In the United States we are on the brink of the American football season. America’s team, the Dallas Cowboys have a new stadium with a new high definition scoreboard that is 60 yards long, 24 yards high and has 30 million bulbs (it’s Texas huge). Unfortunately, in the first game the scoreboard became part of the game when a football hit it during a punt. In fact during the pre-game warm-ups it was clear the scoreboard is so low that when a punter kicks the ball there is a potential it will hit the scoreboard.
At a cost of $40 million dollars you would think that the scoreboard would have been installed so that punts would not hit it. The owner of the Dallas Cowboy’s Jerry Jones said "they did all kinds of analysis and the scoreboard meets the NFL’s specifications."
Sometimes Analysis and Specifications are not enough
As a project manager the question you should have for Jerry Jones is that with all the expert analysis and specification review... "did you take a punter out onto the field and see if he could hit the scoreboard when he actually kicked the ball?"
Regrettably, this scenario occurs in project management all the time. Expert analysis and specifications indicate no problem; but when it is time to go operational all kinds of problems show up that were not part of the analysis or specification. I worked for a NASA executive that had one plaque that sat in isolation on his large desk. When he retired I asked him for the plaque and he gave it to me. The plaque reads:
“One Test is Worth a Thousand Expert Opinions”
Everyone usually agrees testing is a good thing. The challenge is that testing is sometimes expensive and time consuming. Additionally, if (more likely when) a project gets under schedule or cost pressure testing is something that everyone looks to cut to solve the immediate problem. This often happens to a degree that you cannot find rationality anywhere you look. You could fill a book with the excuses and rationale that justifies eliminating or scaling back testing.
When in doubt, test. When not in doubt, test.
When the Space Station that is now in orbit was originally under construction there was a powerful group of people that did not want to do integrated testing on the ground. They justified this position by using the logic that there are interface control documents, rigorous specifications and change control. Integration testing on the ground was hugely expensive. Fortunately cooler, more logical heads ultimately prevailed and the price was paid and the testing was conducted. If you live in the real world I don’t have to tell you they found lots of unanticipated problems that would have been remarkably difficult (and embarrassing) had they occurred in space. As is usually the case…
Testing is cheap compared to the alternative
Don’t be surprised when stakeholders don’t think of testing, say there is no time for testing or don’t want to pay for testing. You must do everything in your power to preserve testing to ensure the integrity of your deliverable. This includes selling the value of testing to stakeholders at the start of the project. The earlier you can incorporate testing into the schedule the better.
If they choose to cut testing in any way make sure they understand the risk, they accept the risk and they are accountable for the risk. (Note the emphasis on they…not you.) Remember, in project management they try to take everything from you and then they give you all the blame when something is not right. They are not evil, that is just the way the system works. Forewarned is forearmed.
Copyright 2009 SEBA Solutions Inc. firstname.lastname@example.org