As a podcaster, I could tell you exactly how many copies of each edition of For Immediate Release: The Hobson & Holtz Report are downloaded. I can also tell you which podcatcher or distribution services downloaders use. I can even tell you whether downloads are direct or via RSS subscriptions. Plus quite a few other useful facts.
These are the types of fundamental stats Shel and I get from Libsyn, the service we use to host our show’s MP3 files. For example, those stats currently say that the average total number of downloads for each edition (the “average audience” as Libsysn describes it) is 570. So that’s 1,140 downloads per week. The most popular way to get the show is iTunes according to the stats, currently accounting for roughly 75% on average of all downloads.
But there’s one significant statistic we can’t yet determine – how many people actually listen to each show. We also have a link on each show notes post where you can click and listen there and then rather than download the MP3 file. And as Shel and I cross-post each show’s show notes to our own respective blogs, that’s three different places where you can click-and-listen. No way to measure that, as far as I know.
This is a problem every podcaster has. Downloads of the MP3s is one thing. Who listens, how many listeners there are and where they are, is another. Yet that’s the most important statistic of all to aid you in creating and tailoring the content in your show to your listener demographic.
Via Techcrunch comes news about Podbridge, a service that aims to connect podcasters with advertisers and serve as an advertising network for audio advertising in podcasts. That wouldn’t have grabbed my attention were it not for this text in Podbridge’s service description:
Accurate measurement of podcast listens – not just downloads … Audience profiles with demographics, listening habits and locations
My feeling about including advertising in our show is a) I’m not sure how our listeners would feel about it, and b) I’m not sure how I feel about it either. Note that Shel and I haven’t discussed the notion of advertising at all.
Yet I’m drawn by Podbridge’s claims to provide podcasters with accurate and valuable information about listeners to a podcast:
[…] Podbridge technologies reach through the desktop to measure behavior right at the portable device. Your Podscore measures reach, frequency, time spent listening and more – concrete data advertisers demand. Prove who listens to your podcast, not just who downloads it. Know them by gender, age, geographic location and more – ten metrics you can use to fine-tune your content, prove your performance and build your advertising revenue.
Techcrunch explains clearly how Podbridge’s service works:
[…] The service for publishers and advertisers they have recently launched allows them to track not just how many times their podcast has been downloaded, but how many times it has actually been listened to. It does this through software on the client side that the listener must install once the first time they listen to a Podbridge-wrapped podcast. This has long been a long problem for both parties and previously advertising in podcasts has been sold on estimated numbers based on the number of downloads.
How Podbridge works is the publisher wraps their feed in Podbridge code that will allow it to sit in between the publisher and the listener. From the listeners perspective when they download the Podbridge add-in the first time they will be asked a few questions so that advertising can be better tailored to them (how users will react to this I am not sure).
I’m not sure about that last point, either. I think many people (me included) will need some major assurances regarding exactly what data Podbridge will capture, how it will be used and how they treat the confidentiality of that information. Plus overcome strong resistance by some people to installing such software on their computers. I certainly wouldn’t want to use a service that resulted in losing listeners to my podcast.
Still, I want to find out more about Podbridge so I’ve requested a demo. Once I’ve seen that in action, I’ll discuss some thoughts about it, and the overall concept of podcast advertising, in an upcoming edition of FIR.
Meanwhile, especially if you’re a listener to FIR, what do you think of this listener-measurement and advertising concept?
12 responses to “Measuring listeners and podcast advertising”
I know I am missing altogether too much good blogging and podcasting out there — and plenty of print titles — to enjoy going over a hurdle for content. If I did not know you and Shel, know your integrity, and know your content, there’s little or no chance I would install the thing just to try the podcast.
I’m betting I’m not alone.
As for the ads, I’d wonder why you do it, unless someone is going to pay big bucks. You and Shel have raised your professional profiles tremendously, which should be paying off in consulting gigs that outstrip any ad revenues you might get. I don’t have near the readership you two do, yet I’ve made enough paid speeches this year alone to cover hosting costs for the next 20-30 years. Why mess with your damned successful format?
All that said — you two could run banner ads, pop-overs, pop-unders, and sell my name to the Nigerian scammers and I’d still show up.
Allan
Thanks, Allan, I really appreciate your comments.
I’m not focused on podcast advertising at all, certainly not for our show. I am very interested, though, in finding a way to measure listenership. So far, I’ve not come across anything as interesting-looking as Podbridge in this context.
I don’t yet know enough about Podbridge, how advertising would work, would it be obtrusive, etc. If some kind of ad arrangement was the only way to get meaningful listener stats, would we participate in any programme? Again, don’t know yet.
My interest is specifically to do with knowing who our listeners are and how we can find ways of providing the content that listeners want to hear. That would likely increase our coverage, which is something I am definitely keen on doing.
So I’m keeping an open mind about the Podbridge concept until I’ve seen a demo and can get a fuller understanding of what Podbridge is about.
Stay tuned, so to speak.
Neville,
It’s that same age old question in any advertising – How can we tell the message gets there? Newspapers have been attempting to find out exact numbers of who actually reads the ads. They know how many people buy the paper, but can only guess at who actually reads them.
The only way I would put any value in actual readership (including listeners) is to have them respond directly in some way – ask for information, buy, enter the contest, etc. Even then you won’t get exact numbers as some people forget to respond later, choose not to, etc. Web sites are offering some very good advancements in that area, as we can now tell when someone clicks a link, how long they read for, where they jump off from etc.
Thought that brings me to an interesting programming point – it should be possible for the java code you use to record start, finish, fast forward, replay, jump points on the podcast. The way I assume it works it pulls the feed from the server, so that information is already part of the code. It *should* be trivial to log it… Can’t do that with itunes, realaudio etc (yet).
For me, I would avoid podbridge – it is invasive, requiring me to download and install code (read adware) and offers me no advantage other than to be advertised to more… And I am strange in that I choose to be advertised to often…
You make a good comparison, Michael, with newspapers and readership. I agree that getting an exact number is not feasible. Yet we do want to know who actually listens to our show in a more precise way than we currently know, which is by people spontaneously telling us they listen. For instance, if the FIR Frappr community is a good measure, then we know we have at least 135 listeners as of today.
There’s a gap, though – 135 listeners versus 570 average downloads. Does that mean 435 episodes are downloaded but no one listens to them? I don’t believe that, but there’s the issue: I don’t know. Or are there multipole listeners to a partucular show? Do people copy the MP3s to friends and colleagues? Unknown.
Is a service like Podbridge the way to go in order to more precisely measure listenership? I’m hesitant because of the advertising connection plus, as you point out, the notion of installing some software. I don’t like that idea at all (frankly, I doubt I would do it). So there would need to be something pretty compelling about Podbridge to persuade listeners to install their software.
But this is still just an idea at the moment. I need to know more about Podbridge and I hope to have a demo sometime soon.
I suspect there are a number of people like me who subscribe and download every episode but only have time to listen to some. When I get more than about three episodes behind I dump the older ones and start from the most recent.
That’s what I do with podcasts too, Stuart – usually miss listening to ones I’ve downloaded so move on to later episoides when I have time. My iPod usually has around 20 podcasts on it and I rarely get a chance to listen to even 5 or 6 of them.
Especially the lengthy ones ;)
My initial reaction as a podcast listener — and a loyal FIR one, at that — is to say that I’m definitely not crazy about the idea of having to download more software (level of invasiveness still unknown, it seems) just so that the podcaster can know whether or not I actually “tune in” to that show once I’ve downloaded it. It’s not a step I — and many others, I would suspect — would be willing to take.
And even if I did download Podbridge, I can’t imagine it would be able to measure when I listen to a podcast on my iRiver while riding on the train or taking a walk.
Still, as you note, Neville, it’s certainly an interesting concept. It’s the quest for that holy-grail piece of information — am I listening?
I’ll be watching for your report after you’ve used the demo.
The difficulties of measuring podcast audiences…
Measuring listeners and podcast advertising at NevilleHobson.com. But there’s one significant statistic we can’t yet determine – how many people actually listen to each show. We also have a link on each show notes post where you can click and liste…
It’s installing the software that perturbs me most (and everyone else by the look of it). Yet as I understand it so far, that’s integral to how Podbridge works its measurement service.
Shel and I will be talking about this in today’s show.
[…] Neville Hobson makes an interesting point: “But there’s one significant statistic we can’t yet determine – how many people actually listen to each show. We also have a link on each show notes post where you can click and listen there and then rather than download the MP3 file. And as Shel and I cross-post each show’s show notes to our own respective blogs, that’s three different places where you can click-and-listen. No way to measure that, as far as I know. This is a problem every podcaster has. Downloads of the MP3s is one thing. Who listens, how many listeners there are and where they are, is another. Yet that’s the most important statistic of all to aid you in creating and tailoring the content in your show to your listener demographic.” […]
[…] 89:36 Sebastian Keil’s view on Edelman/Wal-Mart are that if you work for a PR-agency you should disclose that in a way that is not ambiguous; some thoughts about Podbridge; and asks what happened to the FIR listener survey we talked about before? (50 shows ago, says Sebastian) […]
[…] This post was triggered by Neville Hobson and later this Britcaster discussion. […]