The DAOO Loop

Everyone should be familiar with John Boyd’s OODA loop — Observe, Orient, Decide, Act. There are a number of variations (pick a favourite!) but the basic premise remains the same: you gather information before acting.

Witnessing recent transformation projects that have run into difficulties, or foundered completely, I’ve seen people trying to use a twist on this — the DAOO loop.

Transformation projects are complex, and as I’ve discussed before, they should never be driven by technology or implementation decisions. Senior stakeholders — the people who will lose their jobs in a very public way if it all goes wrong — need to buy in, to be heard, to define the requirements that feed into the target we should be aiming for.

(Let’s be clear, I’m not talking about small projects operating with uncertain parameters — those are an ideal fit for Agile processes. In fact, those sort of projects were why Agile came about in the first place.)

This problem appears when — for whatever reason (tight timescales, lack of budget, etc.) — the project ends up being driven by a specific choice of technology, or a specific implementation date.

The usual process becomes twisted around — the initial focus is how to implement a technology, instead of working out what it is that needs to be done, then planning out how to get there.

Processes and requirements end up being cherry-picked to suit the chosen technology, or glossed over because they raise uncomfortable implementation questions.

This drives the DAOO behaviour — Decide, Act, Observe — Ooops.

Addressing this isn’t rocket science, and isn’t new. We’ve got bookshelves of tomes covering this, as well as established frameworks like TOGAF that really drive it home. Yet there are still large, complex projects which gleefully rush into product selection and implementation without the most basic of requirements from the business.

I’m saddened at the number of senior stakeholders I talk to, especially at the C-suite level, who are surprised (and gratified) that you want to know what they want and what they think.

It’s never too late to involve these stakeholders, and you’ll find that things go much more smoothly when you do. Computer Weekly will have less front page stories, but clients and end-users will thank you when you do.