[an error occurred while processing this directive]

 


Opinion
- Value for money - (December 1999)
IT managers nowadays want to see a good Return on Investment (ROI) after any purchase. Can Windows 2000 ensure it? Mark Vernon finds out.
.

Return on Investment (ROI) is arguably the issue for Chief Information Officers or their executive equivalent. For too long, the argument goes, IT has got away with it. But now consultancies issue ROI reports on an almost daily basis. Typically, they show that a large proportion, if not a majority, of companies fail to realise the benefits they expected from implementing technology solutions. With the consequence that the challenge now facing IT departments is to enhance the flexibility and functionality of what they have and develop a more business-friendly mode of operation.

The trouble is that IT is a victim of its own success. When it was a plumbing issue, what happened underneath the sink did not matter as long as when the tap was turned, the water ran. But it has now become strategic, and that also means subject to endless scrutiny. Over half the workforce is now occupied with information. The price of many goods would be virtually zero but for the value that information adds to them. If IT can take credit, it must also be ready to take the rap.

The highs and lows


However, the trouble is that ROI is notoriously difficult to determine. It is often either meaninglessly high, as firms find when implementing intranets, say; the initial costs are nominal but, with some thought, the benefits can be substantial and far-reaching. Alternatively, the investment seemed unavoidable but very expensive. ERP is the obvious candidate here. A recent Spikes Cavell report noted that two thirds of companies have failed to deliver satisfactory ROIs on SAP, though they would not have thought not to buy ERP. Windows 2000 Professional has recently been subjected to ROI scrutiny. The immediate question at this stage is whether or not to upgrade, but this is closely related to the complex task of producing a methodology to access potential ROI.

The consultancy research group Giga has published its take on the issue, believing that Windows 2000 provides enough of an improvement to warrant deployment shortly after launch, under certain circumstances. However, they begin from a low base line. "For Giga to advocate a platform, an OS doesn’t have to be perfect. It simply has to solve more problems than it creates," says analyst Rob Enderle.

An important issue is clearly the cost of the upgrade. A less than complimentary report from GartnerGroup has been clarified by Giga in this respect. For example, Giga points out that a clean installation depends on all necessary drivers being available and close at hand during the upgrade. For some hardware, drivers are already available. For less Microsoft-friendly vendors or older equipment, these tools might not be ready for months after Windows 2000 ships. Systems that are likely to work well with Windows 2000 need to be identified against those that won’t. It is in determining details like this that ROI stands or falls. Incidentally, taking this into consideration, and given Year 2000 issues, Giga recommends that in-place upgrades should be delayed until the second or third quarter of 2000.

Is it worth it?


Similar points are picked up and developed by Arthur Andersen. So, in addition to the costs of the software itself, other up-front costs the firm believes will be evident include those resulting from the need for new hardware, administrator and end-user training, configuration and deployment of the new operating system, compatibility-testing, and possible reengineering of application software. Clearly these issues are common to any upgrade and will vary according to the particularities of organisations concerned. However, Andersen does conclude that, "For many companies, Windows 2000 Professional’s improved features compared to Windows 98, Windows 95, and Windows NT Workstation 4.0, will offset the up-front investment. It is likely that larger companies that are running native Windows NT-based environments, and that do not utilise many third-party tools for services such as security and synchronisation, will see greater benefits and more positive impacts on ROI."

However, there is also a warning. A dearth of Windows 2000 servers in the infrastructure, and the presence of stand-alone applications that have already addressed some of the benefits that Windows 2000 seeks to bring, notably in respect of management and administration, will reduce the potential ROIs and might make the case for an upgrade far less convincing.

Technology should be held to account since today a business can stand or fall on the IT investment it makes. There is perhaps a tendency for some organisations to treat technology as an automatic competitive edge. But IT is not an optional extra for the risk taking few. It is a necessity. And that only means more care should be taken when it comes to ROI.