Guild Companies

Hewlett-Packard and Microsoft announced last week that they would be throwing about $25 million each into a joint venture to establish a team of sales and services professionals that can create, push and support .NET-based applications for enterprises. Under the deal, HP has been named the first worldwide prime integrator for .NET--it certainly will not be the only vendor with that title--and HP has promised to push .NET as the main application platform for its ProLiant Intel-based server line.

By jumping in so early in the game with .NET, HP is hoping to get the inside track on all of the consulting and sales that will come when Microsoft's .NET tools are readily available and companies once again are in a mood to spend money and create the Web-based services applications that are spoken much about these days but rarely actually built. HP, by virtue of its acquisition of Compaq, is the largest peddler of Wintel servers in the world, and is neck-and-neck with IBM for being the number one server supplier in the world. HP has the support contract for Microsoft's own internal networks and help desks (inherited from Compaq), but IBM has the largest services business in the world and the most lucrative and numerically large server base in the world, with about 400,000 to 500,000 customers.

It's easy to see why Microsoft agreed to split the cost that HP will have to bear to build a .NET practice, since HP could have picked another middleware stack, based on a mix of things, and sold that across its Unix, Windows, and Linux server lines. This would, in fact, be more consistent with HP's triple platform policy. So Microsoft has paid HP to buy into the .NET religion and train some 3,000 services professionals and another 5,000 sales reps to peddle .NET solutions on ProLiant machines. HP just chopped 16,800 people after the merger (a bit higher than the 15,000 it expected), and it is looking for a target rich and complex environment in which to build a services practice that can reap profits and drive product sales. Exactly what IBM will do with .NET is unclear, but it is harder for IBM to be all WebSphere all the time and talk about .NET too. HP is clearly planning to beat IBM over the head with .NET in those customer accounts where .NET will sell--small and medium businesses that are allergic to IBM's proprietary midrange and mainframe platforms and who want something more mainstream than one of the big Unixes that will also run on Intel iron--while IBM takes the high ground with its WebSphere campaign. IBM's SMB people and Global Services organization will undoubtedly be rattled by the Microsoft-HP deal, and will soon get IBM rolling towards building a similar practice.

Microsoft has apparently agreed to hire 1,000 people for its end of the deal to support HP's .NET efforts and part of that $50 million kitty will go to pay for co-marketing of the HP-Microsoft .NET solutions and services as they evolve. Ann Livermore, who is in charge of the HP Services unit, says that HP plans to focus on four core areas in its drive for .NET Web services: business intelligence and data mining; Web-enabling existing applications; integrating information stored in multiple and usually incompatible applications, and Web-based messaging and collaboration. These are the heart of Web services.

Just because HP is getting behind .NET doesn't mean that it is abandoning Java-based alternatives for those customers who want to go the J2EE Web Services route rather than the .NET way. HP has strong ties to Oracle and its own views about best-of-breed middleware for supporting Unix and Linux applications as Web services. It is also a supporter of Java, and is not about to abandon that support--and for practical reasons. Because .NET is a Microsoft-centric technology, it can't be ported to HP-UX and Linux, which means HP has to keep alternatives to .NET, just like IBM has to.