We're writing to let you know that you will soon see a change in your listens and demographics numbers.

Why?  Our Data Sciences team is continuously working to refine the tools and methodologies that we use to collect, filter, and analyze the listener data we report to you from the many available sources.  As we learn, we improve.  Minor adjustments occur without fanfare.  Occasionally, however, a more significant adjustment of our site-wide metrics reporting baseline is warranted ... we call that a "rebase".

We audit our numbers frequently, to reconcile them across all sources and align them with evolving industry standards.  A year-long internal audit of our baseline numbers has informed us that we need to rebase.  Specifically, we've identified how "Cortex", our intelligent data platform, could make a finer distinctions between certain bot and human activity that had traditionally thought to be non-organic.  Cortex has now been re-tuned, and as a result many hosts will now see a modest increase in listenership.

Many months (and cups of coffee) went into achieving this new milestone in Cortex data intelligence.  Our metrics computation algorithms reflect current best practices in bot filtering and listening classification in podcasting audiences.

For example, based upon new learning we've:

  • recovered additional listening events that had previously gone uncounted due to incompatibilities and non-standard behaviors seen with certain clients.
  • enhanced our algorithms for distinguishing organic repeat listens as opposed to duplicate, non-organic, "false" listens.
  • implemented more sophisticated and nuanced routines for uniquely, yet anonymous, identifying player sessions associated with each IP address.
  • updated browser & podcasting application (client) identification techniques, to better understand which platforms and players your listens are coming from.

We hope to release this Cortex counting change over the next few days, at which time it will apply toward both historic and future numbers being reported.

Did this answer your question?