“There are three types of lies — lies, damn lies, and statistics.”
― Benjamin Disraeli
Providing audience numbers for media has never been an easy task, but when it comes to online metrics, it is a challenge that has been getting increasingly difficult. Let’s just look at podcasting, which has skyrocketed in popularity since we first started doing it in 2005.
When we first started ZimmCast in 2005, there were no podcast hosting services, so we had to host it on our own server and get statistics by manually putting our codes into a tracking system. Eventually, we moved to Libsyn about ten years ago. Using Libsyn provides hosting of the audio files which is good for our own server storage and they provided more information about the podcast episodes than we were getting. They provided not only download numbers, but also geographic data and more.
A few years ago Libsyn started tracking two different sets of numbers for podcasts – Unique and IAB, with IAB being Internet Advertising Bureau Podcast Measurement Compliance Guidelines. The difference between the two metrics was generally pretty significant in our case, with the unique being by far the largest of the two numbers. The difference could be in the thousands.
As far as IAB is concerned, they work for the advertisers who want to make sure they are getting a true measure of the audience for a podcast they might want to sponsor. So they filter out any suspicious IP addresses that might be linked to bots or web crawlers and essentially report almost exclusively downloads tied to subscriber apps like Apple Podcasts. In our case, most of our podcasts are downloaded via a url on a web page, social media post or email, so we can have very high unique numbers.
But IAB has started cracking down on reported podcast download statistics and in October Libsyn made a major reporting change that resulted in the Unique numbers suddenly dropping to mirror the IAB numbers within a few downloads. An audio file that would have had 1000 unique downloads in September may now only get about 100!
Why were the Unique downloads so high before and now almost the same as the IAB numbers? Were they ever accurate in the first place and are they now? How do we know and how do we explain it to our clients?
I have been back and forth with Libsyn technical support and management and discussed it with other podcast producers and providers – what do you think?