The rollout of HealthCare.gov, the federal government website through which individuals can sign up for health insurance as part of the Affordable Care Act, has been a disaster. The site has frequently been inaccessible to would-be users, and even when people have managed to get into the system, they've found it to be painfully slow -- with many unable to complete the sign-up process because of timeouts and frozen screens.
Looking for the proverbial silver lining, however, the flawed deployment may serve as a useful refresher course for business intelligence (BI) and analytics teams on how not to implement a large-scale system, particularly when it comes to getting the business and IT alignment right.
Obviously, there are more differences than similarities between the national health insurance site and an analytics or BI system, but a good deal of the problems that have plagued the HealthCare.gov rollout are common to other IT implementations that require buy-in from multiple departments.
A recent Washington Post article outlined the reasons for the bumpy site launch, which included arguments over who should lead the project, disagreements over what kind of qualifications were necessary in leadership positions, and a lack of common goals. For example, the Obama administration was dealing with political considerations, which were very different from the motivations driving the IT team. Lack of communication between the administration and IT vendors also impeded progress.
All too often we hear about how an analytics or BI implementation failed to deliver on its promise because business users were unable to communicate their needs to the IT team or vice versa. The question about who should lead an implementation is another common one. Should the users own the project, or is it the responsibility of the IT and BI staff?
READ MORE ON IMPROVING BUSINESS AND IT ALIGNMENT
Learn how important working together is for IT success
See why better alignment remains elusive for some
Read about why IT and business are still on different pages
It will be worth watching the continuing efforts to clean up the HealthCare.gov mess to see if any pearls of wisdom emerge regarding how to get the IT-business alignment right. Maybe the administration will belatedly build a team made up of some combination of technical nerds and policy wonks that can rescue the project and serve as a model for how to manage successful implementations.
But for now, the main lesson learned here is that these types of situations should be avoided from the beginning. The complications that arise from IT and business departments not working together are so predictable that you wonder why they still happen. BI and analytics professionals would be hard-pressed to make it through a single industry conference without sitting in on at least one session on how to work productively with the sales and marketing teams. Yet the fact that these problems are so frequently discussed suggests they frequently happen.
The specific solution may vary from business to business. Some may want to hire a full-time project manager who has experience on both the business and IT sides. Some BI managers and consultants have suggested that Agile development practices are a good fit for BI projects, as they bring developers into close contact and communication with end users.
Either way, the main point is that project managers should be thinking about the right level of business and IT alignment, and should plan for it from the beginning. After spending millions of dollars on a BI system, a company shouldn't have to scramble to figure how to get that system to create reports that salespeople find useful. That kind of stuff needs to be clear at the outset -- or at least long before the system is turned on. And it may be just as central to the success of the initiative as the selection of the specific BI and analytics technology that will underpin the system.
It's clear that waiting to properly structure and plan a project until after a system is in place is preparing to fail – big time. As we're seeing with the HealthCare.gov rollout, addressing problems only after they've emerged may be courting disaster.