While 59 per cent of mission critical applications in the US are built from scratch, in 1996, some 40 per cent of these projects failed at a cost of $82 billion because they were too ambitious, according to research by consultants at the Standish Group.
In 1994, only 16 per cent of all development projects succeeded, while 31 per cent failed and a futher 53 per cent were "challenged". By 1996, the success rate had improved to 27 per cent, while only 33 per cent of initiatives were "challenged", but many client/server projects were also cancelled as users started evaluating the potential of the Internet.
Speaking at the Fifth Annual ABT Project Leadership Conference in San Francisco this week, Jim Johnson, president of the Standish Group, said: "The best path to success is keeping the project simple, with a minimum of people, and maintaining open lines of communication. Mix this with a good project manager, tools, which are important to keep track of projects, iterative development processes and staff adhering to key roles such as the business sponsor taking responsibility, and your chances of success increase. Optimally, a project should take no longer than six months, have no more than six people on it and cost no more than $750,000."
He added that a 65 per cent reduction in the scope of projects between 1994-96 had had a direct bearing on improved success rates, while cost and time overruns had also improved. This was particularly marked in large companies, which in the past had been the worst offenders for wasting money.
But, after questioning 23,000 US companies on what criteria gave them even a 50:50 chance of success, he concluded that user involvement, firm executive support, an experienced project manager and clear business objectives respectively were top of the list.
Freshly launched 11nm Qualcomm silicon will come with Adreno 612 GPU
Are pinning down the exact rate of expansion of the Hubble constant
RISC OS 5 to form the basis of RISC OS Open after Castle Technology sells to RISC OS Developments
A smartphone maker fiddling its benchmarking scores? That's unusual, isn't it?