Analytics audit update

Periodically, we perform an audit of our analytics data to ensure we’re doing the best job we possibly can to keep your analytics free from junk/bot listens. With the rise of AI, this work has gotten increasingly difficult, as it’s become easier and easier for bad actors to download episodes covertly.

We don’t normally announce analytics audits, but this one is notable because it includes the most substantial changes we’ve had in a long time.

1. First, we’ve dramatically expanded our corpus of data used to determine whether a listen came from a legitimate listener device or from a bot running in a data center. We’ve more than 10x-ed the number of rules that we run against each and every listen.
2. We’ve worked to classify even more podcatchers and browsers, including ones that are not well known. For podcasters, this means you’ll see more names appear under your “Listens by agent” view.
3. We’ve combed through our old rules and weeded out ones that aren’t useful anymore. This is important, because some old rules are doing more harm than good. For example, apps claiming to be Google Podcasts are definitely illegitimate: Google Podcasts was rolled up into YouTube Music quite a while ago.
4. Many of our existing rules were refined, helping to mark somewhat-suspicious listens from obscure devices as definitely legitimate or spammy.

Process change

Historically, for big updates like this, we’d re-process existing analytics data. While this has been our policy (it gives you the most accurate information possible), it would be very disruptive for us to do this to all Pinecast customers today given the substantial amount of bot traffic we’re seeing this year. Starting today, we’re changing that policy. If you’d like your analytics data filtered using the new rules, we’re happy to do so upon request. Please reach out to Pinecast support through the chat in your dashboard or at support@pinecast.com and we’ll be glad to assist you.