Fat versus Lean
A new research report coming out soon is titled Money Pit: the true cost of a project. In this advanced research report we are using Fat and Lean for the names of the two projects profiled in the report. Let us first define what we mean by fat in the context of this paper. Fat to us is the part in a project that provides little to no value. Many people justify larger projects based on claiming the project is feature “rich” when the project is full of marginal functions and features. Now, Lean is the area in a project that provides value. Lean is also used for the process, which for Lean is the agile, Scrum method.
Fat does not mean that everything is predictable and known. However, that is the hope for using the waterfall process. In this paper we show Fat is full of lots of process bloat. We are trying to change the discussion around from that the project is Feature/Function “Rich” to Feature/Function “Fat.” This is a new initiative for us. If you look at CHAOS Manifesto 2014 on page 14 the chart compares success driven verses value driven. Here we used “Rich” to describe Success Driven Feature/Function, but now we think this is too nice a word for the meaning. If you look at the new Factors of Value report on page 7 you see an updated chart with the word “Fat.”
Now lets consider that CGI bid $97 million for healthcare.gov based on some type of formula such as function points or KLOCs. The fact it cost somewhere between $600 million to $1 billion and took 3 to4 years is most disturbing. In contrast four guys in garage in Northern California wrote healthsherpa in one year. The cost was around $600 thousand to a million. Heathsherpa is very similar to healthcare.gov. This is enormous gap, but not uncommon. Another interesting project in the Commonwealth of Massachusetts was a housing application. The Commonwealth went out to bid and the lowest bid was $12 million and two years. They decided to try hiring 3 MIT students – they completed it in 9 months for less than a million dollars.
Now you might consider these projects abnormalities, but they are in fact, very common. They are also two more examples of fat versus lean. We just introduced Stanmets, our cost normalization system for sizing. The Standish Adjudicator assigns Stanmets by looking at the whole project profile with the 25 basic elements, which includes skills, types, methodologies, industries, and maturity levels. We have over 100,000 projects as comparisons. Before we created Stanments we could not see this in our database. Now we see this all the time in our database. Same project - enormous gaps, double, triple. 5-fold, 10, 20, and even 100 times. If you want to read Money Pit become a member. This research report alone is worth price of admission.