodology
Audion provides a set of metrics that is fully compliant with the podcast measurement standards version 2.1 defined by the Interactive Advertising Bureau (IAB), ensuring all data is consistent, comparable and accurate across the industry.
Human audience
We filter out requests from known bots or data centers using exclusion lists of:
- IPs from common data centers
- User Agents from bots, crawlers and spiders
- User Agents from blacklisted devices that automatically trigger duplicate downloads (like Apple Watchs)
We update exclusion lists every 90 days. We do not have inclusion lists.
Unique listener
We identify unique listeners by the combination of their IP address and User Agent (the identifier for their device and listening app). We hash this combination using the SHA-1 algorithm. Before hashing, we truncate IPv6 addresses to the first 64 bits, replacing the interface identifier that changes constantly by ":0000:0000:0000:0000".
Metrics
Stream
We count a stream when a unique listener downloads at least one minute of an episode's content.
To ensure accuracy, only one stream per unique listener for the same episode is counted within a rolling 24-hour window.
Daily listener
We count a daily listener when a unique listener has at least one stream within a rolling 24-hour window.
Impression
We count an impression when a unique listener fully downloads an ad's content during a stream.
Self-audit record
2024-05-15: IAB 2.1 certification
Updated on: 20/10/2025
Thank you!