|
|
|
| mn :: comp | |
As of mid 2003, the hosting market's health is not improving at all. As predicted in last quarter's notes, the mathematics of revenue that fails to exceed expenditure is driving more companies to exit the business. If you didn't read those before, please go do so now as this document builds on that one. Expect the continuing recession to dampen enthusiasm further: consolidation will be the watchword until someone finds a radical way to change the whole market.
Last quarter we discussed all the reasons for putting the company eggs in someone else's building; this quarter we will revisit some assumptions. While the likelihood of a third or fourth generation datacenter going completely dark for technical reasons is still acceptably low, the risk of that same datacenter losing financial viability for its owner is growing higher by the day. It may be wise to review the real uptime requirements of your Internet operations. While it is clear that highly redundant systems and facilities are required for many operations, is it truly worth the tripling of costs for the average medium-sized business? And if your business is already accepting the high risk levels of server failure, where is the value add in hosting at a data center instead?
For most companies that go through this analysis, the value of colocation at a hosting center is clear; the value of managed services at a hosting center is somewhat less so. While the managed hosting provider does offer a greater level of expertise, use of this expertise for functions like troubleshooting or development is not free and the cost savings versus using less-capable internal employees or external consultants is minimal. When observed through the lens of an 18-month or 24-month return-on-investment anaylsis, the surface benefits of outsourcing become less compelling for business that do not need the redundancy or flexibility offered by an external partner.
The market is moving to molo -- colocation with a selected set of managed services, typically a rudimentary monitoring offering and a full-service enterprise backup solution, perhaps a managed remote access solution or managed SAN/NAS offering. This movement is being driven by the need to reduce monthly costs or cut headcount and services. Simply put, managed services are a luxury that many can do without.
Since colocation at market rates doesn't pay the bills, this means that vendors unable to successfully transition to a la carte molo or otherwise change the model will continue to go out of business. The large hosting company which focuses solely on hosting has never been a successful model. It may become so if this period of consolidation leads to two or three standing players who are able to collude market rates back to something that provides revenue greater than expense, but this seems unlikely. After all, the high expenses come from the fact that data centers are so large, and the fact that data centers are so large comes from the fact that big Internet pipes are few. Considering the facts that 75% of fiber capacity in North America is unused and that high-speed wireless keeps getting faster and cheaper, it seems easier to imagine a future in which every mid-sized business district in the United States has its own 500 square foot data center. If data centers cost $200,000 to build and $10,000 a month to operate, selling the space for $30 per square foot per month begins to work. Selling franchise rights to small entreprenuers makes even more sense.
Franchised molo is probably some years away, depending as it does on greater ubiquity of high-speed bandwidth. In the meantime, hosting providers will continue to bleed.
Copyright 2003, Jack Coates. This document may be redistributed freely as long as it is not altered.
|
Last modified: Nov 25, 2005 12:48 pm.
|
||
|
|
