Vivek Kundra: Engineering a Radical Transformation

Vivek Kundra, chief technology officer of the District of Columbia, spoke with Digital Communities editor Blake Harris about his management approach, IT governance innovations, cost-cutting measures and the pace of innovation. Today, President Barack Obama named Kundra the "federal CIO."

by / March 5, 2009
Vivek Kundra A Radical Transformation Photo by John Harrington

Vivek Kundra, chief technology officer of the District of Columbia, drew national attention for his innovative style and his drive to use technology to greatly increase government transparency and citizen involvement. A few weeks ago, he spoke with Digital Communities editor Blake Harris about his management approach, IT governance innovations, cost-cutting measures and the pace of innovation. On March 5, President Barack Obama named Kundra the federal CIO -- a new position that will oversee technology investments and technology spending by the federal government.

Q: What are some of the new approaches you've introduced in Washington, D.C.?
I usually point to three areas in terms of what we are doing here. It starts with a shift in philosophical thinking, but there are three pillars. First, there is the question of how you lower the cost of government operations by moving toward consumer technologies rather than technologies focused within the enterprise. What I mean by that is, for too long the public sector has thought for some reason that it's so special that it must have custom applications, it must have radios for the public safety unit that cost $6,000 a piece, it must have applications targeted for the enterprise. And the government chooses to pay for things that it can get for free.

Second, there is the question of how you drive radical transparency in the public sector. We are doing that by democratizing data and opening up the warehouse of government information to the public so people can innovate; they can slice, dice and cube that data to create amazing and creative applications for citizens and government.

And the third big pillar has to do with how you fundamentally rethink IT governance. What I'm doing with IT governance essentially is simplifying it, getting away from complicated methodologies, and getting more focused on process rather than outcome. Here we've moved to a simple "stock market" model where every technology project is treated as a publicly traded stock.

Q: Your last point brings up how you've approached running IT projects -- which is quite different.
When I came to the district, it had spent more than$1 billion in IT. Yet it was very difficult to say which IT projects resulted in benefits and what those quantified benefits were. Or what the project failures were and what caused those failures. There wasn't any real way to do this type of analysis. And frankly, we did not get a billion dollars worth of technology. These followed old approaches in government and people would not get fired if they were incompetent or they didn't perform. We were also relying too much on consultants to drive the government agenda. So what I said was, if on my iPhone as I'm walking down the street I can check the price of any publicly traded stock, and at the same time, get all the news coverage around that stock, why can't I do that with public-sector spending?

I decided to get rid of my traditional project-management team and I hired five people. The reason I hired five was that I had five major clusters of government operations: public safety, education, health care, economic development and government operations. I began treating each IT project like a stock. I literally built this stock market floor, and with each project we look at the "happiness index." This indicates how happy the customers are how happy the people who implement the project are, and how happy the stakeholders are. We micro-poll to get the happiness index. We see how they are doing in terms of schedule [and] then of course, evaluate the management team. When that's put together as "stock," we categorize them into buy, sell or hold.

For stocks that fall under the sell category, we actually bring in the project manager and we conduct an Oxford Union-style debate, where the project manager of that initiative has to argue in the affirmative for keeping the project going and somebody on my team argues in the negative, to close it down. That way, we can see what's going on with that project. Based on the outcome, we decide to buy, sell or hold those projects. Recently we just "sold" one and saved $3 million. For another that was in a lot of trouble, we put in an emergency team to take over.

To give you an example, in contrast to this model, of what was happening before: Three years ago, D.C. public schools spent $25 million deploying PeopleSoft applications. Now this project failed. Not a single person was fired, and not a single company was sued. They just flushed $25 million of hard-earned taxpayer dollars down the toilet. And what this stock-market model allows us to do is rethink IT governance.

Q: Radically increased transparency really ties into what you mean by a shift in philosophy. Can you elaborate on that?
It's a very personal thing for me because I grew up in Africa, in Tanzania to be specific. I grew up in an environment where there were scarce resources and you had to get by with very little in life. Those early years really shaped my thinking about surviving on very little. But more important than that, it left a lasting impression of the difference between good and bad government, and what bad governance looks like.

When you have corruption, when you have a system that's broken, when you have people in the public sector who care more about themselves rather than "we the people," you end up with a failed state or failed organization. I brought that with me when I moved to the United States at the age of 11. And our bedrock principles of democracy -- freedom of the press, open government, sunlight -- all those resonate with me. You realize that government and democracy die behind closed doors. It's very important that government is practiced in the public square, so to speak, and not in the halls of power behind closed doors amongst just a few people. By democratizing data and information and public processes, making them open and available, you are able to engage citizens in a way you never could before. I think the information revolution has enabled a lot of that.

Q: What are some examples of the kind of public data that you have now made available?
A few examples: One is on every consultant I hire in the government, I've actually put out as a data feed who in the government hired that consultant, how much we are paying that consultant and an hourly rate, which company he or she works for, and the company's address. That's to democratize information about procurement in the public sector. A second example: Every 311 call that comes into the District of Columbia -- a service request -- we actually publish that. So whether that has to do with a broken parking meter, potholes or crime -- all that is published and put into the public domain. A third example is a lot of the crime data, whether it's assaults, burglaries or homicides. All that data is put out there, available to the public in real time. A fourth one: When we were deploying computers to the public school system, as the computers were being deployed, there was a live data feed that showed where they were deployed, which school and a [phone] number for the principal.

Q: Your approach is to treat citizens as partners in government, rather than simply as customers to

be served.
Exactly. Back in the days of kings, people would kiss the ring, and the king would decide all the rules and regulations and how to promulgate them. But the purpose of democracy is that you move away from being treated as subjects and become active citizens. That's what participatory democracy is all about. It's not about being a subject to a centralized, autocratic government, but it's about "we the people" being able to hold our government accountable. And also ensuring that we are creating a better government. My view is that citizens are co-creators of government, not just subjects of it.

Q: One example of turning citizens into real partners is your Apps for Democracy project. Can you talk a little about this?
That followed after I published all these data feeds related to government. I thought how do you really embark on a technology revolution in the public sector? And the best way to do it, in my view, is to actually engage the citizens. We basically said we're going to put out $50,000 -- a modest sum -- for a competition, out of which $30,000 went to run the competition and only $20,000 went to prizes. The highest prize was $2,000. I put a limit of 30 days on the competition, and the challenge was for citizens to create applications for citizens themselves and the public sector. To my delight, we ended up getting 47 fully functional applications in 30 days for $50,000. By our calculations, that would have cost us about $2.6 million and would have taken us two years.

Q: What kind of applications did contestants create?
There is one called This application allows you to put in your home address and based on that address, it lets you know crimes that happened in your neighborhood, the closest bank and things of interest near you. It lets you find the buses and trains and when they are coming. So it really takes all those data feeds and gives them to you in the context of your address.

A second application is for people who love bikes. It lets you know where the best bike routes are, where bikes are getting stolen, where you can buy cheap bikes, and it lets you create a community of bicyclists if you want to go on long rides.

A third application is called Point About, created specifically for the PDA. We are using it as one of the official inauguration apps because, using your iPhone and your BlackBerry, based on your GPS coordinates, it lets you know the closest Metro station and when the next train is coming or where the closest restaurants to you are. It takes those data feeds we have and geocodes them. People can look all these up on the Web at That URL, by the way, is part of my vision of creating a digital public square.

Q: Did you face any obstacles moving away from the custom or enterprise software and tackling the notion that government is different or that the security has to be so much better?
Yes, but I think it's the 80/20 rule. I would agree with the critics who would say classified information can't be moved to the cloud or if you have very sensitive information, you need to deal with that differently. I totally agree with that.

Q: So public safety applications, for example, are a separate case with very legitimate concerns. But those concerns don't then have to extend right across all government.
Right, and that's my argument: The sensitive or classified data represents 20 percent of the public sector. But there are cost-saving possibilities for the 80 percent, where most of the information isn't highly sensitive. It's

training videos, collaboration or rules and procedures -- all of those things. They aren't classified. They might be a little sensitive, but they are not classified. So why would we pay millions of dollars more to secure them when, at the end of the day, you have a bunch of consultants managing that infrastructure anyway? What often happens is that 20 percent of the government ends up driving 100 percent of the decisions, rather than being pragmatic about it. There are certain things where it doesn't make sense to buy your own applications. Why buy blogging software when you can get free blogs? Or you aren't going to go out there and re-create Twitter, Facebook, Picasa or Flickr when you can get it all done for free.

Q: Was this the argument that you used to sell this approach?
Yes, but more important, we started with a pilot. I started with my agency of 700 people. With those 700 people, we moved forward, tested it, made sure all the security requirements we had were well documented, and then scaled it from there. And all along, Mayor Adrian Fenty was extremely supportive.

Q: What advice would you give to a CIO who was trying to adopt a similar approach?
One of the biggest things that has been very helpful for me was the creation of test labs. I've created labs within my organization that are the test bed for innovation. We try out new ideas, we look at models that can be scaled quickly, and we throw hundreds of ideas against the wall and try to fail them fast. And the ones that resonate, we scale them very fast also. It keeps the pace of innovation moving forward. Then philosophically, the question is why government has to be a laggard when it comes to innovation and technology. It used to be that government was the leader -- government was the trendsetter. And we've lost that. Now everybody's taken these hyperconservative positions and that hurts us now when it comes to innovation, efficiency and delivering services. From a philosophical approach, I would say, "Be bold. Be aggressive when it comes to innovation. Don't just turn it into a word that you talk about -- execute."

The lab I have consists of three people who I work closely with, and we basically experiment. We try to think through where the technology going, what it is going to look like. And we try to look at where the county is headed, where the world is headed, what's happening in the developing world, what's going on in the public-sector space. We try out new ideas, whether that's in the social-media space, or whether that's artificial intelligence, whether that's in robotics. We look at all those technologies to see how they could transform the public sector.

The promise of technology was not that it would cost millions of dollars and always be behind schedule. The promise was that it was going to fundamentally change the way we worked. So how do you think this through? What did the light bulb do when it became part of our cultural fabric? It transformed the way we worked, played and lived. In the same way, what can these new technologies do, what transformation will they bring? What did the locomotive and steam engine do in terms of transforming our economy? That's what we need to focus on -- going back to the basics and determining how you increase value at lower costs, rather than asking for these big, multimillion dollar enterprise projects. And finally, I would say that there's no better time to innovate than during tough economic times. This is a great opportunity.

Blake Harris Editor
Platforms & Programs