I had lunch with Tony Virdi from Atos Origin this week. I know Tony as he is a fellow director of the National Outsourcing Association and so we see each other at the monthly NOA board meetings, but we had decided it was time to catch up over a slightly longer meeting where we could exchange a few ideas.

I knew that Tony had wanted to get some of my views on the industrialisation of the IT industry. This is a familiar refrain to those of us who have been involved in software development in the past. The ‘software factory’ was often mooted in the past as the next logical progression as software development languages moved ever closer to natural language, yet it is a concept that is now only just approaching reality.

Tony described some ideas to me. He asked me to imagine all the usual logical steps in an IT process lifecycle from left to right as columns; analysis, design, specification, coding, testing, cutover, support and so on. He felt that we are starting to approach a level of maturity where the input and output from each column is beginning to be clear. Overlaid onto these columns though would be the various subdivisions of the IT sector, such as software development or application maintenance, creating a grid of IT ‘components’. The entire grid would represent the industry as a whole.

What someone expects from a business analyst can then be more clearly defined and should function very much as a black-box component – something familiar to software developers. And the same expectations can be made for testing; it should be possible to create a test factory that can test any software, provided the standard inputs are all satisfied.

Some of these concepts, such as the industrialisation of the IT industry, have been discussed at length before – probably even before I was born – but the difference now is our experience of outsourcing. Companies that have experimented with outsourcing have found that the process-driven nature of the sourcing process does introduce efficiency. Even those companies who experiment with outsourcing and then decide to bring processes back in-house continue with some form of measurable internal market – they don’t discard outsourcing in favour of the chaos they used to have. It creates defined inputs and outputs, just as Tony describes in his industry grid model.

So I think Tony is onto something with his component model. What is even more interesting though, is how Tony described the potential issues related to skills in countries such as the UK. If you create a component-based model of the entire industry in this way then you can observe where different geographies can add value. It should be possible to view the industry model with different areas shaded out, meaning they are of less value in this geography. Though the selection of component specialists is going to be based more on company rather than country, it is likely that some regions will specialise in particular components – and UK higher education institutions need to be aware of this. If they are not then we could find the doomsday scenario of highly-trained people leaving university with the wrong skills at the wrong time in the wrong location.