Can You Trust That “Billion” Download Figure?
A while back, Libsyn issued a press release claiming it had delivered one billion downloads. Is that figure real? It’s not much of a secret in the industry that Libsyn’s stats have been suspect for quite a while.
Now a news item appears in Podcasting News saying that Libsyn is “updating” podcasting stats.
Libsyn admits that they’re just now fixing problems that occurred as far back as July of last year. They also admit they are “toning down” their stat package updates - i.e., doing fewer of them. What? They know they have a problem, announce the fact that they are working on it, but then say that they can’t work on it too fast?
So if that’s the case, how accurate are any of their stats claims, especially the one billion download mark?
Libsyn employee Rob Walch used to routinely call the rest of us to task when we made such errors, often intimating that there was some conspiracy afoot. I wonder if he’ll be as vocal about his own employer’s foibles?
Stats are tough. No getting around it. And just like in radio, television and print media, verifying the stats is ALWAYS going to be tough. The lesson here is that we all need to be suspect of the numbers we hear. It’s time to realize that we won’t ever get this completely right.
The other media have figured this out and don’t worry about it. They just agree on a standard, right or wrong, that everyone abides by. While I used this post to poke a little fun at Rob and Libsyn, the larger point is that we can’t let ourselves get roped into providing stats that our competition doesn’t have to.
I’ll finish this post where I started it. Come on Libsyn - fix your problem please. But more importantly, consider blowing your own horn less often until you do. And don’t be so quick to worry about stats that our competitors can’t or won’t provide.
Popularity: 38% [?]
Scott Bourne is an author, lecturer, teacher and new media pioneer, and the president of Podango Productions in San Francisco, CA. He is the founder of the world's first Internet-only radio network, Netradio, and is a co-host on several prominent podcasts including MacBreak Weekly, Apple Phone Show, the iLifeZone and GMT. Bourne has also been a featured speaker at events such as Macworld Expo and the Podcast & Portable Media Expo.
Scott
Should we trust Wizzard’s numbers? For that matter - should we trust Nielsen’s numbers or the circulation numbers for your newspaper?
In Elisabeth’s interview with Wizzard CEO Chris Spencer, they dig into the subject of how they count podcasts. Spencer says that Wizzard tries to err on the conservative side. Spencer also says that podcast downloads really don’t tell you anything about podcast consumption.
That said, marketers shouldn’t rely on someone else’s numbers - they should rely on their own.
No James we shouldn’t trust anyone’s numbers. That’s the point I am making. It doesn’t even matter what those numbers are to the marketer. The marketer should care only about cost to acquire a new client.
Until we measure our product in that manner, we’re talking about the wrong things.
And I agree that downloads don’t tell you anything about consumption - just as Audited Bureau of Circulation numbers don’t tell you about consumption of magazine article, or Arbitron numbers don’t tell you about consumption of radio. And that’s the larger issue. Why should we be forced to provide numbers our competition won’t?
And Libsyn is in the spotlight because they were out there bragging about downloads. They brought it up first. It’s hypocritical of Spencer to say on the one hand - “Hey look at us - we got 1 billion downloads,” and on the other “downloads don’t really tell you about podcast consumption.”
And taken further - it looks bad for the industry to have a leading company bragging about its downloads while a few weeks later saying it has problems with its stats package.
Scott
While you bring up some interesting points, I think your logic is a bit flawed. While on the surface most of that seems true, those who have studied media planning or stats might argue.
As Chris Spencer said, he is trying to show the traditional media buyers that there are numbers behind podcasting; yes, they are not perfect and don’t always reflect consumption, but this is what buyers want to see (and is the norm of trad’l media).
I don’t think it’s hypocritical for Wizzard to say “We have 1 billion unique download requests this year, versus only XX,000 last year” because the change in numbers DOES represent a shift. Media buyers and planners understand that numbers must be taken with a grain of salt, but nonetheless if you’ve ever sat in the room with Fortune 500 advertisers, these numbers have meaning. In fact, these types of stats and research are required.
Numbers do hold meaning, if considered in a metric sense against various parties. As for ‘the numbers say anything about consumption’, that’s why there is quantitative and qualitative research; any media person will tell you that no one element holds the power.
Also,Libsyn numbers and Wizzard numbers are two different things anyway. Libsyn’s stat machine has been a pain in podcaster’s rear because it requires back end work to get the right numbers; it wasn’t working in the front end. SO the numbers were there, then for safety measures rounded down, and on top of that they brought in Nielsen.
As for Libsyn and Wizzard, I think you’re drawing lines that are not there, in terms of why their stats would be flawed. From what I understand they are slowing the system for more authentic numbers instead of frequent flawed ones. I do agree there should be some clarifications because right now it looks like Wizzard is pulling those numbers from a flawed system.
John you\’re certainly entitled to your opinion - but please don\’\'t change the point I am trying to make. I said Libsyn was hypocritical to brag about lots of downloads and then to say but that doesn\’t really matter because they aren\’\'t a good measurement anyway.
As for studying media planning and stats - I\’ve not only studied them - I\’ve been involved in buying and selling from them for three decades.
The stats that are provided to traditional media buyers go nowhere near the level of information that some podcasters are asked to provide. I for one won\’t cave in on that unless I receive additional compensation.
The broader point isn\’t really whether or not Libsyn/Wizzard stats are accurate. We know they are not. Neither are Arbitron\’s stats, etc. But there is a standard and once agreed upon, everyone plays along.
By Libsyn/Wizzard talking out both sides of their mouth, they make it harder for everyone else to play.
I think you are both hitting a sensitive nerve. I just finsihed a big campaign with Wizzard and we have our own tracking to compare with but we got paid using their numbers. The numbers were WAY off and in fact we didn’t get paid what we first thought we were. They did in fact use Nielsen but they also used their own via Libsyn.
We have had so many problems with Libsyn that I’m looking for another media agency to work with. We sit on an insane amount of inventory so tracking is very very important for us. Underpromise and over deliver is my motto.
[…] The question is what will Wizzard do with Libsyn over the long haul? Last week some controversy erupted around the 1 Billion download claim that Wizzard/Libsyn made. Scott Bourne pointed out that a press release from Wizzard last month stating that 1 Billion podcasts downloads had been achieved in 2007 was followed up just last week by a statement that Libsyn was “fixing” issues with its stats engine. Frankly the total number is irrelevant, whatever it is, it is big, but once again the stats issue raises its ugly head. For true advertiser support/adoption of podcasting, stats need to be quantified and agreed upon. […]
[…] The question is what will Wizzard do with Libsyn over the long haul? Last week some controversy erupted around the 1 Billion download claim that Wizzard/Libsyn made. Scott Bourne pointed out that a press release from Wizzard last month stating that 1 Billion podcasts downloads had been achieved in 2007 was followed up just last week by a statement that Libsyn was “fixing” issues with its stats engine. Frankly the total number is irrelevant, whatever it is, it is big, but once again the stats issue raises its ugly head. For true advertiser support/adoption of podcasting, stats need to be quantified and agreed upon. […]
[…] Wizzard published that they supported over one billion downloads last year from the over 8,500 podcasts that use their hosting services. Some have questioned how that could be, so we ask Chris to help us verify those numbers and to understand how 8,500 podcasts produce so many downloads. Chris explains how they count downloads and filter out the spiders and bots. He says that there are some blockbuster podcasts that do really high volumes and just as in other media forms there is a long tail of podcasters, so looking at mean based average per podcast just does not make sense. It’s the old 80-20 rule once again. Chris also clarifies that some recent problems with reporting to podcasters has nothing to do with how the count their download figure. […]