In part 3 of this series on metrics, I stated that measuring unique downloads, while still an imperfect podcast metric, was one of the most popular ways to measure a podcast’s audience size. While filtering your web logs by unique IP cuts out multiple hits from one user, you still don’t know if that user received all or just part of the file. Measuring downloads that are not only unique but also complete results in a more conservative metric than measuring just unique downloads.
This raises two questions:
- Is it possible/feasible to measure complete downloads?
- Do you need to measure complete downloads?
Is it Possible/Feasible to Measure Complete Downloads?
Yes it is possible. I spoke to Dan Kuykendall, creator of Podpress and an avid web programmer, to find out how to do it. He listed two ways:
- Use a script on your server to monitor the download
- Access the complete logs and use a script to filter them
Dan also pointed out that each of these methods requires more resources than your basic stats. In the first case, every time a file was accessed for download, a script would feed the file out, track the progress of the download and mark if it was completed. This requires increased server processing resources.
In the second case the limitation is that you may or may not have access to the needed server logs. This shouldn’t be a problem for those with a dedicated server. If you do have access to your logs, one script that will count complete downloads is Podalyzer.
Do you need to measure complete downloads?
To answer this, you need to know what your objective is. What are you trying to track? Ask yourself if knowing that a download was complete tell you something important about your consumer?
If someone has downloaded the entire file then there is an assumption that he has a higher level of commitment to the content.There’s a greater likelihood that the whole episode was listened to (including any ads at the beginning, middle or end). But you still don’t know for sure.
If you just need a relative measurement of the size and growth of your audience, measuring unique downloads should be adequate. However, if you have the resources to measure complete downloads, then it would certainly be an interesting metric to compare and contrast. Some advertisers may demand the more stringent complete downloads measurement.
My personal opinion is that it is not absolutely necessary. A partial stream/download could still be someone who found your content and engaged with it. If we really want to get accurate, then let’s devise an open standard for measuring consumer behavior. Let’s track exactly what parts of a show are listened to. Let’s get them to take action and measure the response.
No matter what, numbers such as unique or complete downloads are merely a quantitative analysis of an audience. Really what is needed is something more sophisticated and qualitative to accompany the download numbers. Some have started to call this the Return on Influence (ROI) metric. Some call it "engagement". But what is that and ow do we measure it? That is an ongoing discussion for another time.
In the next installment I’ll discuss other metrics that relate to the size of a podcast audience such as bandwidth.
December 1st, 2006 at 9:37 am
Hi Jason. This post brings up an interesting point and, one I admit that drives me crazy. For some reason, many of the ad agencies I am dealing with have decided that they need only pay for “complete downloads.”
When I ask them if they only pay magazine publishers for ads when the magazine staff can prove the subscribers did a “complete read” of the magazine, they give me a blank stare.
When I ask them if they only pay television producers for ads when the TV staff can prove the viewers watched the “complete commercial,” they give me a blank stare.
I could go on, but you get the point.
For those of us who are monitizing podcasts with advertisements, we need to make sure the playing field stays level.
Magazine sales are based on Audit Bureau of Circulations numbers. They tell the advertisers how many people SUBSCRIBE to the magazine, not how many people will read it. Television and radio have even less information about their audiences. They sell ads based on a very, very small random sampling of prospective audiences via a diary system.
I see no reason to let the advertisers get away with holding us accountable for “complete downloads,” when they don’t hold our competition to the same standard.
And lastly, since many podcasters front-load their shows with ads, it doesn’t matter if the listener downloads the entire show to the advertiser. All they should care about is having their ads heard.
Sorry for the rant, but this subject hits my button. Good series.
December 1st, 2006 at 10:31 am
Scott,
Very good points. I’m glad you ranted.
Podcasters need to be sure they sell the qualitative nature of the medium and not let things get hampered by the quantitative analysis.
If an advertiser only pays you for complete downloads, then you’re probably underselling. The one case where this might not be true is with post-roll ads, but most sponsorships are mentioned early on in the show.
Thanks for the insights.
Jason
December 1st, 2006 at 10:43 am
Scott is absolutely right – do NOT let agencies and buyers hold podcasting to a higher standard than any other media.
Since most host-read ads are at the beginning of a podcast, chances are that many of the incomplete downloads are still impressions on the audience and worth every penny as much as a complete download.
April 8th, 2007 at 9:58 pm
hey i share the same last name. ther is a jason vanorden in california as well