So I was listening to Sirius XM’s broadcast of CNN the other day while driving in my car, when I heard yet another report on the troubled HealthCare.gov site, which is used for people to sign up under the U.S.A.’s Affordable Care Act. That’s what the political pundits call Obamacare.
Without getting into the political by-play over this massive revision of the health care system in this country, let’s limit the discussion to a site that simply doesn’t work. No ifs, ands, or buts, and President Obama has admitted as much, while promising fixes. Whether those fixes come soon enough, and are done correctly, is another issue beyond the scope of this column.
What really concerned me was the almost mindless fashion in which CNN’s reporters attempted to deal with how Silicon Valley might fix the site. So with all the highly sophisticated portals out there, such as Google, Face-book, or even Apple’s iCloud, where did they go? Why to Matt Mullenweg’s Automattic.
Did I say Automattic?
Well, in case you don’t know, Matt Mullenweg is the inventor of WordPress, the world’s largest blogging platform, and Automattic is the company established to manage that property, plus a number of other services, such as Akismet, which handles comment spam, and Gravatar, a site that manages your online avatars.
If you’ve ever tried WordPress, you know that it’s open source and free. Yes, you may pay to set up a site hosting WordPress, but millions simply use the free blogging portal at wordpress.com. Indeed, one in five sites are built with WordPress, so maybe CNN assumed that would be an ideal source for advice on how to fix the health care site. They also compared the $30 million in venture capital received by Automattic over the years to the hundreds of millions of dollars it took to build the failed health care site.
Now someone who doesn’t really understand technology might find this an apt comparison, a prime example of government waste. But it’s not. For one thing, WordPress is, at its heart, a lightweight blogging platform. There are plugins that add limited commerce and content management features, but they are still relatively easy to set up and use.
The health care site is a highly complex commerce site that is designed, at its core, to help you buy a health insurance policy. That may seem simple enough, but the site also has to interact with other sites, government and private, to make the process work. So there must be an active connection with the IRS and Social Security sites, for example, and the individual health marketplaces or exchanges in the some three dozen states that didn’t build their own. That’s hard enough, but once you actually submit an application, the data has to be sent to your chosen insurance carrier, with whom you complete the transaction.
Each of these services has their own organization and setup, and the sites are coded differently. Just communicating successfully would probably be as close to a miracle as you can get. Add to that the process of setting up, verifying and managing an account, and you can see where problems can easily arise even if everyone got involved with good intentions and did their jobs properly. After all, the U.S. Centers for Medicare & Medicaid Services, which manages the Obamacare site, cannot change the programming of all the other sites with which they communicate and exchange data. They have to cope.
So the comparison between Automattic and HealthCare.gov is not just inappropriate. It’s totally wrong. They are designed to handle totally different functions. More to the point, government-run sites are often designed by large firms that get the contracts by submitting the lowest bid. Having the best idea may not win out, if the company who presents the proposal charges a lot more money. It’s not the same thing as choosing an iPhone over a cheap smartphone from another company.
What they ended up with was a monstrosity that wouldn’t have survived a load test, or even a traditional beta test to evaluate how well it worked. The design is also backwards if you apply even basic logic to the setup.When the site first debuted, you couldn’t simply check out insurance plans and rates in your locality until you actually applied for coverage. Talk about walking backwards. However, the site was revised just this week to allow you to actually browse the offerings before signing up.
I would like to think that the performance problems with the Obamacare site will be fixed in a few weeks. I also understand that applications can be taken by phone or in person for those who just don’t want to confront a badly designed site. I would also hope that the people in charge of the department that created this godawful site will be given their walking papers once the fixes are in place, but don’t bet on it. And as I said, a poorly designed site has little to do with whether the underlying service has any value.
It’s also true that, even the properly coded sites have problems. I’m writing this just hours after a major Face-book failure, and there have been notable failures at sites run by Amazon, Microsoft and, of course, Apple. Yes, the health care site had more visitors than the developers anticipated, but that’s no excuse for poor design.
As for CNN, their inability to pick the right Silicon Valley company for advice is all-too-typical of mainstream journalists. They really ought to have people on staff who do understand how to cover such a story. The eternal quest for ratings, however, means that anchors and reporters are hired because of their appearance, performing ability, and the ability to get ratings, rather than their actual knowledge on how to do their jobs.