Wednesday, November 08, 2006

Beware the hidden costs (FT)

Beware the hidden costs

By Alan Cane
Published: November 8 2006 11:27 Last updated: November 8 2006 11:27

The challenges faced by information technology managers in meeting their business obligations within tight budgets can be encapsulated in a few telling sentences.

First, the cost to the average organisation of a personal computer is about $10,000 a year, of which less than 10 per cent goes on the hardware. A few per cent more account for the operating software, but the bulk of the cost is absorbed by administration and operational support including installation and training. For larger systems, the same iceberg-like ratio of seen to unseen costs obtains.

Second, IT managers are increasingly being asked by their business colleagues to develop innovative systems to provide efficiencies and competitive advantage within budgets that are at best static and often declining in real terms.

In general, between 70 and 90 per cent of their existing budget is already committed to what Lisa Hammond, chief executive of the consultancy Centrix, describes as “keeping the lights on”, in other words, running and maintaining existing systems.

So this is the conundrum: funds are limited and the cost to the business of new investment in technology will be significantly greater than the price list might suggest.

“The chief information officer has a twofold task,” says Ms Hammond. “He or she has to keep the current systems alive and handle the demands from the business. The question is how they can use their budget to create some cash to spend on innovation.”

The situation has opened the floodgates to a range of vendor financing schemes for hardware, software and services (see feature, Page 6) of varying degrees of novelty and imagination.

For example, Banca Intesa, the Italian financial institution, has made use of vouchers offered with Microsoft’s software assurance scheme to put its IT staff through training courses without draining the group’s conventional training budget.

And Smiths News, the UK’s largest magazine and newspaper wholesaler, is taking advantage of EMC’s scheme for balancing its customers’ storage needs: EMC provides more memory than the customer initially requires but only charges for it when it is used.

Large and small companies make use of vendor financing. ING World, the sports and media group, uses Microsoft Financing to pay for its software licences: valuable when it is going through a growth phase and incurring new IT costs to integrate businesses into the group.

But how should a company approach investment in IT? It is close to 20 years since the concept of “total cost of ownership” revolutionised the way vendors and customers thought about the cost of acquiring new computer systems.

Formalised by the consulting firm Gartner Group as a family of software tools, TCO has been widely used to help make purchasing decisions. Essentially, a financial estimate based on the direct and indirect costs of a capital investment, such as a new computer system, it traditionally includes the cost of hardware, software, installation, training, support, downtime, staffing, maintenance and infrastructure.

Nevertheless, even today, surveys have shown that many purchasing decisions are made purely on price, without consideration of the other factors enumerated in the TCO.

Moreover, a more sophisticated approach is already evolving. “You would have to go back five years or more to find people using pure TCO models,” says Peter Critchley, strategy director for the consultancy Morse. “People are beginning to think about the whole computing environment as an asset and considering how to manage that asset throughout its life cycle.”

CIOs today, he argues, are positioning themselves as businesspeople rather than technologists, taking responsibility for translating between the technology and the business: “CIOs almost take it for granted that a good adviser on technology can help them have an effective, efficient and optimised data centre.

“What they really want is for that data centre to be highly flexible and changeable as the business changes. The cost of change, if you don’t have the right architecture in your data centre to be able to do that, could be huge.”

Philip Dawson, a senior consultant with Gartner argues robustly that the term TCO has become devalued by misuse and by misleading comparisons: “When you look at an analysis, it will compare old, unmanaged, distributed stuff versus new, consolidated, managed stuff. So a Unix guy will say ‘Look at these old Windows NT shops. If you put in a Unix box you’ll get X, Y, Z savings. But if you move forward with the NT stuff with Microsoft, you get the majority of the savings anyway.

“Any vendor can show you a TCO comparison which shows their products looking better than their competitor’s products. I don’t mind working with clients on TCO as a methodology for the assessment of their portfolio: what they have now, what they need and where they need to go. The bit I don’t like is where a vendor says ‘I want to carry out a piece of work with you to prove X or Y’.”

Dawson explains: “The value of TCO lies in understanding your portfolio and its cost drivers. In the early days, TCO was about cost of projects. That is the wrong approach. If you have eight projects in a portfolio and you optimise them separately on a TCO model, you get eight separate islands. If you start a new project you have to start a new island.

“Total cost of service is a better approach. If you understand in a horizontal fashion ‘this is the cost of my database services, this is the cost of my security services, this is the cost of the IT equipment’, then that drives down TCO on every project.”

He emphasises the importance of reducing the amount of under-utilised equipment which, in many data centres, grows in an uncontrolled way: “Rationalisation makes things cheaper. If you reduce complexity, then TCO just falls through the floor.”

It is a view which resonates strongly with Lisa Hammond of Centrix.

She points out that what she calls “configuration items” – hardware and software products to you and me – tend to be bought system by system, project by project and just accumulate.

“There will be a load of configuration items that are just not needed. The first job is to map the IT services in the data centre to the business processes that they support. You will find a lot of configuration items that do not map to your key processes. Rationalise out the ones you don’t need. That one exercise can save between 5 per cent and 30 per cent of the budget,” she argues.
She believes there is a growing trend towards “buy rather than make” through the use of subscription services.

“Nowadays, we can subscribe to existing services that are 80 to 90 per cent fit for purpose,” she says, pointing to the options open to the CIO of a large international bank shopping for a claims processing system.

Inhouse, the projects would have taken 80 staff two years to complete. Outsourced abroad, it would have taken the same time, almost the same number of people but one quarter of the cost.
The third option came from an Italian company which proposed to charge £11 a claim for processing through its own system, giving the bank’s CIO the opportunity to get into the market in six weeks.

“You can buy services from smaller, innovative companies who have it set up already. And if you don’t like it, you can exit cleanly,” she argues.

If speed to market is important, so is time to return on investment.

Paul Loftus, who is in charge of IBM’s infrastructure support worldwide, echoes Peter Critchley’s views: “The CIO’s role today is being transformed from the classic IT specialist to business manager.

“They are being challenged to spend far more of their time on business transformation and integration, in particular, decreasing the time it takes to realise maximum value from an investment. That’s the fundamental play.

“What CIOs and business leaders around the world are asking me for, in addition to the basic technology, is to focus those value-added activities in the enterprise so as to augment their role as a change agent.”

Mr Loftus says that IBM has completely refreshed its service offerings in a number of areas to address the new environment and to return value to the customer more rapidly: “Our classic approach was labour-based, where each and every customer would sit down and define their individual plan and we would build it.

“What we have seen is a maturation of that thought process and today our portfolio of offerings not only has a skilled labour base but a set of recurring hardware and software assets that have been built or acquired.”

He shrugs off the idea that TCO analysis can eliminate risk, however: “Are there costs of complexity, of integration that one might not have clear sight of in going into a new installation? Certainly. That could be the case on any major project.

“But there’s both a realisation that this is not a precise business and a renewed determination to get on with it. Investment in technology is, at the end of the day, just as or more important than it was 10 years ago.”

Copyright The Financial Times Limited 2006

No comments: