The Forces of Long Tail Software and ServicesChris Anderson’s book

November 29, 2007 – 5:00 am

The Forces of Long Tail Software and Services

Chris Anderson’s book on “the business of selling more of less” has enlightened people’s thinking about current market economics where modern day technology and the Internet has enabled long tail ecommerce to flourish. For this blog post, I thought it would be interesting to examine how the forces in his book can be applied to the long tail software market.

As Chris mentioned in his book, the long tail market is really about niches and therefore long tail software are not the top selling applications.  Niche software could range from specialized business applications that are interesting to only a small number of audience to simple online social games for connecting with people. For example, the event management software your kids’ soccer club uses to manage their league matches; and the custom project management application the construction company who remodeled your kitchen uses to schedule their suppliers and sub-contractors.

Niche software applications have always existed since the early days of personal computers. Many of these existed as DOS and Windows freeware/shareware that are posted at online computer bulletin boards. Some of these applications are developed by PC enthusiasts who have regular day jobs but spend their evenings writing small applications for fun and extra cash. Others are written by “micro-ISVs” who hope to make it big when their applications pop into the top 100 chart some day.

Those were the days of niche software - it didn’t have the three forces of Chris’s long tail economics to ride on. Today, the creator, buyer and seller of long tail software are all better off because of the three liberating forces, and a 4th emerging one that I have added to the list. Let’s examine them one by one.

 

image

Force #1: Democratize production

Whether it’s building a business, consumer or social application, the tools for developing software applications have come a long way. RAD tools such as Ruby on Rails and Microsoft Access 2007 enable data-driven applications to be created without writing a lot of code. Furthermore, hosted application builders from startups like Coghead and Longjump provide hosted tooling and application runtime environment so that application developers can be productive immediately and do not have to worry about deploying their applications. Using mashups platform like Popfly, Web enthusiasts can easily create Web applications that overlay data and services from multiple sources. The current version of Sharepoint designer simplifies the tasks of building composite applications by producing application templates and Web part components that can be composed with other software modules to create custom business applications. In addition, Sharepoint designer’s integration with Office Live enables custom business applications for small medium businesses to be designed, built and deployed quickly. No more fussing with the complexity of coding and deploying three-tier Web applications. (For those interested in diving deeper, here’s more videos and demonstration on how to build custom applications for Office Live: http://www.innovateonmicrosoftofficelive.com/LearnBuild.aspx)

This point should need no further elaboration - the improvement in application development tools is key to the wide selection of software choices we have today, and the tools are getting even better as we read on.

 

Force #2: Democratize distribution

Compared to physical goods, software has some advantages when it comes to storage and distribution - thanks to it’s digital form. In many ways, software has as “birth right” the capability to be distributed via the Internet. However, this ease of duplication is a sword double edged with the problems of software piracy. So for software, force #2 should really be “democratizing legal distribution”. On the other hand, distributing niche software and physical commodities are similar in the sense that they can both use aggregators to lower the cost of selling. For software, the cost of selling is much more than transporting the bits from one computer to another - there are also the issues of consumer trust, brand and quality association and effective marketing to potential users. In the digital world of computer viruses, phishing attacks, mal- and spy ware, a lesser known micro-ISV has much to gain by selling their long tail applications through well established aggregators. A small business owner is much more likely to buy a project management application listed on the Windows marketplace than buying directly from an unknown micro-ISV’s Website.

Although distributing software online does not always require trucking physical goods (except when selling shrink-wrapped software, which is still a common kind of commerce transaction) from stores to buyers, the delivery of hosted software services do require server infrastructure and data center operation support to host and run the application services. In many situations, it is not cost effective for an ISV to build and run their own operational infrastructure. Instead, it is frequently beneficial for SaaS ISV to have their applications hosted by professional SaaS hosting companies. Software service delivery is the subject of our recent paper. Look around at the companies providing application hosting services, and you will notice that many hosters are also aggregators of software services who provide the market channels for the ISVs they host.

However, the tasks of coordinating service delivery with SaaS hoster may still be too tedious for ISV producing long tail applications. This is why we are seeing new breeds of startups like Coghead and Longjump, where they provide a one-stop shop for ISV to develop, host, market and deliver their long tail applications.

 

Force #3: Connecting supply and demand

Technologies that connect consumers of long tail applications are what help drives increasing demand for niche software. Chris calls these types of technologies “filters”. Search engine results, reviews, user recommendations, blog opinions, application usage data - all these information influence application adoption and can eventually introduce new user behaviors and habits. Take social apps on Facebook for instance, various filters at that site not only help reduce the cost of search, but also ultimately influence the applications’ success and adoption:

 

image 

As you may have guessed, a one-size-fit-all filter will not do the best job of connecting users with the applications they want to find. We see from the Facebook example above that depending on the user’s goal, he may choose among the multiple filters to find applications categorized according to popularity, activities, active users or published date. Not to be neglected is the observation that the purpose of the Websites also act as filters: Facebook is the filter for social network applications; Office Live is the filter for line-of-business applications built for small-medium companies etc.

 

Force #4: Democratize Capitalism

This particular force is my contribution to long tail software economics. I believe that enhancing the opportunities to make money can further increase the supply of long tail software. The recent proliferation of Facebook applications illustrate this point very well - the authors of Facebook applications get to keep 100% of the ad-revenue attributed to their applications. Furthermore, incorporating an ad-funded software revenue model shifts the economics of when and how software is paid for, which can be explained through a fishing service metaphor. Using ad-funding is as if the long tail aggregator is running a “fishing service” business and software applications are hooks which baits (advertisements) are attached to entice the users to transact for some goods and services. The merchandise and service sellers are willing to pay for the “fishing service” when there is fish action. Effectively, hooks are essential components of the fishing service and the aggregator splits the ad-revenue received from the merchandise sellers with the long tail application providers. The ad-funded model not only provides opportunities for micro-ISVs to get paid, it also reduces the up front cost of trying and using software, thus removing some barriers to drive demand further and encourage software adoption.  

Ad-funding is not the only model to monetize long tail applications. Traditional licensing, subscription and transaction models are all still valid means for the ISV to get paid. The key point here is that the supply curve of long tail software is likely to grow when there are effective mechanisms for the developers to be compensated for their work.

Most long tail applications will never command the kinds of price tags that top selling software are sold at. However, with production tools and distribution democratized, it may only take a very short time and almost no capital investment to build long tail software. In fact, when the long tail software economy takes off, it is very conceivable to see many more micro-ISVs creating applications as fast as them can (in matter of days or weeks) and live off the the aggregate income from multiple long tail applications.

+++

Amazon and Netflix have already proven the profitable business model of selling long tail physical goods and entertainment content. The same cannot be said about the business of selling long tail software yet. However, there are existing and emerging businesses that are standing behind the forces we described above, as noted below in the table:

Force

Business

Current Examples

Democratize Production

Long tail application development tools, mashups and composite application platforms.

Microsoft Access, Ruby on Rails, APEX, Popfly, Sharepoint designer, Coghead, Longjump.

Democratize Distribution

Long tail aggregators, application service delivery platforms.

Office Live, Facebook, AppExchange, Windows Marketplace, BT Application Marketplace.

Connect Supply and Demand

Long tail filters.

Facebook ranking and search, Sourceforge marketplace search for open source software, Windows marketplace rating.

Democratize Capitalism

Long tail monetization engines.

Microsoft Adcenter, Google Ad-sense.

 

To prove that the market for long tail software is in fact a profitable one, we’ll need to orchestrate the integration of long tail development tools, software marketplaces, filters and monetization schemes deliberately.

May the force be with us…

Test With… Silverlight/CoreCLR

I have just released a new version of TestDriven.NET (2.7 Beta) with support for Silverlight/CoreCLR. Simply right click on the method you would like to execute with the CoreCLR and select 'Test With… Silverlight'. You can use this to execute any public methods (ad-hoc tests). I recommend you use 'Console.WriteLine' for any output messages. 'Trace.WriteLine' is supported by Silverlight's version of 'System.dll' but you would need to use something like DebugView to see the output.

Here is an example method being executed with Silverlight:

As shown above you do have access to the 'System.Windows.Browser.HtmlPage' static methods and properties (from the 'System.Silverlight' assembly). Hopefully people will find this feature useful for spelunking the Silverlight functionality.

I am working on proper unit testing support for Silverlight projects. At the moment none of the unit testing frameworks supported by TestDriven.NET (NUnit, MbUnit, MSTest or Zanebug) are compatible with the CoreCLR. I have had some success getting unit tests running using NUnitLite. You are welcome to contact me if you would like to try an early version of this.

Buzzword gets word count, easier sharing
Adobe’s online, Flash-based word processor has been revised and now includes word count, handy for journalists and essay writers: It also has easier sharing: simply copy the URL at the top of the document you are editing to create a link. I intended to demo this here; but cannot because Buzzwo…
Tim

Equivalence Class Partitioning Part 3 - The tests
In the last post we decomposed the set of characters in the ANSI character set into valid and invalid class subsets for use in a base filename component on the Windows Xp operating system. The second part of the testing technique of equivalence class partitioning is to then use this information in tests that will adequately evaluate the functional capabilities of the base filename parameter used by COMDLG32.DLL in Windows Xp. ( Note: It is appropriate here to differentiate between a test case and…(read more)

JC Penney Catalog! Circa 1977! What A Great Year!

Now, I was born in 1977, and never regretted it, until now!

There is an email going around, and it is pure gold (and green)!

Tonnes of ideas for Christmas! Including matching “Stretch Terry Jumpsuits”…

It’s here on my SkyDrive for all to enjoy… so, enjoy.

http://cid-2acaee0f398245b2.skydrive.live.com/self.aspx/Funsies/JC%20Penney%201977%20Catalog.pdf

Technorati tags: JC Penney Catalog 1977

You must be logged in to post a comment.