Those who work in IT are accustomed to being invisible most of the time. Users expect the technology to just work so that they don’t have to think about it and when it does work, they don’t think about it. When it doesn’t work, it’s a whole other story, usually an unpleasant one. But just imagine what life would be like if your latest failed IT project was on the front page of newspapers across the country.
One failed project is getting exactly that kind of attention—Healthcare.gov which collapsed when millions of Americans tried to sign up for insurance under the Affordable Care Act. It’s been an uncomfortable few weeks for Marilyn Tavenner who leads the Centers for Medicare and Medicaid Services which was responsible for HealthCare.gov. As she struggles to awaken from her current nightmare, consider the lessons any IT leader can draw from these events.
1. Sometimes you just can’t win.
HealthCare.gov never stood a chance. First of all, at this scale, only about 6 percent of projects succeed, according to research by the Standish Group. But even leaving that aside for a moment, consider the challenges even a perfectly built site would have faced on its first day, when seven times as many people as had visited Medicare.gov at its highest traffic peak all tried to access the new site at once.
With all these factors against it, the site crashing was a foregone conclusion.
2. You should take the blame anyway.
When you’re in a no-win situation, resist the temptation to place the blame on someone else, especially when that someone is less powerful than you. That’s what Tavenner did when she gracelessly told Congress that her office works with contractors and that the problems occurred because “a subset of those contractors for HealthCare.gov have not met expectations.”
Phooey. As we’ve seen, problems were inevitable, but even if they were caused by contractors, the head of IT is responsible for the contractors IT chooses to employ. Pointing the finger at them, or anyone else, just makes you look weak and won’t get you out of trouble anyway.
3. ‘One throat to choke’ may be a good idea.
Tavenner’s agency was coordinating the work of 55 different contractors on this project, and has not designated any of them as the project lead. According to a New York Times report, “numerous people involved in the project said the agency did not have the expertise to do the job and did not fully understand what it entailed.”
That may be one reason this behemoth weighs in at half a billion lines of code.
4. Whatever you do, don’t panic.
Obama has pledged that the site will be fixed by the end of November. Reportedly, he wanted it fixed by the end of October but was told that wouldn’t be possible. The end of November may not be either, as some anonymous sources have told the Times that the troubles run much deeper than mere log-in difficulties.
Now, observers are sounding warnings that rapidly fixing the site will inevitably leave it open to attack, and obviously it’s a juicy target. Adding to the concern is a leaked memo warning that full security testing was not completed before HealthCare.gov went live.
The best course now is to make sure security is well in place, even if it means missing that end-of-November deadline for solving all problems. If some users still find glitches after December 1, it will be annoying but soon forgotten. If the site gets hacked, that’s a whole other nightmare