Search and Rescue for Troubled Initiatives

What's gone wrong with

Let’s face it, when corporate IT projects fail they usually don’t end up in the news – but when politics and public spending are involved, they often do. Managing politically charged government IT initiatives is rife with risk. Recently, we’ve seen coverage of three states with troubled  e-government projects by Deloitte, and a Xerox software outage that halted electronic access to food stamps and other aid for shoppers in 17 states. And, the launch difficulties of have become a national fixation.

I’ve run three high profile IT initiatives efforts in two states, and I have some first-hand knowledge of what it is to be on a scaled-down version of the hot seat that Health and Human Services Secretary Kathleen Sebelius is on today. While failure rates of IT initiatives are a concern in industry, there are factors that make government efforts considerably more complicated. Often the details of services, their launch date, budget, and how they’re operated are set in stone. And too often government officials don’t have operational details of a project finalized until well into implementation. This leads to “all or nothing” situations where firms would normally hedge their risks by reducing scale, adding testing or not kicking off until an initiative is well defined.

As a consultant who has taken on troubled projects, there’s an art to telling the difference between a salvageable project and one that needs to be shut down faster than a nuclear reactor about to go bad. At the moment there’s a chorus of pundits providing opinions – some more expert than others – of what went wrong.

Specialists tend to see problems expressed in whatever lens they use to see the world. And the farther away one stands from any endeavor, the easier it appears. Want to respect NBA players? Watch them work from the front row.

The key question is not what went wrong, but what is practical to get it back on-plan. And based on that, what, if any, efforts should create alternative solutions through the web or other channels?

Here are the risk buckets I’d consider to figure that out.

First, there’s the inherent risk of technical invention. I’ve managed initiatives that were plagued by underlying issues from providers, and that worked well in the lab but were glitchy in the field. To determine if recovery is practical the managers need to look at how structural the flaw/flaws is/are and how short the runway is to get it right.

Second, there’s the risk of market adoption. Something can be a great deal, but if the sales process is difficult, or the marketing doesn’t feel credible, and especially if it’s an elective service that can be ignored for a little while – then even good offerings will die on the vine. There are non-technical solutions for this sort of problem: One of my innovative yet-still-hard-to-use programs had a “lead pipe guarantee” from that state’s governor. He gave us this by telling our market of 800 or so school districts that their state funding would be 100% dependent on their use of the system. This was one of those rare cases where managing motivation and culture was more important than design.

Third, there’s procurement risk which comes from the systematic selection of the wrong vendors. At first there was concern that CGI Group, the Canadian firm which built, was one of those wrong vendors because of some earlier legal issues. The jury is still out on that – as reports of the site never being beta tested and huge last minute changes required by the government suggest there is a lot of blame to go around.

Reasons for Optimism:
There are three things (all technical risks) that would doom this project: internal data corruption, inaccurate price quoting, or a security breach. So far, these have not happened at scale. Reportedly, 10 million users hit the site on its first day – exposing latency, user interface, and capacity issues. These can be addressed with an amount of brute force. Further, people have until December 15th to register – and they don’t have to register through a website. Phone centers, in-person services at hospitals, or mobile registration (as is done for voter registration) could all be contemplated.

Reasons for Concern:
For every dollar spent on the front stage of an enterprise web system, two to three must be spent backstage to make it work. If the integration between and insurers isn’t working, or if the data which the site must pull in isn’t ready for use, that will be difficult to remedy quickly.

There is a lack of transparency that leaves outsiders to assume, or assert the worst. Public relations of this failure have been handled badly, and the arrival of Jeffrey Zients as the high-profile manager of this project rescue is the right first step.

Meanwhile, political factors are kicking in. It would seem there is the equivalent of a “keep home the vote” campaign going on with Obamacare. Some conservative groups are buying media to warn young prospects away from enrolling. If younger consumers who are still on the fence about buying coverage sit this round out and pay a one-time penalty, while only motivated sick people bother to register, then insurers may end up with too few healthy members to offset the costs of less-healthy enrollees.

But at this point the question really is no longer about people getting healthcare, or the more tactical question of how to fix a website. The focus is fixing mechanisms can quickly shift to re-litigating the Affordable Healthcare Act. There will be a time of accountability for this snake-bit web launch. But making that a disruption to fixing the service launch isn’t how projects get set right. It’s how policy gets diverted through political theater, which I suppose is yet another risk unique to e-government projects too.

Dave Wieneke directs the digital strategy practice at ISITE Design