« The Long Tail of Open Source | Main | Mobitopian Gift-giving »
January 21, 2005
Revisited - The Long Tail of Open Source
Wolfgang commented on my post and his valid point got me thinking.
The economic model applied to Amazon or iTunes Music store is that (because there is no need for shelf space) it pays off to offer access to basically every song or book ever written. Even if the audience of each niche title is only very small. The collective sum of those small audiences is larger than the mainstream group that goes for the hits, so it becomes economically feasible. However, this is only possible because of the "storefronts" Amazon and iTunes providing access to both the mainstream part and the long tail.
But I think there's more to it than just the economics. The original example in Chris Anderson's article talks about an obscure book becoming popular because of reader recommendations.
What happened? In short, Amazon.com recommendations. The online bookseller's software noted patterns in buying behavior and suggested that readers who liked Into Thin Air would also like Touching the Void. People took the suggestion, agreed wholeheartedly, wrote rhapsodic reviews. More sales, more algorithm-fueled recommendations, and the positive feedback loop kicked in.
A first step - when applying this to Open Source would be to create a similar storefront for the access to Open Source projects. Sourceforge is already part of that storefront, but it doesn't include any of the projects under the Apache umbrella or in the other Open Source repositories. The unified "storefront" would also need to have the same level of participation that users give back to Amazon or iTunes by providing their personal feedback in form of wishlists, comments, playlists etc. "Readers who bought this book also bought that book". Could this be applied to Open Source? "Users of this project downloaded that one", "Developers on this project are also working on that project", "User Max releases his top ten Open Source all time favorite project list", "Here are the last 10 endorsements for developer Peter" etc. Sourceforge already has "most active" and "top downloads".
This would help increase the visibility of Open Source projects that - to date - may have received little attention. If I were looking for an Open Source based CMS with integrated BPM for say a financial application, then it would be great if I could see whether there were other people doing exactly that and what Open Source projects they were using.
And maybe some of those projects that are currently hidden in the Long Tail would become popular, attract developers and users and become the JBoss or MySQL of tomorrow.
Posted by Matthew at January 21, 2005 12:13 PM
Comments
You're thinking what I'm thinking you're thinking! :-)
I was about to answer Wolfgang that distributtion == awareness+reputation+QA here, but that you still need a "distribution channel".
Also, that his view is (too) supply oriented "... it pays off to write a book ...", while, for instance, John Kennedy Toole, author of "A conspiracy of dunces" committed suicide before seeing his novel published, and his mother had to keep tryng before it got published and won the Pulitzer. This is a good example of the problem looked at from the offer side. The books are there, just like the software projects.
Posted by: Santiago Gala at January 21, 2005 01:47 PM
"A confederacy of dunces", of course. The point is that "distribution" of non-scarce/cheap-to-copy goods is significantly different from distribution of scarce goods. (The internet as a fractal shelve-space)
Posted by: Santiago Gala at January 21, 2005 01:53 PM
@Santiago: Good point, Amazon can even deliver books which faild to become bestsellers (and have never been intended to be niche products). But as soon as there is a global long tail channel, publishers might consider producing for small markets (especially e-books or MP3s, where you do have a low entry barrier).
@Metthew: I agree with you that there are interesting parallels between distributing digital media and open source software ("pareto distribution").
But there are also major differences. One might be, that I don't have to care about how many other people bought a certain book or song. It's enough if I like it (technical books are a different story). When evaluating an OSS project however, the size of the community (including the number of users) is very important. Open Source is subject to power laws (see slides from the O'Reilly conference: http://conferences.oreillynet.com/presentations/os2004/conklin_megan.ppt), where the rich tend to get richer. So as a user you probably jump onto the bandwagon of a successful project, even if a small one better meets your demand. In effect, when choosing free software, in most cases you would shy away from the long tail and prefer the hits - its a safer bet (much like "nobody has ever been fired for buying IBM").
P.S. I tried to format my comment, but all HTML tags got filtered out :-(
Posted by: Wolfgang Sommergut at January 21, 2005 10:23 PM