The Wall Street Journal ran a story today about the U.S. Customs and Border Protection department deciding to put off their grand plans for building a futuristic “virtual fence” between the U.S. and Mexico. The fence is being built by Boeing and was originally designed to cost $20M but has become much more expensive over time: the project is currently priced at $8B and could triple through 2013!!!
Ok fellow product managers, I think that we can all agree that something has seriously gotten out of hand here. What’s interesting is that it didn’t seem to start so badly. When Boeing won the original contract, they had stated that they intended to create “…a network of off-the-shelf technologies that would be a powerful surveillance system that could rival what the military uses.” That sounds reasonable, doesn’t it?
Things started to go bad awhile ago. The virtual fence system was going to be tested on a 28-mile stretch more than a year ago; however, even this piece of the project is not yet operational. Now the virtual fence’s debut is scheduled for sometime in 2009.
So what went wrong here? There are probably lots of factors; however, from a product manager perspective I suspect that we can probably put our fingers on a number of the key ones. First, it looks like the original product was to be created by a relatively well defined $20M project. That was probably a manageable product. I’m going to guess, but I’ll bet you that it was for a proof-of-concept and that it would have resulted in a small test bed (maybe a 1 mile fence?) Clearly, something happened to cause this initial project to grow out of control. I’m going to blame money and time.
Somehow the U.S. Customs and Border Protection department seems to have gotten their hands on a LOT of cash if they can now fund an $8B fence. I’m going to guess that they got some funds and then went back to Boeing and said “Let’s add some bells and whistles”. Oh, and let’s make the pilot project 28 miles long. Apparently nobody on the Boeing team (hello Mr./Mrs. Product Manager!) stood up and said “Let’s complete the first pilot before we start adding all of this additional work.”
The issue of requirements probably deserves a quick discussion also. If a project has grown from $20M to $8B, do you think that the requirements are all up-to-date? Based on the fact that there appears to be no phased delivery of the virtual fence, just a planned “Ta-Da” moment when it all gets delivered, I’m going to predict that there are multiple versions of requirements for different parts of the fence and that they don’t really fit together very well. Just wait until it comes time to get the different pieces of the system to work together. Oh yeah, I guess they already know that – that’s why the project is over a year behind schedule.
I could go on, but I think that you get my point. If I had to make one more guess, I’d be betting that Boeing has cycled through Product Mangers on this project more than once. In fact, I’ll bet there are a bunch of product managers working on it who keep turning over. We’ve all seen things like this happen and should know how to spot the warning signs: poor or too many product requirements, pilots that get abandoned in favor of more lofty goals, sudden funding that must be used NOW that causes the project to grow to quickly, etc.
If you ever find yourself working on a product that starts to show any of these characteristics, then perhaps you might want to remember the Taco Bell jingle and make a run for the border…!