Fear as a root cause for IT failure
For an explanation of why large information technology projects are dysfunctional, look no further than a recent Institute for Defense Analyses report on Defense Department enterprise resource planning projects.
Those ERPs, it's well known, are over budget and behind schedule. A recent Rand Corp. analysis of Navy ERP found extenuating circumstances outside the control of program officials to excuse some cost overruns--but still noted that even without outside business process instability, Navy ERP would have still likely overrun its initial cost estimate.
IDA report authors come up with a number of reasons that have direct applicability to large IT projects in general, including lack of business-side leadership and the heavy burden of legacy systems.
But this in particular caught my eye--an inability by program managers to deliver a factual status report to leadership if there's negative information in it.
To do so "is perceived as weakness in execution," IDA report authors say, and program managers fear that perception could lead to the cancelation of their project.
This fear against negativity also leads to increasingly sanitized status reports forwarded up the chain to the point where, by the time they reach senior leaders, "there are no problems."
Though report authors don't use this exact language, this eventually will result in a "come to Jesus" meeting during which program managers, finally no longer able to hide the true status of the program, confess all. As I've noted elsewhere, attempts to prevent these meetings through earned value management may have even increased their likelihood since project managers can keep indicators positive by postponing difficult problems into the future, where they compound until a difficult task becomes impossible.
The big problem with letting problems build up until they explode is that post-explosion, opportunities for salvaging the program are limited. Even with additional money thrown at a project, its scope most likely will significantly shrink (see FBI Sentinel), meaning that the final result falls far short of the intended functionality. (In an indictment of how bad federal IT is, even this will be portrayed as a "big win.")
Unfortunately, program managers' fear of reporting bad results isn't irrational; they're correct to be fearful for the status of their programs (and their careers) if they churn out reports containing bad news.
This state of affairs is the result of many factors, some of which may be impossible to change, since they include a dysfunctional budget process, a very complex federal contracting system, a dearth of IT expertise in federal agencies, as well as the natural inclination of any large organization to confuse candid discussion with dissent.
High officials within the DoD appear aware of the problem, but their latest move doesn't inspire confidence that they're doing much about root causes. Instead, the DoD appears to be on its way to even further institutionalizing the problem--as of the start of the new fiscal year, an investment review board chaired by the deputy chief management officer will review all DoD business systems worth more than $1 million over 5 years, whether they spend development and modernization or operations and maintenance dollars.
Those reviews will subject "virtually all" DoD business systems to DCMO IRB review--who wants to bet that the board will read a lot of positive-sounding reports? - Dave