views:

347

answers:

10

Just like human beings have vital signs (e.g. temperature, heart rate, blood pressure, and breathing function); what vital signs are essential to collect and track to aid in the evaluation of the overall health of a project?

No clear winner yet based on the voting to date (June 2009).

+1  A: 

Programmer happiness, daily builds, and bug reports/fixes all come to mind right-off.

warren
+1  A: 
flukus
+1  A: 

From a programmer's point of view (i.e., a non-manager), I'm most interested in working features. Tracking progress in terms of the implementation of features is possibly one of the only meaningful statistics that apply; anything else can be too easily gamed.

Jason Etheridge
+1  A: 

Number of user complaints.

The ability for users to report errors is a prerequisite.

JosephStyons
+1  A: 

Bug count.

Matias Nino
+1  A: 

(number of original developers still on the project) / (number of original developers at the start of the project)

;-)

Steven A. Lowe
What is a good score?
Brian
Seriously? So you don't want people to roll off and move on to new challenges? Why should creativity be rewarded by being forever assigned to the same code base?
Brian
@[Brian]: of course not (a) i am assuming a project still in development, not in maintenance and (b) did you miss the smilicon? ;-)
Steven A. Lowe
+1  A: 
  1. Project team confidence that project scope and quality can be achieved in time and on budget. Can be easilly checked by asking casually what people feel (as opposed to think) about current progress and goals. Also listen to and address when people ranting and complaining.

  2. Continuous active support from project stakeholders. Once stakeholders start distancing themselves from the project, do not want to be involved in the decision making process, spend as much of their time as previously, or take responsibility for parts of the project it means things have started turning sour. Successful, worthy projects tend to attract new stakeholders not loose existing. Another check is to keep asking what they feel about the thing and are they happy with the progress so far. Address any unhappiness.

Internal measures such as earned value, velocity, LOC written and tested, outstanding issues, function points completed, degree of deviation from original plan etc are all useful for monitoring the health; however neither of them tells whether project is going to survive. The two main things that are required, in my experience, are active support from the project sponsor and other external stakeholders and team's confidence that the thing can still be achieved.

Totophil
+1  A: 

The Iron Triangle... scope, schedule and budget. Focus on just one, and you've got a problem.

Mark Nold
+1  A: 

My suggestion probably sounds strange, but if you think about it, its a pretty good indication of a healthy project. How good are your unit tests?

Think about it. It means: 1) you HAVE unit tests 2) you probably have requirements 3) you probably have a design 4) your code is broken up into "units" that can be tested 5) your units are defined well enough to be tested

AND

6) you will have an immediate feedback on future changes

dviljoen
+1  A: 

You want to measure the progress of the project in terms of value delivered to the (prospective) customers so I would include a measure for that in your vital signs.

One that comes to mind (we are using scrum and using story points to measure feature size)

story_points_completed / remaining_story_points

During a project both will grow: story points completed due to work done by team members, remaining story points due to new or changing requirements. What you want to monitor is that story_points_completed / remaining_story_points is increasing and not decreasing.

Jeroen Huinink