Skip to content Skip to navigation

Waterfall or Agile?

August 22, 2008
by anonymous
| Reprints

Just the other day I founds myself in the middle of a large conference room, at a meeting where the CIO was facing the COO, CXOs and a team of consultants from a strategy firm. The strategy firm had been engaged to tell the hospital system what was the state of its IT and what could make it work better.

Interestingly, one of the observations the consultants made referred to an absence of methodology. It got my antenna up.. It was clearly a legitimate question in a large development shop. But in a hospital system where most projects are rather small in size, the question seemed more from the book the consultants may have read rather than an experience of the hospital systems development environment.

That said, this meeting was not alone in raising that specter. I have been witness to these arguments across IT organizations. Most of them had a sizeable IT budget, however. That prompted me to take a sample out of several projects that go on at any time in hospital IT systems. Not surprisingly, over 80 percent of them take less than 40 hrs of efforts at the most.

Methodologies were designed for really large projects where it was far from easy to keep a view of the project for any single person. At under 40 hrs, they may find it tough to justify and implement standard project management practices, let alone a methodology overarching it.

So I thought of asking our readers, what do you think hospital IT organizations should do? Invest in old, huge, unwieldy, expensive methodologies? Or should they be looking at some lower order management systems that yield the results they need to deliver? Is it waterfall? Or is it Agile? Or something in between?

Topics

Comments

Joe, So true!
Waterfall worked when the wisdom of the development team could prevail! Or they had absorbed enough and reviews were mostly about conformity. These were large projects, mandated, huge RFPs and were needed at a time when we had to build the arteries that make up teh technological eco-system.

Now, most of the effort goes on to address issues that are dynamic, as they are happening and typically a good number of projects have a tendency to become a little "wicked"! Moreover, in a small project, waterfall does not work. If agile slows it down by the order of magnitude you seem to suggest, clearly bringing in better project management practices may greatly help!

Thanks so much..
Where are you based? I am in Washington, DC area and you may reach me on 301 841 7422.

let's keep the conversation public :)

Satish,

I think your focus on time-to-anticipated completion is appropriate.

How well the requirements are understood and the social dynamic context are the other two dimensions that come to mind as very important. So, how 'wicked' is the underlying problem being addressed ( http://cognexus.org/id42.htm ) is an important question.

If there a moderate likelihood that the understanding of the problem will change, and/or priorities will need to be changed, then a waterfall approach will fail. It will also fail if conflict management is absent.

Many hospital IT projects I've seen done successfully had a very talented (analysis skills) person both managing and doing most of the work. These were often of the 40 hour variety, but the 'iteration tail' work often approached 200 hours and 4-6 months of calendar time.

It should go without saying, but, independent of methodology, good, strong project management is like oxygen. You really miss it when it's gone!

Satish, Thanks for raising one of my favorite topics. IMHO, the question boils down to this:

Is the Project Wicked?

(For our readers, 'wicked' definition below comes from here: http://cognexus.org/id42.htm)

"A wicked problem is one for which each attempt to create a solution changes the understanding of the problem. Wicked problems cannot be solved in a traditional linear fashion, because the problem definition evolves as new possible solutions are considered and/or implemented. The term was originally coined by Horst Rittel.

Wicked problems always occur in a social context the wickedness of the problem reflects the diversity among the stakeholders in the problem."

The tools of Agile recognize that understanding of requirements change, as do the priorities of the stakeholders. Generally speaking, Waterfall assumes the opposite.

anonymous (not verified)