TAURUS is a good example of how a project can go badly wrong even when the original goal makes sense. The London Stock Exchange wanted to replace paper share certificates and improve the speed and reliability of settling trades. The problem was that this was never simply an IT project. It required changes to technology, regulations, legal processes, and the way brokers, banks, registrars, and other market participants worked together.
I think the biggest mistake was trying to satisfy too many stakeholders at once. Each organization had its own systems, requirements, and preferred way of operating. Rather than establishing a limited first-stage objective and building from there, TAURUS continued absorbing additional requirements. Many of those requests probably seemed reasonable when considered individually. Taken together, however, they created a system that became extremely difficult to design, integrate, test, and manage.
There was also a basic accountability problem. The London Stock Exchange sponsored the project, but it did not completely control every organization that needed to participate. This allowed stakeholders to influence the project without necessarily being responsible for the additional cost or risk their requirements created. The project needed one clearly accountable sponsor, stronger decision rights, and a disciplined process for evaluating changes. Instead, the scope continued growing while the original budget and timeline became less realistic.
I do not think projects usually reach this condition because one project manager makes a single terrible decision. It happens gradually. People confuse activity with progress. Teams complete reports, hold meetings, write software, and spend money, so the project appears to be moving forward. At the same time, bad news is often softened as it moves up the organization. Project managers may be reluctant to challenge executives, and executives may not want to admit that their original commitment was unrealistic.
The amount already invested also becomes part of the problem. Once an organization has spent millions of dollars and attached its reputation to a project, stopping can feel more damaging than approving another delay or budget increase. That is the sunk-cost trap. Everyone hopes that one more change, one more deadline extension, or one more round of development will finally solve the problem.
In my view, TAURUS was less a software failure than a failure of governance, scope control, and honest decision-making. A better approach would have broken the work into smaller stages, required working demonstrations before releasing additional funding, and forced every proposed change to be evaluated for its effect on cost, time, and risk. Project managers cannot prevent every problem, but they are responsible for making problems visible. Their job is not just to keep a project moving. It is also to tell leadership when the evidence shows that the project is no longer viable.
Sources consulted:
Project Management Institute, “Avoiding the Successful Failure”
SWIFT Institute, research discussing the suspension of TAURUS and development of CREST
