I've recently been working with an international project abroad with an interesting challenge, No requirements as suggested in TechnoVision from Capgemini.

The problems are well known. Lets say that there are improvements to be done.

We have a very clear management decision about to use best practices within the industry and standard systems unless it gives a very clear competitive advantage. Take my word for it, you are not going to solve their IT-issues within time if you start development of a new bespoke system.

We have the business need defined, i.e. what you need to do improve your business. We are also defining KPI's and capabilities for the business within the project. But we are don't define IT-requirements.

The hypothesis we have is that standard processes and standard IT-capabilities will be sufficient in most cases. This will be a huge change compared to a heavily modified COTS system that is in use today in several places.

We will now run workshops with stakeholders and users to verify our hypothesis. Any deviations will then be handled as constraints, issues and risks in the final evaluation of the solution.

We are planning to run the workshops with the process owners. The input is based on business events and their existing product groups so this will be real word examples and not high-level fluffy stuff.

The outcome of our part in this project is a transformation plan and an initial estimate for a COTS implementation be be part of a decision material for an implementation at a later stage.

This is no requirements architecture. To succeed, you need a reference architecture for the associated type of business and architect who can understand when to use the standard or not.

Disruptive architecture