Have we misinterpreted the business value of the long tail?
I've been thinking about the economics of the long tail and its impact on the business of Silicon Valley's largest Internet companies and discovering some very interesting issues, which have lead me to some startling conclusions. But please check my math...
My thinking about the long tail was triggered by a conversation earlier this week with David Scott, the CEO of 3PAR, which provides highly automated data storage systems for massive server farms. This is data storage for what these days is known as cloud computing, which was sometimes called utility computing, and which also supports Software As A Service companies (SAAS), and enables Services Oriented Architecture (SOA) IT.
3PAR customers are among the world's largest online services companies such as MySpace, and Savvis, which runs a global network of 24 data centers providing IT infrastructure services on-demand. They service hundreds of millions of users.
3PAR enables what is known as thin provisioning, it is the ability to quickly reconfigure data storage on-the-fly. It is a just-in-time approach to IT systems in the same way manufacturing lines were streamlined in the 1990s and made more agile and efficient through tight control of supply chains.
In addition to cutting IT costs through better utilization of data storage resources, 3PAR systems also eliminate the tedious and expensive human admin tasks required to run data storage systems. Labor cost cuts provide customers with huge savings that can be kept to improve the bottom line or invested in generating new business.
Mr Scott was telling me that 3PAR focuses on the transactional data storage market rather than the low end of the market, such as providing data storage for long tail services such hosting a photo of his grandma.
I asked why? "I might view a photo of my grandma, and that might be once a year or less. Yet that data still has to be hosted, secured, managed, and backed up. What is the business case for that photo? The cost of keeping that photo and others like it, will continually force companies to cut their storage costs and that will lead to them to storing the data themselves on cheap disk drives."
That makes perfect sense. Clearly, there is little money to be made, and nothing to gain from 3PAR providing storage for long tail data because it eventually leads to the loss of business rather than a returning customer.
It is better for it to target the high margins in providing storage for transactional data.
That's what got me thinking about the long tail and reaching some startling conclusions.