Recently, Microsoft's top executives disclosed that the company will embark on a new business model: shifting from selling shrink-wrapped software to offering applications on a use-as-needed basis. The move copies application service provider offerings that have sprouted up all over the Web. Under this model, Microsoft would no longer depend exclusively on sales of licenses but would instead draw additional income from the delivery of applications and support over networks to produce a more predictable stream of cash.

Bill Gates' notion of a megaserver-centered architecture confirms that such a service strategy has been in the works for some time. In the opinion of a leading financial analyst who follows Microsoft closely, "They would eventually like to get a cut every time their software is used."

Trillion-dollar potential This would amount to one of the largest instances of outsourcing in history. Microsoft's 1998 revenue of $10.5 billion from Windows and desktop applications would be augmented with a hefty share of customers' spending on desktop support - a market that I estimate to be worth at least $200 billion (100 million global desktops multiplied by $2,000 in displaceable support costs per desktop). As I see it, this could raise Microsoft's revenue beyond anything it can potentially earn from software sales. Profits from real-time desktop support could propel its market capitalization beyond $1 trillion. Microsoft would evolve from being a seller of capital assets to the world's dominant information-services vendor.

Microsoft's largest impediment is executing this strategy without monitoring every keystroke on every desktop, while retaining its stranglehold on the desktop. This helps explain why Microsoft is insisting on tight integration of a proprietary browser and programming protocols with its operating systems: It's central to its ability to pursue the rental-services option. That Microsoft is pursuing this tactic is evident from several other clues, including an elaborate scheme for tracking the configuration of what its customers have installed on their machines. I suspect that some of the best brains in Redmond are spending much time trying to figure out whether the entire concept can be executed given legal, privacy and security considerations.

From Microsoft's standpoint, selling services instead of one-time licensing is of vital importance. Currently, it derives 79 per cent of its revenue - and by far, most of its profits - from the sale of Windows and desktop applications. As resistance builds to paying additional license fees for unceasing upgrades, the existing Microsoft economic model is becoming increasingly vulnerable.

Quality benefitsl There may be some good news in these pricing shifts. A customer's relationship with a vendor that sells a product outright without warranties is at best tenuous. When a customer purchases an item, most of the risks of the total cost of ownership are unloaded from the vendor to the customer. Usage contracts will motivate suppliers to offer higher quality, improved availability and superior reliability, because every improvement should decrease their costs and increase profits.

Usage contracts are also a good deal from the customer's standpoint. They offer cost predictability, automatic technology refreshment and relief from the complexities that are becoming unmanageable for just about everyone.

Technologically complex solutions will increasingly be rented and outsourced instead of being owned. This is true in other technology marketplaces: Witness GE and Boeing's recent $20 billion deal, in which Boeing aircraft with GE engines will be provided to airlines for a fixed operating cost, instead of being sold outright. Networked computers are ideally set up to follow this model. IT managers will have to carefully examine this option and discover whether this initially attractive service may pave the road to a new form of electronic bondage.