Digital Service and agency IT: Advice for Mikey

Last week, FCW and others reported on a White House announcement of the establishment of a digital service team -- people with commercial IT experience, from outside the Beltway and running out of U.S. CIO Steven VanRoekel's office in the Office of Management and Budget -- to provide advice and consulting to agencies on executing IT projects. Then on Monday, I was surprised to see in the business section of the New York Times an interview with Mikey Dickerson, a 35 year-old ex-Googler (and IT person for the 2012 Obama campaign), who will be running the team with the title of deputy CIO. It is unusual for the Times to get into the weeds of federal IT, so the interview caught my eye.

What struck me most about the article was that this outsider's perspective on how government must change to improve its IT success record is very close to what many people inside government have been saying. The summary version of that message? Go agile and off-the-shelf.

"Having a multiyear project with no checks along the way and the promise of one big outcome," Dickerson says in the interview, "is not a highly successful approach, in or outside government." Responding to the reporter's question about whether agency systems must " deliver complex information or payments to millions of people across the country starting on Day 1," Dickerson stressed that scale does not preclude iterative development.

"It may take two years to do a government project, but you have to check in on it, find ways to see if there is a problem," he said. "If you, a contractor, have to deliver some smaller thing in four to six weeks while the system is being constructed, you'll act differently."

Dickerson's other theme was off-the-shelf. "When possible, use things that have already been invented. Use off-the-shelf products, not custom software. There was things that work in the private sector; there's no reason we can't use these in government too."

The big challenge for Dickerson and his team will be political, though probably not of the sort the Times reporter suggested by noting that Dickerson's team was outnumbered "30 to one" by "lobbyists for existing contractors used to big-ticket, long-term contracts." The question had a nice populist ring (and Dickerson unfortunately took the bait), but lobbyists have little influence over agency acquisition strategies. Small agile spurts don't mean small contracts, but rather many task orders on big contracts. And many if not most government IT vendors are smart enough to know that it is in their own long-term interest for the government to buy IT intelligently.

Instead, the political challenge to Dickerson involves resistance to "outsiders" and to OMB (as a government outsider) from within the agencies. This was in fact signaled by one comment posted on the FCW story about the Digital Service team:

"Typical WH play: effete, elitist, and working around the political appointees and career officials in the agencies. Experts from the WH will come in and tell u what is wrong and how to fix it. Yes, the very experts who have made this administration what it is today."

Dickerson needs to understand that there is a natural resistance to the White House telling people in the agencies what to do. He needs to disarm people with collegiality, respect and un-arrogance. He needs to listen as well as preach.

I would suggest he consider convening a Frontline Forum for working level IT professionals from each agency, which would meet regularly at a fancy room in the White House complex, to share ideas and provide feedback. (I did this as administrator of the Office of Federal Procurement Policy in the 1990's, and I am still in touch with some of the forum members today!)

My advice to him would be to look for agency volunteers for his team's help on projects before imposing himself -- after all, agencies hire outsiders (consultants) all the time to help them, many should be happy to get help for free. Finally (I assume he has already thought of this), it's much better to insert himself into a project at the beginning rather than when it is underway, or off track, when the "assistance" may well be resented.

I believe we may be heading toward a tipping point for a major paradigm shift, after so many years, away from big-bang, waterfall IT development and towards a modular, agile approach. With the right mixture of enthusiasm and respectfulness, Mikey Dickerson and his team may be able to play a key role in that.