IEEE Spectrum has had a number of articles over the last two years regarding the failure of major software/IT projects and the reasons why they fail. One major reason is the lack of a well defined specification for the project. This means a six page max list of requirements as to what the software must do and will not do. They noted one example of a project with an 80 page spec that got down to specifying the color of every button on the page but did not really state what anything did. Next, everyone wanted a different page with a different look and different functions. Allowing different departments or organizations to dictate each page made sure it would be a mess. And then they had different contractors working on different parts reacting to making major changes whenever the sponsor thought about them. And they expected all of these parts to work flawlessly when this whole mess was merged together. To bad the contractors didn't know they each were using incompatible software.
From: email@example.com on behalf of Butner, R Scott
Sent: Wed 9/3/2008 9:59 AM
To: firstname.lastname@example.org; email@example.com
Cc: Butner, R Scott
Subject: Please read if you have ever supported a web/database/information technology project for US EPA
Imagine the setting:
it's the EPA's semi-annual Web Working Group meeting, a gathering of 200 or so project managers, web masters, and contractor scum, err….valued partners from the contracting community -- which convenes every 6 months to discuss the effective use of web technology to get the environmental message out.
A breakout session on the second day of the workshop advertises a session entitled "An EPA Project Manager's Guide to Insuring that your Web Project Will Fail."
The sole presenter mentioned in the agenda is the mysterious "Contractor X." No bio is provided.
Day two of the workshop dawns, and 75 EPA program managers show up, lattes in hand, to find out what this mysterious stranger has to say to them.
Minutes before the session is scheduled to begin, the room lights dim. The mournful sound of a solitary bagpipe drifts from the PA speakers, taking just a few moments to converge on the tune of "Amazing Grace."
Because this is Seattle (and because Contractor X is deeply, tragically hip) it's the version recorded by the Dropkick Murphys. Grunge-laden guitars displace the bagpipes, and 60 seconds of raucous music electrifies the room.
As the last cymbal crash fades to black, a spotlight comes on, focused on the podium. A "solidly built" (yes, it's a euphemism) gentleman of indeterminate age walks into the room, a bag over his head concealing his true identity.
Because it's an EPA meeting, the bag is neither plastic NOR paper, but canvas. Two eye holes cut out where the "w" and "y" once helped spell out "Safeway."
[don't worry, dear reader -- the bag will be put back into traditional service at the end of the day. Turns out that Contractor X is handy with duct tape. ]
Contractor X clears his throat to get the room's full attention, and signal the start of his talk. The abundance of reverb and upper register rattles suggests that the voice has been electronically disguised to protect the mysterious stranger's true identity.
The room is hushed as Contractor X begins his talk -- a litany of "do's and don'ts" -- but mostly, "don'ts" -- for working successfully with Information Technology contractors…….
[/end of scene]
Pretty cool scenario, huh?
So here's the question: what should Contractor X tell the crowd?
I invite your responses, anonymous or otherwise, public or private. I invite responses from both contractors, and from EPA project managers who have learned from bitter experience.
I even invite responses from those who didn't.
Pacific Northwest National Laboratory
902 Battelle Boulevard
P.O. Box 999, MSIN K7-28
Richland, WA 99352 USA