Scott Stewart and US Federal CIO, Vivek Kundra
When US President, Barack Obama, decided the White House needed a CIO to run the IT for the entire country’s administration, I am sure the recruiter didn’t use Monster.com to find the right candidate. Imagine how the advertisement would have read:
Once in a lifetime opportunity!
Our client, a country, a leading economic, political, and cultural force and a world super power, is conducting a national search for a CIO to head up its ICT function. This new role would suit a self starter who has a lot of initiative and some ideas on how to be the CIO of a country that has never had one before.
The incumbent will be responsible for a total IT budget of $76 billion and oversee 10,769 systems that support more than 1.9 million staff, servicing 300 million customers — most of whom are very unhappy because despite the 10,769 IT systems at their disposal, employees still hand out paper forms and make customers stand in long queues. The Federal CIO role will have more than 200 CIOs as direct reports (yes, that is 200 CIOs banging on your door asking for more budget, resources and time) and the successful candidate will be responsible for delivery of 794 major projects costing close to $40 billion. An immediate challenge is dealing with $27 billion dollars worth of projects that are currently running years over schedule and billions of dollars over budget.
You must have experience in changing the culture of very, very, very large organisations and be comfortable being under the unforgiving scrutiny of 300 million people who have learned not to trust their government or the people who serve them. Strong project management and people skills are highly desirable and the successful candidate should expect that long hours will be required.
The term of this gig may depend on the re-election of the President but there is a great dental plan and you get to work in the White House ... a lot.
Before you get too excited and start dusting off your CVs, you should know the job has already been filled. I had the pleasure of meeting the first Federal CIO, Vivek Kundra, during a recent trip to the US. The setting was the Harvard Business School in Boston where Vivek was addressing nearly 60 global CIOs who had gathered for a week to study and discuss several case studies around IT projects and delivery as part of the HBR course ‘Delivering Information Services’. It was a career highlight and one of the best learning experiences I have had. The line up of speakers and facilitators not only included the finest Harvard Professors but also Fortune 500 industry leaders and executives.
Kundra’s presentation was a highlight. I was impressed on several fronts, but especially with his thinking around how to attack the technology gaps that exist in the US government and his plan for greater visibility and accountability in the effort to reduce project failures and over runs.
Here are some extracts from his address.
Tell us about your first days in the role
Kundra: After everyone congratulated me, they came up and gave me a huge big stack of PDF documents and said: “Congratulations, here are $27 billion worth of IT projects that are way over budget and way behind schedule.”
So where did you start?
If I take you back to right after the election in November of 2008, I was honoured to serve on President Obama’s transition team and, for the first time in history, we had a transition team that focused on technology, innovation and government reform.
Before that, transition teams focused on healthcare and the economy, defense but never technology.
What was truly exciting was we got deep and we looked at the challenges that confronted the nation and the role of technology.
I want to share with you the sort of scale that confronted me on day one: More than $76 billion dollars in annual spending, over 10,000 systems across the US Government, over 300 million customers and of course, 1.9 million government employees. And the challenge of how we arm everybody with the power of technology and how we make sure that we close a technology gap that exists today.
How behind is the government in the technology stakes?
If you think about our day to day lives, we go online and book our favourite restaurant within minutes. You are able to book an airline reservation within minutes, to set up an accounting system for a small business in a matter of minutes. Yet, when it comes to dealing with your government you have to wait in line, hold on the phone or send in a paper form.
Looking at the difference in culture in our everyday lives we see this culture of ‘there is an App for that!’
The point has been raised about the distrust people have with the government. Why do you think this is so?
Think about the challenge we face as an institution when we try to address this, similar to some of the challenges that you face in your companies. When people interact in their day-to-day lives, a lot of it has a very low what I call the ‘coefficient of friction’ but when they interact with their government the ‘coefficient a friction’ is very, very high.
How do we start to address some of these challenges? One of the first things I did was to say: ‘Look, we have to be able to shine light in terms of how we are spending over $76 billion on an annual basis.’ So we launched the IT dashboard that gives the American people transparency to how their dollars are being spent. I put a picture of every CIO right next to the project they are responsible for and how it is performing on cost, on schedule, and whether they are hitting the performance metrics that were promised upfront when they received the funding.
What impact did it have?
One example is the Veterans Administration [where] 45 IT projects were halted, [and of those] 12 were terminated. That was about $54 million and, historically, the government doesn’t really terminate poorly performing IT projects. Historically, the government has tended to throw good money after bad money.
The next phase was to say; ‘Shining a light on all is not sufficient, and what we need to be able to do is hold the CIO and agencies accountable for how projects are performing.’
So we started the TechStat Accountability Sessions.Essentially, it’s a 60 minute session but it takes anywhere from two weeks to a month in terms of preparing for the session.
I have around the folks who are responsible for the projects around the table, from the program management perspective, the CIO, officials from the secretary’s office, all the policy councils that need to be there from the White House, all the budget folks and we evaluate, we go deep, using evidence and data to figure out how these projects are performing.
Can you tell us about some of the sessions?
To give you a very small example, we did a TechStat Accountability Session at the small business administration (SBA) and we found the SBA was spending $1614 per smartcard; the same exact smartcard you could get at another agency for $240. We decided to halt that project right there and move towards another model and the project moved 30 per cent faster.
We looked at the EPA [Environmental Protection Agency] and the TechStat Accountability Sessions found the system was $30 million over budget and it was one year behind schedule — it was to do with the financials systems within the EPA.
As we dug deeper, we found financial systems across the United States Government were way over budget and way behind schedule. But beyond financial systems, look at the history in the government: The Department of Defence spent 12 years and $1 billion deploying an integrated human resource system that failed.
The census bureau spent two years and $600 million to develop a customer device, a handheld census device to try and get the census to move to the digital world, yet our 2010 census is being developed with paper and is costing billions of dollars more. In 1998 [Veterans Administration] deployed a financial system that cost approximately $300 million dollars and was terminated in 2004 because it couldn’t execute. Then they started a system again in 2005 and we just terminated again last month. There are consistent failures across the public sector when it comes to managing large-scale programs.
So, I went back and looked at 50 years of history and what was going on. It wasn’t that we didn’t have good policy in place — we had the numerous Acts of governance, we had the E-Government Act… we had countless memos that came out of OMB [the Office of Management and Budget] around what needed to be done to manage these IT projects.
The challenge wasn’t good policy or good ideas — it was a lack of execution, the inability to execute. The market prizes execution above everything else. If you can’t execute forget it.
So what are you doing now to address some of these really difficult problems?
Number one: We ordered a halt on all future task orders and contracts across 30 financials systems in the United States Government. This is about $3 billion of spend annually and $20 billion of spend in the lifecycle costs. I just issued a memo around what we are going to do with 25 to 30 of the most troubled IT projects.
And what we’re going to do there is we’re going to go deep with TechStat Accountability Sessions and halt projects that are not performing well. We are going to terminate projects that don’t produce dividends for the American people and re-scope these projects so you don’t have deliverables that are five to 10 years out.
Part of the problem is that these projects are scoped in such a way that you have a deliverable that is two, three or four years out and, therefore, no-one is held accountable.
Some of the best practices showed that in six months, if you don’t have a customer facing a deliverable, forget it. That project is not going to be successful.
Are you working with private enterprise to learn how they deal with major programs?
In January, the President invited 50 of the top CEOs to talk about some of the challenges in the private sector in IT. One of the most common themes that emerged was that all the CEOs said if there is a project that doesn’t deliver in 18 months, they don’t even bother investing in it.
Yet there is a project that I looked at during the TechStat Accountability Sessions where they spent five years on an architecture document. And they were defending the fact that they have a perfect architecture.
Zero execution. Nothing that is customer facing. But they gave these huge thick books of BPR [Business Process Re-engineering], and guess how much that cost?
Forty million dollars to get that book.
Yet the agency is still backwards in terms of technological capabilities. Part of what we’re trying to do is fundamentally change the culture within [Washington] DC to make sure we focus on execution and ensure that, from a policy perspective, we scope projects appropriately and within a 120 day period. We are trying to unearth some of the best practices around the world — whether we are looking at other countries, local government or states, or the private sector, to figure out what we need to do to address some of these persistent structural issues.
Contractors aren’t celebrating when systems fail because they’re judged on past performance. The government doesn’t win when systems fail and, most importantly, taxpayers lose billions of dollars when the systems aren’t implemented well.
How do you sum up your role?
What we are trying to do is make sure that we are cutting waste and cracking down on waste wherever we find it. And that we are creating the right conditions and environment to make sure that the investments we are making in information technology produce the dividends for the American people.
Scott Stewart is the consulting CIO at Wilson HTM, where he architected and delivered a 'Trusted' Cloud computing strategy to deliver flexibility and agility to the organisation. He regularly blogs on cloud computing and is undertaking a professional doctorate in Information Technology and the Queensland University of Technology around business and IT convergence.