IT Success!: Towards a New Model for Information Technology

1.9. No sacred cows

The fundamental error of reasoning in the traditional IT business model is to assume that software can be conceived upfront like a house, and subsequently scoped, spec'd and signed off for commitment by both client and vendor – and that the documented business benefits will start flowing once the solution has been delivered.

If you're able to accept this fundamental error, then suddenly you can view a whole lot of things which we normally take for granted in an entirely different light. For example, the following statements would no longer appear outlandish (for many readers some of them probably never were outlandish – just difficult to justify under the traditional model):

• You can't ask users to define requirements and specifications to be contractually signed off and cast in stone.

• You can't give such requirements and specifications to an IT department (or a vendor) and ask it to define a budget and a project plan to be cast in stone, and expect it to contractually meet them.

• For an IT project, nothing is cast in stone – the business environment, costs, benefits, schedules and risk can and will change, both during the project and after delivery.

• Whatever is initially delivered on day one can never totally correspond to actual requirements, and will have to be continuously reworked and enhanced over many years in the form of new releases and versions. There is therefore is no such thing as a 'project end date'. Rather, the delivery date of an application represents ...