The British Broadcasting Corporation (BBC) recently announced the cancellation of a major Information Technology (IT) project intended to update their vast broadcast operations. The project is called the Digital Media Initiative (DMI), was originally budgeted at £81.7 million ($140 million), and was developed to eliminate the outdated filing systems and use of old-fashioned analog videotapes with their expensive archival storage. The BBC is one of the world’s largest and most widely recognized news and media organizations; it is publicly funded and under British government oversight. The DMI project was intended to save the organization millions annually by eliminating the cost of expensive and outdated storage facilities while moving all media content to a modern, digital format. As an example of a large-scale IT project, the plan for the DMI involved media asset management, archive storage and retrieval systems, and media sharing capabilities. The DMI project was begun in 2008 when the BBC contracted with technology service provider Siemens,
with consulting expertise to be provided by Deloitte. Interestingly, the BBC never put the contract out for competitive bidding, reasoning that it already had a 10-year support contract with Siemens and trusted Siemens’ judgment on project development. As part of this hands-off attitude, executives at the BBC gave Siemens full control of the project, and apparently, little communication flowed back and forth between the organizations. The BBC finally grew concerned with the distant relationship that was developing between itself and the contractor when Siemens began missing important delivery milestones and encountering technical difficulties. After one year, the BBC terminated its $65
million contract with Siemens and sued the company for damages, collecting approximately $47 million in a court settlement. Still, losing nearly $20 million in taxpayer money after only one year, with nothing to show for it, did not bode well for the future. Having been burned by this relationship with an outside contractor, the BBC next tried to move the project in-house, assigning its own staff and project manager to continue developing the DMI. The project was under the overall control of the BBC’s Chief Technology Officer, John Linwood. It was hoped that the lessons learned from the first-round failure of the project would help improve the technology and delivery of the system throughout the organization. Unfortunately, the project did no better under BBC control. Reports started surfacing as early as 2011 that the project was way behind schedule, was not living up to its promises,
and, in fact, had been failing most testing along the way. However, although there are claims that the BBC was well aware of the flaws in the project as early as 2011, the picture presented to the outside world, including Parliamentary oversight committees, was relentlessly upbeat. The BBC’s Director General, Mark Thompson, appeared before a committee in 2011 and told them the DMI was definitely on schedule and was actually working already: “There are many programs that are already being made with DMI and some have gone to air and are going to air,” he told members of Parliament. The trouble was, the project was not working well at all. Continual failures with the technology were widely known within the project team and company executives, but reports suggest that these concerns were buried under a flood of rosy projections. In fact, a later report on the project by an outside consulting firm suggests that throughout 2012 the deteriorating fortunes of the DMI were not accurately reported either within management or, critically, to the BBC Trust. For example, the BBC’s own internal project management office issued a “code red” warning of imminent project failure in February that was not reported to the trust until six months later. The CTO, John Linwood, maintained that the project did work and would lead to a streamlined and more cost-effective method for producing media; he did not waver from this view throughout these years. This rosy view hid a deeper problem: the technology was just not working. Different views emerged as to why the DMI was not progressing. To the “technologists,” there was nothing wrong with the system; it did deliver working technology, but the project was undermined by would-be users who never bought into the original vision and who continually changed their requirements. They believed that the DMI was failing not because it did not work, but as a result of internal politics. On the other side were those who questioned the development of the project because the technology, whether it had been “delivered” or not, never really worked, certainly not at the scale required to make it adopted across the whole organization. Furthermore, it was becoming evident that off-the-shelf technology existed in the marketplace which did some of what the DMI promised but which, critically, already worked well. Why, then, was the BBC spending so much time and money trying to create its own system out of thin air? According to a news report, it was not until April 2013 that events demonstrated the ongoing problems with the DMI. During BBC coverage of the death and funeral of Margaret Thatcher, news staff worked feverishly to transfer old archived analog videotapes to a digital format in order to produce footage for background on the life and career of the former Prime Minister. So poorly did the new digital archive system works that it
was reported tapes had to be physically transported around London by taxi and subway system to get to their locations while video transfer work was being carried out by private production companies. All this after nearly four years of working to develop the DMI! The failure of the system during Thatcher’s funeral was the final straw. In May 2013 the new Director General of the BBC, Lord Hall, announced the cancellation of the project and that the BBC’s Chief Technology Officer, John Linwood, was to be suspended pending an external investigation into the management of the DMI project. It was later revealed that a senior BBC manager had expressed grave doubts about the DMI to BBC Chairman
Lord Patten one year before the project was canceled. He had also claimed that there was a “very significant risk” that the National Audit Office had been misled about the actual progress of the DMI in 2011. Other BBC executives had also voiced similar concerns for about two years before the DMI was abandoned. The final cost of the project to the BBC and British taxpayers has been estimated
at about $160 million. BBC Trust member Anthony Fry remarked that the DMI had been a “complete catastrophe” and said that the project was “probably the most serious, embarrassing thing I have ever seen.” Members of Parliament, looking into the failure of the DMI, also had a number of very pointed criticisms of the project, the executive oversight of the DMI, and the operations of the BBC in general. Margaret Hodge MP, Chair of the Committee of Public Accounts, summed up the project in her Parliament report: “The BBC’s Digital Media Initiative was a complete failure. License fee payers paid nearly £100 million [$160 million] for this supposedly essential system but got virtually nothing in return. The main output from the DMI is an archive catalog and ordering system that is slower and more cumbersome than the 40-year-old system it was designed to replace. It has only 163 regular users and a running cost of £3 million [$5.1 million] a year, compared to £780,000 [$1.3 million] a year for the old system. When my Committee examined the DMI’s progress in February 2011, the BBC told us that the DMI was “an absolutely essential have to have” and that a lot of the BBC’s future was tied up in the successful delivery of the DMI. The BBC also told us that it was using the DMI to
make many programs and was on track to complete the system in 2011 with no further delays. This turned out not to be the case. [. . .] The BBC was far too complacent about the high risks involved in taking it in-house. No single individual had overall responsibility or accountability for delivering the DMI and achieving the benefits or took ownership of problems when they arose. Lack of clearly defined responsibility and accountability meant the Corporation failed to respond to warning signals that the program was in trouble.” A lengthy post-project analysis of the DMI’s failure by Britain’s National Accounting Office identified a series of errors that all contributed to the fiasco. As part of its final report, the NAO noted that it was never clear where responsibility lay within the BBC for the
completion of the project; in other words, when concerned parties asked to speak to those “in charge,” it was never clear just who those people actually were. This issue was made more significant as a result of the BBC losing a case brought against it by its former CTO John Linwood. Even though Linwood was fired by the BBC, the judge found that he was not responsible for the failure of the DMI. The NAO report stated that the BBC must make clear who is accountable for projects and define anticipated benefits at the start. Without a clear sense of benefits from a project such as the DMI, it was impossible to evaluate whether stakeholders were receiving value for the work undertaken.
Bad planning, poor corporate governance, excessively optimistic projections, and a cloak of secrecy
regarding the real status of the Digital Media Initiative project all resulted in a very public black eye for one of the most respected broadcasting organizations in the world. It is likely that the causes of the failure of the DMI project will be debated for years to come, but at a minimum, this story should be a cautionary tale for organizations developing sophisticated IT projects.36
1. What does the story of the BBC’s failed Digital Media Initiative suggest to you about the importance of carefully managing not only the project but the “message” of the project? That is, why is “benefits management” critical for project success?
2. Successful project management requires clear organization, careful planning, and good execution. How was the absence of each of these traits shown in this example?
with any paper