New models allow proper assessment of IT risks

Besides labour, capital and nature, software is an important factor in production. One result of this is that IT projects make the front pages with increasing regularity; sadly this is usually whenever the costs get out of hand and the projects actually seem to have become unmanageable. Erald Kulk has developed a method for avoiding this scenario in future. Business people can use his models to take calculated risks instead of random decisions based on gut instincts.

New requirements

One of the major problems when implementing large-scale IT projects is the continually changing context. New requirements are constantly imposed on the software that is being developed, whether through progressive insights or legislative changes. Using IT project data from a range of industries, Kulk has analysed the causes underlying cost overruns. He has developed a model for working out the point at which a project can no longer accept any further adjustments. He has precisely calculated the 'danger zone', where more work is coming in than is being completed, so that it becomes impossible to finalise a project. One of the factors he took into account for this calculation was the size of the project -- a large project can reach this danger zone quite rapidly, with relatively little in the way of further growth.

Contrary to expectations, the businesses whose IT projects ended up 'in the soup' were not the 'immature' ones. Only when the entire organisation was working at a more advanced level were the costs more accurately estimated, so that there was a greater chance of success. It was the smaller organisations that succeeded in making a more realistic assessment of a project's costs and risks. Large organisations often had to deal with too many people, each with separate lines of communication.

EQUITY

Erald Kulk's research forms part of the NWO project 'Exploring Quantifiable Information Technology Yields' (EQUITY), which in turn is part of NWO's JACQUARD research programme. JACQUARD focuses on software engineering and 'software as a service' -- otherwise known as 'cloud computing' -- and places a strong emphasis on collaboration between science and industry. For instance, the EQUITY project is a collaboration between VU University Amsterdam and ABN AMRO. Kulk's research was largely undertaken at ABN AMRO.

The use of practical data means that the models delivered by the research can be applied immediately, and there are a range of available benchmarks, even for businesses that have collected little data themselves. Using Erald Kulk's models, it should be possible to get any projects likely to derail back on the right track in good time.