BBC Digital Media Initiative

~ the last words ~

I have never worked at the BBC and I have never touched a broadcast asset management system. What I recognize in the Digital Media Initiative is the shape of it, because most engineers who have spent time inside a large organization have watched a smaller version go the same way: a program with a strategic name, a supplier picked because there was already a contract with that supplier, a rebrand in the middle, and at the end of it nothing anybody uses. The BBC's version cost ninety-eight million pounds and finished in a parliamentary committee room.

It began in February 2008. Ashley Highfield, then director of BBC Technology, promoted it, the BBC Trust approved it at £81.7 million, and the work was contracted out to Siemens with Deloitte consulting alongside. The goal was stated as a fully tapeless production workflow: everything the corporation shot would be ingested digitally into a media asset management system, then edited and moved around without a physical cassette touching anybody's hands. It was forecast to save the BBC about £18 million.

Siemens had not won a tender. Erik Huggers told the Public Accounts Committee on 15 February 2011 that the company was selected without one because the BBC already held a ten-year support contract with it, and that handing the risk over had produced what he called a "rather distant" relationship with the supplier, which made it hard to tell whether a milestone had actually been hit. The contract was terminated in December 2009. Siemens paid a £27.5 million settlement against estimated BBC losses of £38.2 million, leaving a net hole of £10.7 million, and the project came back in house under a new name: Fabric.

Fabric is a nicer word than Digital Media Initiative and it did not help. The National Audit Office criticized the BBC's handling of the project in February 2011, parts of the work went back out to a consortium of IT companies that June, and the spend between 2010 and 2012 came to £98 million.

~ the record ~

~ the Thatcher tapes ~

Margaret Thatcher died on 8 April 2013, and the BBC did what every broadcaster does when a former prime minister dies, which is reach for decades of archive at short notice and cut it fast. The archive did not come out of DMI. News staff worked off videotape, and some of the transfer work had to be sent to outside production companies because the system built to make exactly this possible could not do it.

A few weeks later it emerged that tape editing equipment might have to be installed at Broadcasting House, in rooms that would need special cooling, because the machines run hot and the new building had been designed on the assumption that nobody would need them. That is the sentence that should have ended the project, and roughly speaking it did.

~ six weeks in ~

Tony Hall became director-general on 2 April 2013. On 24 May, six weeks later, he canceled the Digital Media Initiative rather than, in the phrase that followed the story everywhere afterward, throw good money after bad. The same day, chief technology officer John Linwood was suspended pending an external investigation. His contract was ended that July, which the BBC did not confirm publicly until 24 January 2014.

Then came the part where everyone discovered how long the problem had been visible. In June it was reported that a senior manager had taken serious doubts about DMI directly to the BBC chairman, Lord Patten, a full year before cancellation, and had suggested the National Audit Office was given a picture of progress in 2011 that did not match reality. Other executives had been saying similar things internally for about two years. The Public Accounts Committee held a hearing at MediaCityUK in Salford on 10 June 2013 and accused the corporation of misleading Parliament.

~ one show ~

The committee reported on 10 April 2014 and the language was not hedged. Margaret Hodge, its chair, said the public had "paid nearly £100 million for this supposedly essential system but got virtually nothing in return," and the report itself used the words complete failure.

Two findings in it are worth keeping. The first is that the only usable thing to come out of five years was an archive catalog and ordering system that took around ten times longer to use than Infax, the forty-year-old database it had been built to replace, costing £3 million a year against the old one's £700,000, with 163 regular users. The second is that in the entire life of the project the BBC made one television show using DMI, a popular science show called Bang Goes the Theory.

The committee's other conclusion was the structural one, and it is the reason this story is worth reading if you work anywhere with a program board. No single person had overall responsibility for delivering DMI or for taking ownership when things went wrong, so when the warning signals came, there was nobody whose job it was to act on them. The committee said the BBC Trust had been just as complacent in overseeing the executive that ran it. Of the ninety-eight million, the archive catalog was what remained, running at more than four times the cost of the forty-year-old system it had been built to retire.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »