Measuring podcasts
I’ve long contended that we need to find a way to measure audience for podcasts and vlogs so we can, if we want, attach ads and get support. If we don’t find an open way to do these things, companies such as Audible will come up with closed, expensive ways such as this:
Audible’s technology puts measurement capability in Audible-compatible devices — such as Apple Inc.’s iPod — or in the software for listening to Audible’s content. ….Audible is making its tracking service available to outside podcasters. The company will charge three cents per downloaded podcast to report whether a downloader listened, and for how long. Audible will also offer tools that will stop the podcast from being emailed to others. It will charge five cents per download to track listening and attach the access restrictions. For half a cent per download, Audible will insert an ad relevant to the podcast.
With the tools, “you can build a bona fide rate card” for advertising, says Foy Sperring, Audible’s senior vice president for strategic alliances. The company says it will provide only aggregated statistics; it won’t disclose what individual podcast downloaders listen to.
What we need is an open system that allows any content creator to get audience data pinged back and allows them to attach measurable ads. Today on the text web, these things are free unless I choose to use a premium service for stats or ad serving. We need similar functionality for the multimedia web using MP3s and not just Audible-formatted media.
Look at the Audible economics: They’re charging 3 cents just for measuring listenership. That, in ad math (if I have enough fingers and toes) is a $30 CPM just for measurement — $35 for inserting ad ad. That’s a high rate for advertising online these days — very high. So there’s no profit. That won’t work.
Mitch Ratcliffe, who helped build this, responds in the comments and on his blog. And I add there:
Mitch: What I’m really calling/hoping for is an open version of what you’ve done. I doubt the pricing is your fault but it’s both greedy and dumb, for it will limit uptake so severely as to make the effort not worth it. So what we need is an open version that allows audio and video producers to support their efforts without losing all their possible income to fees such as these and without backing a standard that cannot gain acceptance because it’s just too expensive. If Audible really wanted to play in this arena and become a new standard, they’d put this out for free and build businesses atop that.
We need the commercial MP3. Once that exists, little guys and big guys alike will rush to produce and distribute content by every means possible. P2P will be their friend.
November 11th, 2005 at 9:19 am
“I’ve long contended that we need to find a way to measure audience for podcasts and vlogs so we can, if we want, attach ads and get support.”
Are people tryint to think outside the box and not use the “old media” advertising models to support these efforts?
November 11th, 2005 at 9:25 am
And your suggestion, Michael?
Advertising is one means, not the only means, but it requires measurement to work. Basic fact of media life.
I hope that iTunes will also enable pay-per-sub or pay-per-download models but I’m not sanguine about supporting this medium on that basis.
November 11th, 2005 at 10:18 am
You put an anchor tag inside an anchor tag there, so that only via View Source do I see the link to
http://online.wsj.com/public/article/SB113167835201394489-6RLXo50JXniwqPt59a3cCUJPXsM_20061111.html?mod=blogs; I suspect that most of your readers won’t see it.
November 11th, 2005 at 10:34 am
The Google AdSense model could be applied here - display highly targetted ads based on content and advertisers can pay per action. Right now AdSense parses web page content to display ads based on text, but it’s not hard to imagine a speech recognition engine that could rotate ads to a Flash or Ajax based ad panel embedded within the media player based on audio content as it’s being played in real time.
November 11th, 2005 at 2:38 pm
My comprehensive response, as one of the people who developed the Audible product, is on my blog. But let me add a couple things:
MP3 simply can’t do what Jeff suggests it needs to. Every MP3 player would need to be replaced with devices that supported a new MP3 standard.
Displaying ads in conjunction with audio doesn’t make sense in almost every situation, as you are not looking at the device or application you are listening to. Not to mention that speech recognition is incredibly hard.
Finally, the product offering is designed to support a lot of different types of audio delivery, not just individual podcasters and not just advertising-based—that’s the fiscal reality of building software—and we are doing our best to bring the tools into the realm of every producer of content so that they all can play on a level field for the first time.
November 11th, 2005 at 2:40 pm
Jeff - I just saw their demo at the PME. One *major* point to emphasize - if you charge for your show/content, they take a very hefty 20% (or at least $1) charge for the transaction. Additionally, I asked about their service level guarantee for bandwidth/delivery and while I was assured there was one, no one in the booth could point me to the actual language.
November 11th, 2005 at 5:43 pm
Check out http://www.podtrac.com
They are here at the PME, too. Although, Audible bought all the lanyards, getting the hit on attention.
November 11th, 2005 at 6:47 pm
Jeff — [Extracting from email and Jeff's comments on my blog] I think you’re wrong about the Radio thing, though and will respond in the blog if you reply there (or I can post this, if you like). In 2001, I had 30 readers, compared to about 5500 today. My cost-per-reader back then was in the $50 range…. today it’s $0.027 per reader.
You’re fixating on MP3. It’s just an audio codec and not a particularly capable one—assuming everything has to be MP3 to succeed is locking the industry into something comparable to AM radio. Audible’s .aa file can be delivered via P2P, too. So we could do everything you’re describing wanting to do and with greater distribution reach than MP3 provides (since Audible’s format is supported, with wireless delivery, on Windows Mobile, Nokia/Symbian phones and Palm devices, too).
So, I think we’re all on the same page. The only problem is that Audible can’t go broke revolutionizing the market; it can, however, learn from this kind of feedback and will.
November 11th, 2005 at 8:10 pm
From the same email exchange, which Mitch and I agreed is best discussed online, I said:
Tom Evslin and Tim O’Reilly said it better than I did at Fred Wilson’s and Brad Burnham’s Union Square Ventures event recently: first build the network. Take as little as you can at first to grow as big as you can. That’s the opposite of this trajectory: Make money from the first and you will grow slowly. And in this world, slow growth can be the same as no growth if you don’t get to critical mass first. I’m dying — dying — for the means to commercialize audio and video for those who want to. But a $35 cpm as the starting point won’t scale. And, sorry, but I don’t buy the Radio analogy. I negotiated my employer’s investment in Pyra.
Three cents just to measure one listen is in any measurement excessive. I’ll bet that if you found a way to reduce that — while keeping the ad-serving piece — you’d grow a heckuva lot faster.
And the only ones who can use this are the big companies. That’s not where the real growth is, as we all know. Making this work on the, dare I saw it?, tail is where the size is.
November 12th, 2005 at 12:55 am
A system already exists that works with the MP3 format and its limitations. And it’s by a brand advertisers trust.
See: http://vocuspr.vocus.com/VocusPR30/DotNet/Newsroom/Query.aspx?SiteName=arbitron&Entity=PRAsset&SF_PRAsset_PRAssetID_EQ=102232&XSL=PressRelease&Cache=True
November 12th, 2005 at 2:53 am
Well, no that’s not the same thing. The portable people meter is a device carried by an Arbitron panel member to record their alternative media use. That would be extrapolated to project listenership for the whole population, as explained by the passage “Carried throughout the day by randomly selected survey participants, the PPM device can track when and where they watch television, listen to radio as well as how they interact with other forms of media and entertainment.”
If you need a device that listens for the inaudible codes it could be reduced to be included in hardware and offered as a software component for use in players, but it would take years to reach market penetration anywhere near close to Audible’s. Meanwhile, the technology adds no benefits for the user, which Audible’s does, including the ability to start where you last listened across multiple devices, bookmarking, etc.
You’d also have to distribute tools for producers in order to inject the codes into audio. This is a solution that’s miles and miles from commercial deployment.
Very different than “exists” in any commercial sense, but don’t take my word for it. Check it out and tell me if I am wrong.
May 31st, 2007 at 8:45 pm
[...] to say I told you so: Audible is killing its Wordcast proprietary podcast distribution and measurement [...]