Insert Ads into Podcasts with Fruitcast

Trackbacks/Pings (Trackback URL)

1. James Archer
   

   November 4th, 2005 at 12:04 am

   Hey, Michael, thanks for the coverage!

   Podcasting, by its very nature, does involve higher CPM than radio would, simply because cost (bandwidth) goes up with every listener, which isn’t the case with radio.

   It’s certainly not something that would form the core of any major campaign, but the highly targeted nature of podcasting, as well as the non-casual audience, makes it a great candidate for at least a portion of a company’s marketing dollars.

   In any case, the option is out there now!

2. Todd Cochrane
   

   November 4th, 2005 at 12:17 am

   If they are actually able to deliver that type of CPM, I want it in writing no one has been able to secure that high of a per 1000 listeners. I have talked with at least 25 different mainstream companies and no one is willing to pay that much.

3. Michael Arrington
   

   November 4th, 2005 at 12:19 am

   James,

   What does the cost have to do with the CPM? I mean, I understand supply curves and how cost affects them, but that implies that there is a very limited amount of inventory for advertisers. Do you feel that this is the case, and that podcasting can charge a premium due to the high marginal cost of a podcast relative to radio?

   In my opinion the argument to make to advertisers is that a specific podcast is highly targeted, more so than a radio program. They’ll argue that many people download but don’t listen. You can counter that at least you have some method of determining listeners (downloads) v. radio, which is all BS…But $250 is very, very rich.

4. James Archer
   

   November 4th, 2005 at 12:28 am

   Michael, I think you’re probably right, and that example might turn out to be too high and it’s likely that we’ll revise it to something more realistic.

   I definitely agree, though, that the value of podcast advertising isn’t in quantity, it’s in quality. It doesn’t matter if 100,000 people hear your message if only six of them are interested. With podcasting, you can probably pay for ten and have six of them interested.

5. Michael Arrington
   

   November 4th, 2005 at 1:02 am

   James, the auctioning process that I see described at solutionwatch is interesting.

6. Jon
   

   November 4th, 2005 at 2:15 am

   Hate to sound like a pessimist (realist?) but I really don’t see them being able to maintain a ~$250 CPM.

7. Mike Rundle
   

   November 4th, 2005 at 7:45 am

   Damn that CSS is clean!

8. Mark
   

   November 4th, 2005 at 8:26 am

   I’ve worked in the online advertising business for more than 10 years and it is clear a number of things must occur for this to work:

   1 An effective sales force that has access to large media buyers to drum up business. Forget the adsense inventory auction approach unless you have tons of podcasts identified by psycho/demo groups and multiple keyword filtering that can be easily targeted.

   2. On the heals of that: Fruitcase needs a way to target specific demographic segments (by podcast tags?) How does the sporting goods seller know that “Joe” is reaching buyers of their products? What data is available to media buyers to make that decision? I couldn’t find any such capability listed on their site.

   3. There will need to be enough people downloading the ads to make it attractive to advertisers and financially viable (at least 2-5MM on a weekly basis per podcast) How many podcasts actually deliver that quantity on a consistent basis yet? One of the biggest fallacies in the online world has been that “all you have to do” is pull together the inventory from a lot of small sites(or, in this case,podcasts) and that you can sell them all at a great CPM because they are so “targeted.” The reality is that major advertisers (where the REAL money is) don’t want to buy that way. They may some day, when there is simple, tangible, verifiable ways to do so, but today (and for the next several years) they want to buy large, targeted audinces in single chunks. Ad inventory aggregators end up usually selling most of the inventory they represent at extremely low CPMs or even on a CPC performance basis . . but by performance, this means that a user actually clicks on the ad and goes to the advertiser’s site, not that they were just exposed to it. The only way ad inventory aggregators have been able to justify higher CPMs has been their increased ability to profile, monitor and serve targeted ads to specific users through the use of complex cookie and ping back systems, which Fruitcast doesn’t offer.

   4. Are the podcasts monitored for content accuracy and appropriateness? Will Gillette really want to have to review literally hundreds of podcasts to make sure the quality and opinions expressed and language used, etc., mesh with their desired image? No media buyer or account exec will take the time to perform due diligence on dozens of micro sites . . . it would cost more to do that than any potential benefit the low volume business the ads could offer. Its one thing to buy sponsored text links (without any true product/brand representation, logo. etc. )that are served peripherally next to blog content, and quite another for a company/agency to provide finished audio commercials integrated with audio content generated by who knows who. General Motors can live with a randomly delivered sponsored link for their new GMC truck models served somewhere on the page of an individual’s blog that focuses on construction trends . . . it would be difficult to convince GM not to worry about their painstakingly created audio commercials being played as a billboard to a poorly produced podcast. (Or to dozens of such amateur podcasts)

   4. They will need some sort of tracking mechanism to demonstrate effectiveness, which is VERY difficult with what, in essence, is an audio file that has been downloaded. Is there third party verification of whether the ad was ever actually listened to, or how that has affected user behavior/awareness/action? Online advertisers have come to expect extremely detailed performance data and will demand it here. I don’t think media buyers are going to go backwards and accept mere “impressions” a la 1970s radio (with this model you can’t even identify cumes).

   5. CPMs will have to come way down. Face it, the most expensive CPM for mass market media today would be a MAX of $150 CPM for HIGHLY qualified direct (physical) mail lists. How could any media buyer justify spending $240 CPM for an unproven medium that has no immediate consumer response capability?

   I applaud Fruitcast for being “the first on the block” with a technical solution to inserting ads into podcasts, but the needs of the ad market far exceed that straightforward technical implementation. It needs critical mass audience(s) size that is independently verifiable, highly targetable and that has meaningful performance/tracking data. Oh, and it has to be demonstrably competitively priced.

9. joshua strebel
   

   November 4th, 2005 at 9:32 am

   Good job James, Glad to see our little conversation over lunch came to fruition. Wow.. what a pun. knock em dead.

10. Brian Breslin
    

    November 4th, 2005 at 11:45 am

    why???? why??? i had this idea 5 months ago, and even started working on some stuff. we even got a simple script to insert ads in the beginning and ends of podcasts. ah well. i never thought $250 CPM would be approachable. I figured the rates would average more around $100/CPM (giving podcasters something like $.05 per listener).

11. Michael Arrington
    

    November 4th, 2005 at 11:49 am

    most ideas are a dime a dozen. Execution is what creates value.

12. James Archer
    

    November 4th, 2005 at 12:19 pm

    Mark,

    Thanks for the great feedback and advice. We’ll certainly take it to heart.

    As far as the great CPM debate goes, I want to reiterate that we’re not necessarily shooting for $250 CPM. The example we gave on our site assumed two ads (one at the beginning and one at the end of the podcast), and even that was probably a bit high so we’ve since lowered the number in our example.

    Right now the current minimum bid per download is $0.10. Ten cents a head isn’t bad for a medium in which you’ve got a relatively dedicated and captive audience, and you’re able to pretty closely target the demographic you want by picking the appropriate podcast.

    There are definitely some issues we still need to resolve, but I don’t think that should come as a surprise. It’s new, and it’ll take a while for the community to discuss these issues and settle into a groove.

    Podcast advertising had to come around eventually, and (as Michael mentions above) we simply executed on it. The numbers may change, the methods may change, the technology may change — but Fruitcast is here to help make life easy for podcasters and advertisers, and we’re going to be around for quite a while to make sure it happens.

13. Mark
    

    November 4th, 2005 at 3:16 pm

    James — You’re right . . . this was inevitable, and being in on the game this early positions you well. If you can stay ahead of your competition with innovative technologies, and IF you can PROVE a viable business model for ADVERTISERS, you will at least be able to sell the company in a year or so for a nice nest egg, if not dominate the world

    But remember, right now online advertisers are only used to paying a $.10 (or $.05, or $.50, whatever)”fee” for a specific ad impression that GENERATES A CLICK to their site/product, and NOT for a mere impression. They are used to paying Google ten cents when someone clicks their ad and ends up on their site. They are NOT used to buying what they will see as online advertising on a “per impression” basis unless that translates into a competitive CPM. By charging ten cents per individual audio ad delivered/downloaded (with the assumption that every ad downloaded will be listened to), you are essentially talking about a $100 CPM (to start) for an impression that delivers awareness ONLY and, at this point, that CANNOT deliver a direct connection between the targeted user and the advertiser. That’s mighty steep. I have no doubt you will have some advertisers who will be willing to run test campaigns, but how are they going to assess the success of such a trial? How can they tell if the money spent resulted in ANY consumer action? At least with traditional radio a local advertiser can run a test with a manageable budget and will know very quickly if it results in increased store traffic or phone calls. With such a small audience spread across the nation (world!) no advertiser will be able to point to a podcast audio ad and justify the spending, at least not that that kind of CPM. Can you deliver targeted ads by geographic region?

    This is such a new area that the problem you’ll face with media buyers is essentially the opposite of the problem online advertising faced when it first started. Back then, there were no media buyers that specialized or deeply understood online advertising, and it was hard getting them to make the leap. Today, you will be asking online media buyers — who have become VERY performance/tracking data driven with sophisticated analysis tools — to essentially adopt the old school mentality of buying traditional radio spots. Only you don’t have ANY way to demonstrate results.

    Also, ad sales people and media buyers need to be able to instantly understand and relate to any cost model being proposed to them. They understand mass CPM. They understand cost per click. You don’t want to come across as trying to be both, which you kinda do when you say it’s only ten cents per impression. . . it sounds like you are simply trying to hide a very high CPM. In their minds, media buyers will put you in one category or the other, and since there is NO performance component in your model, they will see you as a straight CPM buy. You can try to explain the difference (which I don’t really buy into), but if you’ve ever tried to pitch a new ad unit/model/distribution channel to media buyers, you’ll find it is next to impossible. They have multiple clients and large budgets and multiple channels to understand and manage . . . if they can’t make your product/service fit the way they buy, it will be easier for them to simply not use you, particularly when you are talking about such a relatively tiny audience. It simply won’t be worth their time.

    I remember a month or so ago reading that either Arbitron or Nielsen had successfully tested a device that their test panel members would attach to their MP3 player that would tranmit/record when specific audio ads were listened to (I think they add an inaudible tone to the commercial that the recording device picks up). You might want to look into this and see if you can’t become a test bed for their technology roll out. This would at least give you a degree of third party validation of impressions actually listened to. It still doesn’t overcome the lack of being able to actually link the user wtih the advertiser and you’ll still find you ‘ll need to lower your CPM dramatically, but it is a start.

    The heady days when advertisers were willing to throw money at emerging online advertising approaches to see what happens is over. Unless there is a way that they can directly prove a measureable and cost-effective impact by podcast audio ads, they won’t use them.

14. James Archer
    

    November 4th, 2005 at 4:57 pm

    Thanks again for the great feedback, Mark. I never turn down advice from someone who knows what they’re talking about.

    You may well be right that $100CPM is too high for this medium overall, and if that’s the case we’ll certainly look at changing those numbers.

    Advertising always finds a way. The medium is new and growing rapidly, so it’ll take time for the natural balance between advertisers and publishers, but balance will inevitably be found. There’s a sweet spot, we just have to stumble around a bit until we find it.

    It’s a good medium (and getting better), and we’re excited to be participating in it during these early stages.

15. Cary Stuart
    

    November 4th, 2005 at 5:14 pm

    You know what I am seeing from reading James’ comments above? This is a company that really cares about us [podcasters], that wants to help us and is obviously, willing to bend over backwards to do so. That right there means a whole lot.

    Let’s give them a chance and see what they can do. I for one, as a podcaster, am grateful to Fruitcast and I’m looking forward to watching them grow and expand.

16. Michael Arrington
    

    November 4th, 2005 at 5:16 pm

    Cary, I agree totally and was thinking about writing the same thing. Rarely have I seen a company take criticism so well. It is a very, very good sign.

    James could have actually defended himself in a number of ways, such as “hey, it’s an auction system”, or “it was just an example on our website, we never said people could expect this”, etc. He didn’t do that, he just said thanks to everyone for the input. Nicely done.

17. Chris Christensen
    

    November 16th, 2005 at 12:45 pm

    While they have quoted rates as high as $.25 it appears from my experience now that rates are more like $.05. But that beats nothin’.

18. Aditya
    

    May 24th, 2006 at 8:30 pm

    Fruitcast has been down for quite a few days now and the feeds are not working either. does anyone know if they are still in business or have they packed their bags?

19. Hargreaves
    

    May 31st, 2006 at 4:57 pm

    They’ve shut down for the summer. They need more bandwidth and money, I think!

20. James from Fruitcast
    

    June 20th, 2006 at 10:24 pm

    Molly (one of the fine podcasters that has been using Fruitcast) speaks truth. We’re on a bit of a planned hiatus while we get some things organized and ready another go.

    We hit the market early, and instead of floundering for a few years and then joining the deadpool, we’re simply pacing ourselves. No reason to burn cash just because it’s the trendy Web 2.0 thing to do.

    (If anyone wants to throw some money and bandwidth at us, though, we certainly wouldn’t complain…!)