In mid-July, I attended Fusion 2000, Microsoft’s annual conference for their development and networking partners, a.k.a. Solution Providers, or SPs. The conference is designed to prepare the sales and marketing groups within these organizations to position and price upcoming solutions based on the Microsoft platform.

It should come as no surprise that the main message was centered around configuring and delivering solutions on the Microsoft .NET platform. But if you look beyond the marketing glitter, one of the major themes that emerged from the conference is that Microsoft now believes “ASPs are for real.”

Piloting for success
At the conference, Microsoft revealed that they had been running a series of licensing pilots over the last year designed to help them determine a strategy for offering their products on a subscription (or rental) basis rather than as a one-time purchase.

These licensing pilots centered on offering infrastructure based on Microsoft Windows NT 4.0 and Microsoft BackOffice and also productivity services centered around Microsoft Exchange 5.5 and Microsoft Office 2000. As a result of these pilots, Microsoft has determined that there are two basic licensing decisions that ASPs need to be ready to offer, and customers willing to accept.

Decision #1: How long does the license last?
The first licensing decision is one of purchasing a perpetual license versus a non-perpetual license.

In the past, the only option from Microsoft has been a perpetual license, i.e. when you buy a copy of Office 97, you have the option of using that license on a single machine “in perpetuity.” Microsoft simulated a non-perpetual license by allowing a customer to upgrade every 18 months for a fee, thus preserving a revenue stream on a purchased perpetual license.

But now Microsoft will offer a real non-perpetual license. For example, you can purchase, (from an authorized ASP), a right to use the current version of Microsoft Office for one month at a price set by the ASP. That price would include not only the license cost of the software, but also the ASP’s fees for managing and maintaining access to the software.

The other key difference between a perpetual and non-perpetual license is to whom (or what) the license is applied. In the case of a perpetual license, the license is applied to a device, i.e. anyone can use the copy of Office installed on a specific PC. In the non-perpetual mode, the license is applied to an individual, i.e. the individual has the right to use Office on ANY DEVICE from which he or she can access a hosted version of Office. This would include their office PC, a home PC, or even an airport kiosk from which they can load a Web session into Windows Terminal Services.

Decision #2: How many licenses do I need?
The second key decision that an ASP must make on behalf of their customers is whether to license certain products by the user or by the CPU. For some products, (like Windows 2000, Commerce Server, and SQL Server), the hosting company has the option of paying Microsoft a monthly per CPU fee for using a copy of the software.

If the hosting company chooses to segregate companies into their own unique, non-shared environments, it will be more cost effective to pay Microsoft and charge their customers by the user rather than the CPU.

The exception to this rule is for applications that require external authenticated access to SQL Server databases or to Commerce Server. Microsoft has finally dropped their maddening and somewhat irrational requirement that externally authenticated Internet users be required to pay for a CAL or Client Access License.

(This was one of the biggest inhibitors to widespread, legal deployment of Microsoft’s Commerce Server product. There was no way for you, as the company deploying an e-commerce solution, to determine how many unique, external users would access your site. It made Commerce deployment on the Microsoft platform unpredictable and, potentially, prohibitively expensive.)

The more compelling case for per-CPU pricing from the hosting company’s perspective is the ability to share large, multiprocessor servers running Windows 2000 for access and authentication, SQL Server 2000 for database access, and Commerce Server for commerce transactions. This allows the ASP to spread the cost of maintaining these services across multiple boxes and to offer hosting services at a lower price than the company can purchase and manage the product without the ASP.

Most of the per-user pricing is determined by making the impact to Microsoft revenue-neutral. That means that the hosting company’s real ability to generate revenue lies in delivering efficient shared services to its customers. In order to keep from cannibalizing their revenue, Microsoft has chosen to stick with per-user pricing only for products like Office, Exchange, and Terminal Server. ASPs will have to add value there by integration and custom development surrounding those products.

Will Microsoft be successful with ASPs?
Microsoft has chosen a different route than their competitors in the software space. Companies like Oracle, SAP and PeopleSoft have chosen to compete with their own partners in the ASP space. Other than their bCentral effort, (aimed at companies with less than 50 PCs), Microsoft is encouraging companies to focus on providing hosted offerings to companies with 1,000 or more PCs. Unfortunately, Microsoft has no clear vision or program to help companies in the 50 to 1000 PC market segment—either the companies with this number of seats or the ASPs who serve them.

Microsoft’s ultimate success here actually rides on two things. First, the rest of the market needs to keep rolling over on the desktop. If Sun, IBM, Oracle, and AOL all decided to band together and focus on providing the Sun StarOffice Suite on a standard Linux/Gnome platform using a rich, hosted Lotus Notes back end for a set fee per month, ($5 or less), then Microsoft would have to dramatically alter their plans to continue desktop operating system and productivity suite domination.

The second key to Microsoft’s success is their own ability to deploy large numbers of well-configured and manageable Windows 2000 platforms, (hosted and non-hosted), in a short period of time. If Microsoft can push enough Windows 2000 and .NET servers out the door in the next 12 months, then the game’s over anyway.
Is Microsoft’s non-perpetual license marketable? Share your opinions with us by e-mail or by posting below.