Monday, 5 November 2007

Why Did TAURUS Fail and CREST Succeed?

In 1989 the London Stock Exchange (LSE) put forward a proposal for a computerised system to ensure that share certificates and cash changed hands between the interested parties after the trading transaction, implicit in this was the dematerialisation of stock certificates. It was a big project with hundreds of staff contracted in and lots of external pressures from various different stakeholders. The project was scrapped in 1993 at a cost of £75 million to the LSE. However, the LSE still needed a computerised system and took up the project of CREST (Head, 2001). This project learnt from TAURUS’ failures and went ‘live’ in July 1996. There are many differences between the TAURUS and CREST projects that led to the eventual downfall of TAURUS and the success of CREST.

One of these problems was that TAURUS wanted to be “all things to men” and with the pressures from the companies that were members of the LSE, they included 21 Corporate Actions or Events to deal with the different business approaches (Head, 2001). Due to the amount of modification that had to be done to the software, this is one of the reasons the system failed. CREST learned from this and included only 2 business functions. However, these business functions made up 85-90% of the business transactions. It was the Pareto effect that was used here: the two business functions were only 10-15% of the total 21 functions set out by TAURUS however, adding the extra functions the system became overloaded (Head, 2001).

Another problem that TAURUS suffered from was the software package that it chose. It needed heavy modification due to the 21 different business transactions. This software had not previously been tested and resulted in a downfall. This was much the same as the problems that FoxMeyer found when installing an ERP system. It used an untried system that it was ill-equipped to install and failed disastrously (Buss, 1997). CREST did not replicate this problem and opted for a tried and tested method in which the Bank of England had experience in (Head, 2001). By drafting in the IT manager from the Bank of England the CREST project had experience on its side.

A further point that separates the failure of TAURUS and the success of CREST was the project structure. TAURUS was marred by ad-hocracy, fragmentation and mismanagement (Currie, 1997). This was due to the hundreds of contract staff it hired in, the consultations with the member firms, the legal aspects drafted by the government and the pressures of a project team repeatedly crossing the Atlantic (Head, 2001). CREST on the other hand was seen to be highly structured and controlled (Currie, 1997). It had a 20 member strong design team with four or five managers to constantly overlook the project. This highlights that although both the projects were high scale and using sophisticated technology, the management and co-ordination of the project is a key to its success.

Additionally, the budget and time constraints of the projects were seen to be a differentiator to their success. Goulielmos (2003) states that of the four concepts of failure in Information Systems is process failure where the project over runs its budget or time constrictions. TAURUS did both incurring increasing media attention and scrutiny, which led to an increase in pressure on the project team (Head, 2001). CREST again, learnt from this mistake by being highly structured and controlled. It was within its original budget and went ‘live’ at its target date. The CREST team also learnt to establish press relations and brief them about the progress of the project (Head, 2001). Therefore, they did suffer from general scrutiny about the project but, had a reduced amount of pressure compared to the previous project, TAURUS.

In sum, the TAURUS project implemented by the LSE suffered many downfalls and eventually was scrapped however, the success of CREST may be largely down to learning from previous mistakes. As highlighted above, from the downfalls of TAURUS, CREST has implemented a solution or more basically avoided these problems.


Buss, D, Nightmare, Context Magazine, 1997

Currie, W, Computerising the Stock Exchange: a Comparison of Two Information Systems, 1997

Goulielmos, M, Outlining Organisational Failure in Information Systems Development, Disaster Prevention and Management, 2003

Head, C, TAURUS and CREST, Failure and Success in Technology Project Management, 2001

No comments: