As promised, let’s talk about a topic I brushed on last month while I dug into the nitty-gritty of podcast statistics data: Listener-listen percentages, the holy grail of podcast statistics data.
By looking at the raw data of the media server logs, we can now calculate exactly how much of an individual media file was delivered to a listener. With a great number of podcast listeners simply clicking “play now,” versus downloading the file first, oftentimes the entire media file is not delivered.
When you click “play now” on most devices, the media is delivered to you in chunks, aka the infamous Byte Serving. Byte Serving is essentially — without getting too deep — how Apple and other mobile providers send you the media in pieces instead of downloading the entire file at once. Depending on your Internet connection and a lot of other variables, a 100 mb file could be broken up into 100 chunks on one request and 500 the next.
So if you listen to 15 minutes of a 30 minute podcast before clicking “stop,” there are many chunks / minutes of the podcast media file that have not yet been served to you. With this data we are now able to get an exact percentage of a file downloaded.
We can stitch those chunks back together and tell exactly how much, how little, or whether the file was served in it’s entirety. We have detailed data on each and every media file request. It is pretty neat when we can see that a listener scrubs forward to, let’s say, the 10 minute mark and starts listening there instead of the beginning.
By now you can see where I am headed. The media delivery percentages really tell an incredible listener engagement story. To do this it takes a huge amount of processing to stitch, calculate and build sensible reports for our corporate clients. This data then allows them to do a lot of cool things. Here are a few:
*Make programming changes based on trends showing when an audience bounced out.
*Determine peak listening for ad placement before drop off.
*Provide accurate billing to advertisers.
One of our vendors had a show that lost about 80 percent of its audience each episode around the 23 minute mark. The producers knew that at that point in their program was a segment change. Upon removal of that segment, nearly their entire audience kept listening through to the 45 minute mark.
In another show, the audience scrubbed up — or jumped ahead — to about the 5 minute mark before they started listening. The show hosts revamped the beginning of their show and advertised the new change at the 7 minute mark, and regained the audience at the intro.
I want to be very clear here: This gives our clients inferred data on what is happening with each and every episode, no one to date is providing a signal that an app has been closed or the listener hit stop. An assumption that they hit stop can be made, but may not always be the case.
The bottom line is that the listening session ended. If they come back later and pick up where they left off, we have other techniques that allow us to account for that action as well.
Simply watching the trending lines of the show’s audience over time has allowed our clients to tweak their shows, gain advertising revenue by better placement, and use a high level of sophistication to understand exactly what is happening with their listening audience.
For podcasters that host their podcast media with Blubrry, we will have an option to opt-in for similar data in their stats later this year, along with some yet-to-be announced data sets that will enable us to “close the loop.”
My goal in these first three articles has been to educate you that measuring media accurately truly is rocket science and we are pretty pleased to be the scientist behind that rocket. My team lives and breathes this everyday, and we hope that all networks and podcasters alike will trust us as tens of thousands of podcasters, networks and radio stations already do in their podcast media measurement.
Next month I want to switch gears and talk about mobile and the trends we are seeing in the utilization of mobile devices and even apps that are trending in the space. I will also cover some of the frustrations we have in tracking some of the mobile apps being used by podcasters today.
At Blubrry, we want to work together as a community to make sure that there are solid, reliable statistics and no misleading numbers in the podcasting space. If all podcasters utilized trusted solutions the space would be much better off in the long run.
Send your comments and questions to email@example.com
Catch my personal podcast @ GeekNewsCentral.com and tune in to our weekly New Media Show co-hosted with Rob Greenlee at NewMediaShow.com.