The FISMA Focus IPD

And it has spurred the adoption of new ways of purchasing and delivering IT, from app stores and shared services to cloud computing. It has even promoted the use of 3D telepresence as a substitute for international travel and live appearances at conferences (see photograph, above).

In shaking up the historical picture of government IT ― and making a strong case for doing so ― Kundra has become a source of inspiration for government CIOs around the world. Dozens are pursuing internal app store programs and opening up their country’s data sources. And almost all, including the European Union, are launching strategies for cloud computing that draw on Kundra’s pioneering vision.

That vision is born out of a set of core frustrations with public sector IT: systemic inefficiency, inflexibility, a lack of demonstrable ROI and a sense that both consumer and private sector IT have overtaken government IT in many key areas.

“Despite spending more than $600 billion on information technology over the past decade, the federal government has achieved little of the productivity improvements that [the rest of] industry has realized from IT,” Kundra wrote on the White House blog in December. At the same time, “too often we hear stories about how the [US] government, for one reason or another, lacks technological capabilities that are commonplace in the private sector and our everyday lives.”

Many of the changes he is now enacting are long overdue, he argues. “We need to make sure that we are not relying on the same old ideas, and the same old approaches in terms of how we invest in IT.”

And for Kundra, the cloud computing model is the springboard for much of that new thinking. In that arena, he looks set to match ― if not leapfrog ― the private sector with the bold adoption he is championing, driven by a conviction that cloud has the potential to deliver a new era of cost-effective, responsive, innovative, efficient IT to government.

Under his leadership over the past two years, the US government has been building a comprehensive strategy designed to cover every core aspect of cloud computing, from security and interoperability standards to the inevitable consolidation and closure of data centers. Now, that strategy is ready for large-scale ― and mandated ― adoption.

Cloud first

We are “shifting the mindset of the US government from asset ownership to service provisioning,” Kundra said at the Cloud Security Alliance Summit in San Francisco in February. “We are looking to drive a fundamental shift in capital allocation.”

That shift will manifest itself in some pretty big numbers. Kundra says that $20 billion of the total US Federal Government IT budget of $80 billion has already been identified for the move to cloud. And to ensure that this 25% cloud migration commitment is not stillborn, Kundra and 27 main departmental CIOs have signed up to a “Cloud-First” policy that all must implement by the middle of 2011.

“Every agency must first evaluate a viable cloud solution before 
they go out and make investments or procure infrastructure,” Kundra outlined in February. “We are setting that as part of the budget process [so] cloud is at the heart of how we provision IT.”

The policy involves an immediate call to action and the setting of some fairly ambitious targets. By mid-year, each government CIO is obliged to “identify at least three ‘must move’ services that they’re going to move to the cloud, and create a project plan for migrating one of those services to the cloud solutions within 12 months and the remaining two within 18 months, retiring the associated legacy systems,” Kundra outlined.

“And these are not just simple systems; these are going to be core to the workflow of these agencies’ systems, and disruptive in terms of budget and savings.”

Ambitious strategy

The ambition runs even further. By the end of this year, all CIOs must have created a strategy for shared services.

As the strategy has been taking shape, there have been plenty of tactical initiatives that have bolstered Kundra’s confidence in cloud’s ability to deliver both savings and improvements in services:

• The General Service Administration, the operational support services unit for all Federal agencies, is moving its 17,000 users to the Google Apps platform, replacing a fragmented infrastructure consisting of multiple email systems spread across 17 global locations that have struggled to interoperate. The shift to cloud email services will result in a 50% cost reduction over five years ― a saving of about $15 million.

• The Recovery, Accountability & Transparency Board’s website was moved from an internally hosted platform to the Amazon EC2 cloud, saving $750,000.

• To coordinate the allocation of grant funding to doctors and hospitals for the implementation of the US government’s new Electronic Health Records systems, the Department of Health & Human Services used a Salesforce.com-based CRM and project management solution. That reduced the time to go live from one year to three months.

• The US Department of Agriculture is in the process of migrating 120,000 users to the Microsoft Azure cloud service platform, a move that will consolidate 21 siloed systems at a saving of $27 million.

What comes next in the Cloud-First program is the identification and prioritization of cloud-ready systems. Aside from collaboration and office productivity tools, areas earmarked include: workflow-based processes such as employee verification, grants management, claims processing and CRM; business intelligence; and select areas of information security such as identity management.

Shared data centers

But arguably the biggest shake-up is coming in IT infrastructure. As Kundra emphasizes: “Agencies don’t want to be in the infrastructure business, they want to abstract their entire infrastructure [and] consume it very much like [they consume] electricity ― as a utility. They want to move away from managing data centers to actually working on what really matters ― serving their customers.”

Data centers are, in fact, a particular target. In the last decade, the number of facilities operated by US federal agencies grew from 432 to 2,094, even as CIOs in other industries were starting to consolidate their estates.

For Kundra, that translates into large-scale duplication in terms of spending, a fragmented infrastructure that has serious security implications, a lack of agility when the need arises to ramp up and scale back IT, and a decoupling of private sector innovation from government IT because government IT is too often custom built.

“In 2010, approximately 30 cents of every dollar invested in Federal IT was spent on data center infrastructure,” said Kundra. “Unfortunately, only a fraction of this investment delivers real, measurable impact for US citizens. By using the cloud computing model for IT services, we will be able to reduce our data center infrastructure expenditure by 30%.”

The aim is to bring an to end the siloing of data centers to ensure spare capacity can be shared across different departments ― in essence, to create an internal market for data center capacity.

“Within the next 18 months, [we] will create a government-wide marketplace for data centers that will match agencies with extra capacity to agencies with increasing demand, thereby improving the utilization of existing facilities,” he says. Utilization rates currently run at 10% to 27%. Ultimately the plan is for almost 800 (or 40%) of the federal government’s current 2,094 data centers to be eliminated by 2015.

Increasing visibility

The broader cloud strategy has been in gestation since Kundra’s early months as CIO. But a prerequisite for that was an early move to dramatically increase visibility into how the US government spends its $80 billion IT budget.

“I remember my first day on the job, when I was handed a static PDF doc that said ‘here are $27 billion [worth of projects] on the management watch-list, that are way over budget and years behind schedule.’ There was one IT project, where a billion dollars had already been spent, that was 12 years behind schedule.”

Reacting to such runaway projects, he launched a public dashboard “to shine a light on the performance of this portfolio.”

“The way this portfolio was being managed was in a very secretive, very opaque, very closed way,” said Kundra. “[With the IT Dashboard] we moved away from a culture of faceless responsibility, putting the picture of the [relevant agency] CIO next to the project they were responsible for, making sure everyone knew it was red, yellow or green.”

The analytics portal tracks the progress of investments over time, displaying data on over 7,000 investments and detailed information on the status of over 800 “major” projects. Investments are ranked out of 10, and flagged as “normal,” “needs attention” or “significant concerns.”

But when introduced in June 2009, it was more than just a show-and-tell exercise. Bad dashboard results were followed by TechStat Accountability Sessions ― face-to-face, evidence-based reviews that end with a decision to turn around, halt or terminate an IT investment.

Data-driven innovation

Another of Kundra’s early initiatives that kicked off in mid-2009 was Data.gov, a facility providing third-party access to the masses of raw data generated by government ― a move that has subsequently been copied in Canada, the UK and elsewhere.

It was launched with 47 machine readable data sets; it now provides over 270,000 data sets on every aspect of US Government operations from healthcare and education to energy and geology. As well as creating unprecedented levels of openness and fostering greater public participation in government, it has also encouraged innovative use of that data by the private sector.

“Part of what we have seen is the ability to unleash creativity and the ingenious ways that were structurally impossible before,” he told an audience in Brussels recently.

For example, when one company turned government agency data on consumer product recalls into an iPhone app, the agency “co-innovated” with the developer to extend the app so that citizens could take a picture of a product’s bar code and see instantly if that product was the subject of a recall.

That is just part of a wider revolution. By the end of next year, as initiatives such as Cloud-First, data center markets, software-as-a-service, and other services move from pilot to adoption phases, the transformation will be much more pronounced.

This is a one-way street, says Kundra. “When you can save $20 million on something as simple as email, imagine what happens when we go after that $80 billion and the opportunity to crack down on wasteful spending [and], most importantly, unleash the innovative spirit. Cloud computing will not just be more innovative than we imagine; it will be more innovative than we can imagine.”