When I was in the Air Force in Canberra, there was a famous project called the Supply Systems Redevelopment Project, the SSRP. I knew people on the team.
Unlike the projects I worked on, this one was public. It was in the Canberra Times and before Parliament.
I remember going into the Cameron Offices. The analysis was paperwork, metres and metres of data mapping, up on the windows. It was amazing.
The SSRP was Defence’s attempt to put the supply systems of the Army, Navy and Air Force onto one common system. In 1990 the Canberra Times put the size of the job at more than 1.5 million line items, worth over $5 billion. One steel washer had been found recorded more than a hundred times across the separate systems. Before three systems could become one, someone had to agree what each record meant. That was the paper on the windows.
A decade of studies
The studies began in the mid-1970s. When the project was given the go-ahead in February 1986, the Army newspaper reported that “nearly a decade of preliminary studies” had already taken place. By then it had a 150-strong team from all three Services at the Cameron Offices in Belconnen.
In September 1987, after what the Canberra Times called a “four-year attempt”, Defence brought in Arthur Andersen to provide a cost/benefit analysis of the project.
By July 1990 the same paper said the project had been “around for longer than most of us care to remember”, and that some of its own team had taken to calling it the Tricentennial Project.
I remember it as a project that suffered from analysis paralysis, and as one that was cancelled. Strictly, what was cancelled was the original plan. Because no one else had built a common supply system like it, Defence decided early that the software would have to be written from scratch. By 1990 a commercial package was being trialled instead.
The Canberra Times writer in 1990 said possibly the biggest problem was human relations, between Defence’s computing policy areas and the Services.
In 1992 Parliament’s Joint Committee of Public Accounts reviewed the project. It called its report “A champagne appetite but only a beer income”.
When paper stops being cheap
In Agile never got rid of the business case, I argued for iterating on paper first.
The SSRP shows what that argument depends on. Paper is cheap while each draft settles a question. When an early decision or the politics keeps reopening the questions, the paper becomes the expense: years of salaries, offices and computers.
By late 1994, Navy News reported, the people doing the work had replaced most of the Services’ old supply systems, nearly two decades after the first studies, with work still to finish. That progress matters. Defence had a hard problem to solve, and it would be wrong to call the whole undertaking a failure.
Delivery doesn’t tell you whether the earlier years were well spent. Some analysis goes on because the problem is hard. Some goes on because the organisation can’t agree on a decision. A reviewer needs to know which is happening.
What a reviewer can ask
Is each round of analysis closing a question, or opening the next one?
Does the same question come back in each new draft?
If the questions keep reopening, is it because nobody has decided, or because an early decision or the politics keeps undoing the answers?
Who owns the decision that ends the study, and when is it due?
Want more like this? Rick writes about the go/no-go decision: for the people who assess business cases, and the founders who have to make them.
All writing