The Bloor Perspective: Compaq's direct intentions, SAP's turnaround and the lack of apps for Windows 2000

Last week Compaq dug deep into its pockets and spent almost $370m on the distribution arm of Inacom Corporation, an acquisition Compaq sees as key to its own operations and one that will help it achieve its target of reducing inventory and speeding cycle time. But the move has hit a number of other high-profile organisations right in their key areas.

Among the casualties are the likes of Dell, Hewlett-Packard and IBM, all of whom have relied heavily on Inacom for the distribution of their PCs. But IBM may well be hurt the most. Having had a long-standing relationship with Inacom that saw the distributor building, configuring, distributing and servicing IBM's PCs, it would now appear that the relationship is going to end.

It is unlikely that this will have a dramatic effect on IBM's PC business, but for Compaq it suggests that fairly major changes are afoot. Certainly we knew that Compaq was keen to shift its distribution to a more direct model - but perhaps not everyone was aware of just how serious it was about this venture. This move confirms it. Compaq is going direct in a major way and with Inacom's wealth of experience it should be doing it very well in little to no time.

** SAP rising **

Preliminary results from SAP were announced last Friday and they showed fourth quarter (Q4) licence revenues of over $800m. This was a staggering 40 per cent year-on-year increase over Q4 1998 and as a result SAP's stock price on the New York stock exchange leapt, by 33 per cent, to a new 52-week high of $60.

A significant factor in SAP's rally was undoubtedly the launch of its mySAP.com Internet solution in the middle of last year. SAP has been quick to realise that ecommerce puts a strain on the logistics part of any business and that if these functions could be integrated more efficiently with front office functions then a much more competitive ebusiness could be created.

It's probably too early for SAP's front office components to have contributed significantly to licence revenues, however. It's more likely that the mySAP.com vision has repositioned SAP's traditional ERP solution in the customer's mind from a "necessary evil" to an important pre-requisite for building a successful ebusiness - a view that, if proliferated across the ERP industry, could spell a general unblocking of the ERP sector.

But SAP is still not out of the woods completely. Let's not forget that this superb Q4 number follows three consecutive quarters of consistently poor performance - during which time SAP has lost a lot of talent to more "glamorous" companies like Siebel, Oracle and Ariba. It does however show that the longer term prospects for SAP are looking up.

** Windows 2001 **

A recent survey of US developers from Evans Marketing Services shows that two-thirds of developers plan to delay application writing for the Windows 2000 platform until 2001 or later.

Reasons for this drift were various but the growing popularity of Linux and other open source software was seen as key to the sea change in developer backing. All of this will come as bad news for Microsoft which, despite its size and dominance of the OS market, cannot afford to overlook the increasing impact that open source is having on the developer community.

In the past, customers were prepared to wait for new releases from Microsoft rather than go with the "incompatible", though existing, competition. This time around, however, the landscape is very different. Most customers already have Microsoft software, and so can look at their own installations rather than those of their neighbours. And most have solved the problems of the desktop or the workgroup, both areas where Windows NT has already proved its worth.

But as the debate moves into the machine room, companies are less and less willing to use Microsoft software without significant proof that it works, does so cost-effectively, and can interoperate with legacy platforms and new applications alike. Certainly Microsoft can argue all of these points, but the fact is that Windows 2000 does not yet have the roll of honour that it requires to become a de facto standard in the server space.

Linux on the other hand is already being used. Its reputation is growing as a stable OS for general use, and while it is currently seen as unsuitable for mission-critical applications, it is at least in a position where opportunities to use it can be clearly identified.

This brings us to the issue of application support. Windows NT is already out there and will remain a suitable OS for many applications. But Linux installations are now in sufficient quantities to merit the attention of developers too, leaving Windows 2000 with an uphill struggle, against Linux on the one side and NT on the other.

End-user organisations will only move to Windows 2000 if there is sufficient reason to do so. Unlike Windows NT, which rode the coattails of the client/server revolution, Windows 2000 requires a migration from the installed operating system base.

Any benefits of the migration must outweigh their costs. Benefits will be measured in terms of new applications which cannot be run on existing platforms, including NT. If Microsoft does not have the application companies on its side, Windows 2000 will find itself lacking sufficient reasons, in terms of convincing killer apps, for end users to migrate. While it is waiting for its market to materialise, Windows 2000 may find itself submerged beneath the rising tide of Linux.

For more analysis, see http://www.it-director.com