Podcast Measurement: A RAD Emerging Standard

• Accord­ing to research conducted in 2018 by the Reuters Insti­tute for the Study of Jour­nal­ism, 34% of indi­vid­u­als 18 years old and older across 22 devel­oped coun­tries have accessed a podcast in the last month
• Accord­ing to a study conducted by the IAB and PwC in 2018, U.S. podcast ad revenue will reach $659 million by 2020 (from $169 million in FY 2016 to $314 million in FY 2017 — this is pretty serious growth that is expected to continue)
• Spotify recently spent report­edly over $200 million to acquire 4‑year-old podcast produc­tion company Gimlet Media (and acquired Anchor at the same time)

Like most emerg­ing media (although, when you’re looking at more than one-third of the popu­la­tion engag­ing with the medium, is it really “emerg­ing?”), robust measure­ment tends to be lacking. In the early days of websites, we counted “hits,” which were a reason­able approx­i­ma­tion of scale and volume, but that was about it. In the world of podcasts, adver­tis­ers and publish­ers are still largely in the “hits” era, in that the only univer­sally avail­able metric is “down­loads.”

While there have been various efforts to at least shore up the defi­n­i­tion of that one metric (see the IAB Podcast Measure­ment Tech­ni­cal Guide­lines version 2.0 and the Public Radio Podcast Measure­ment Guide­lines version 1.1), stan­dards and mech­a­nisms for actu­ally measur­ing how much of a podcast was listened to, includ­ing whether mid-roll ads were consumed, has been a chal­lenge. This is because podcasts operate on a pretty old tech­nol­ogy: RSS (remem­ber when we used to “subscribe” to blogs and use a news­reader? It’s the same tech­nol­ogy!). Regard­less of the podcast app someone uses, the app is just reading the RSS feed and then down­load­ing an audio file if one becomes avail­able. The publisher can detect that that file was down­loaded but has no visi­bil­ity into whether the file was actu­ally ever listened to (or how much of the file was listened to).