VSJ – March 2002 – Work in Progress

Allen Woods, MIAP, managing director of JIT Software Ltd and author of a range of organisation modelling and management software has written three articles for us about a system requirements approach adopted by his company. He points out that a common reason for IT project failure is the often tenuous link between requirement definition and business strategy. This month, he outlines the rationale underlying his approach.

A primary stumbling block to identifying key information needs is a lack of communication between business and IT. From the perspective of the IS/IT system designer, the communication block can be broken down into two broad forms. The first is the knowledge gap between the IT/IS technician and business. Traditionally, IT/IS system design skills have been seen as a black art best left to the professionals. Equally, the IT/IS professional has often lacked detailed knowledge of how an organisation works. IS/IT designers may provide what they think is required, rather than what is actually wanted.

A second problem is the passage of company policy from the boardroom to the shop floor in a form that everyone in the organisation can understand. For instance, the CEO of a car company may wish to capture an additional 10% of market share and so make an extra £100m profit. However, that may mean little to shop floor workers, who may not know what ‘market share’ means. They do know that they have to make 10 axles to the required quality standard, though. So we need to translate ‘improve market share’ into ‘make 10 good quality axles’.

One way to overcome these gaps is for business and IT to adopt a common set of tools that allow system sponsors to specify a project, using current business ‘best practice’ so that the requirement can be mapped into an IS/IT method like UML. Where are the tools and standards we can adopt that lend themselves to IS/IT design principles?

The Government Background According to the Treasury, over 40% of UK GDP is in the hands of civil servants in some form of taxation. So the biggest single market is central and local government.

Government spending is tightly controlled and, for major contracts at the tender stage, potential contractors must demonstrate compliance with government sponsored standards. For instance, ISO 9000, which is about process definitions, is one factor that can determine whether or not a bid will get past the initial acceptance stage.

Over the past few years, there has been a sea change in the way central and local government accounts for how money is spent. One aim of the changes is to identify the ‘whole cost’ of a product or service.  The main driver of ‘whole cost’ or Resource Account Budgeting (RAB) is ‘best value for money’.

RAB is an accruals accounting technique that identifies and quantifies everything that impacts on the real financial cost of a product or service. For instance, a vehicle has its initial purchase cost, but other associated costs include insurance, fuel, a driver etc. Many organisations that have applied RAB believe that judging things from a solely financial viewpoint misses much of the intrinsic value of a product or service. Intrinsic values include popularity, reliability and ‘fitness for purpose’. These values should form part of the ‘best value for money’ concept. For example, one machine may cost £1000 and an alternative £1500, on the face of it a potential saving of £500.00 on the purchase of the former. But ease of use, maintenance etc., may be lower for the more expensive option and therefore have a greater impact on long-term budget planning. RAB can lead organisations into collecting much more data on a task or piece of equipment than might have been the case in the past.

Measuring aspects of the value of an item generates problems of its own. Can we define suitable performance indicators for use in identifying success or failure? How do we group such indicators logically for reporting purposes? Can we collect relevant data on which to measure performance objectively?

In practice, Government forces standards. They may not be the most appropriate, but because over 40% of GDP is in their hands, most organisations will either adopt them to get business, or be aware of them. ISO 9000 is a case in point.  From a policy perspective, the organisations responsible for the promotion of such standards include the National Audit Office, the Audit Commission, and perhaps more importantly the Performance and Innovation Unit at the Cabinet Office.

All these organisations currently favour the following techniques that can be adapted to suit a systematic approach to statements of requirements:

  • The European Foundation for Quality Management Model
  • The Business Balanced Score Card
  • ISO 9* series of standards
  • ISO 14* series
  • Investors In People (IIP)

The Business Background The main imperative of business is, of course, profit. To improve profit we sell more product, and/or reduce costs. The use of money is easy to measure. There are sound historical reasons for this. For many years the only real indicator of success was the profit and loss account. However, accounts do not measure intrinsic value and they only give a historical view.

Business is coming to realise the need to measure other aspects of business activity, like customer satisfaction, that affect how a company fares. With improvements in data capture and information delivery, organisations have to be more responsive to non-financial pressures. Increasingly, concepts associated with ‘Performance Measurement’ are being adopted. Performance Measurement relies on the identification of ‘Key Performance Indicators’ and the subsequent collection of data on which to calculate how well an organisation is doing.

Performance Measurement is a valid approach that is gaining credibility. Its adoption is facilitated by the proliferation of powerful computers (particularly PCs) combined with the Internet. Improvements in technology aside, the validity of performance management is easy to defend. All the Wall Street Fortune 500 companies that have been in the list since its inception use it.  One, DuPont, is seen by many as authoritative in the field.

Performance measurement depends on an analysis technique coupled with a reporting mechanism. Analysis techniques vary from company to company. However, there is one technique associated with performance measurement, lending itself to computing, which describes information delivery and is gaining a lot of support. This is the Balanced Score Card.

Conclusions I have described some factors that can affect the definition of system requirements and have suggested that it is sensible for IS/IT practitioners to adopt business best practice where possible, so that system sponsors can define requirements for ease of mapping into IS/IT design disciplines. Next month, I’ll describe the tools and techniques we have adopted and how they can be used to define structured requirements.

If you can’t wait for next month’s instalment, visit www.jit-software.com. You can contact Allen at Allenwoods@jit-software.com

Interesting project or development? Let us know at eo@iap.org.uk!

Comments are closed.