When Technology Fails

This kind of debacle, the critics snorted, is just what we’ve come to expect from government technology projects. (Remember the repeated crashings and burnings of ambitious IRS system upgrades before that agency finally seemed to get it right?) The ultimate snort, of course, centers on the question of how the various intelligence agencies can ever hope to connect dots dispersed among their respective silos if a single key player like the FBI can’t even connect the intrasilo dots to enable sharing among its own agents.

Setting aside for a moment the worthiness of that last question, the sorry truth is that technology projects fail in high percentages everywhere, regardless of whether they have public- or private-sector roots.

That failure rate has been tracked for years by The Standish Group, a research firm based in West Yarmouth, Mass. And it’s not a pretty picture. Numbers from 2003 suggest that only 34 percent of the projects evaluated in Standish’s research succeededand that represents a vast improvement over the group’s first survey, nine years earlier, when only 16 percent succeeded in meeting their goals. Of the projects tracked in the 2003 report, 15 percent were flat-out failures, but a disturbing majority (51 percent) fell into a category that Standish calls “challenged,” meaning that they were over budget, took longer than promised and lacked some critical capabilities. That means that two-thirds of all projects fall short, to varying degrees, of fully satisfying their requirements.

Projects tank for many reasons, but in a short article on the Softwaremag.com website posted early last year, Standish Group Chairman Jim Johnson attributed most failures to faulty project management. And the bigger the project, the harder it falls. Virtual Case File is a cruiser-class system.

My colleague Allan Holmes, Washington bureau chief for our sister publication, CIO, has followed government technology projects for many years. And he notes that there are certainly some factors peculiar to federal bureaucracy that make matters more challenging, including the frequent turnover of political appointees (and their clashes with career employees); the byzantine procurement policies that enforce a glacial decision-making pace (and limit the universe of vendors willing to bid on work); and the fabled inflexibility of civil service workforces. Nonetheless, Allan believes, the government is no more inept than the private sector when it comes to deploying IT systems. It just feels that way. “The government is more accountable. Every failure ends up in the public record,” he points out.

So, for IT projects, success, at 34 percent, is slightly better than new Baseball Hall of Famer Wade Boggs’s lifetime batting average (.328). In baseball, .340 would be sensational, but in business, not so hot. And yet that’s pretty much the norm.

And this is a good thing to keep in mind when we start pinning our hopes for the early detection of planned terrorist acts on vast information-sharing systems. If the FBI can’t get it right within its own puzzle palace, what is the likelihood of an IT-powered success that cuts across all of the puzzle palaces, both here and abroad?