Saturday, December 19, 2009

Standish Chaos reports miss the point

The Standish Chaos report of 1995 is probably the most cited work on project success and failure. The latest incarnation is quoted as saying:
  • 32% of all projects succeed
  • 44% of all projects were challenged
  • 24% of all projects failed
where succeeded means 'delivered planned scope on time and on budget'. Scope, Time and Budget are the classic legs of the 'Iron Triangle' ('What about Quality?' 'We don't talk about quality, that's a given.' 'Right.') which every project leader is taught is the holy grail of a successful project. But is that really success?

I teach my Scrum students about 5 perspectives on success, depending on who you are:
  • As a developer: success is fun and learning. A good working climate and personal development.
  • As a PL: satisfy the constraints of the Iron Triangle
  • As the line manager: business units are happy - line management gets to deal with escalations. (Or for a programming shop: the salesperson who is happy when the customers are happy).
  • As the user: "the product helps me do important things better and easier than before. I am thrilled and want to use the product."
  • As the investor/sponsor/paying customer: a good ROI (however defined
Which of these is most underrated? My vote: Fun and Learning. For the reasons why, check out this video from TheFunTheory.com or Dan Pink's talk about motivating people.

Which of these is most overrated? My vote: Satisfying the Iron Triangle. Why do I say this? So many participants in my Product Owner courses, when asked what was their worst professional disaster, responded 'We built the product, but no one wanted it. The whole effort in time and money was for nothing.' Focusing on the project leaders definition of success misses the user and sponsor's view of success completely.

Want your project to be a success?
  1. Figure out who your customers and users are. Make sure they will be thrilled with what you build.
  2. Figure what what the product should earn for the company and what it should cost.
  3. Assume it will produce half of what you estimate and cost twice as much. Is it still profitable? If not, think twice about doing the project.
  4. Figure out what is the minimum set of functionality to get your customers excited about your product. Build that (more or less - you are probably going to learn more about/from your customers and your product along the way, which will change your understanding of what you need to build)
  5. Stay focused on delivering value to your customer. Do Scrum.
Did I say 'Do Scrum?' Yep Scrum helps you do all the rest.

2 comments:

Dave Rooney said...

Peter,

I agree that the criteria used by Standish are somewhat archaic, but the Success/Challenged/Failure Rates are only part of the CHAOS Reports.

What I find much more useful is their weighted Top 10 list of success factors. An example of this, from the 2004 report, can be found on the InfoQ site.

That list matches much more closely with the criteria for success that you've observed (and mine as well, for that matter!).

Dave Rooney
The Agile Consortium

PM Hut said...

Peter,

You can see how project success evolved from 1994 to 2009 in this CHAOS Report.

Things are definitely looking better, success rate has doubled...