White House Forum on IT Management Reform

Uploaded by whitehouse on 09.12.2010

Mr. Zients: Hi.
For those of you who I've not met, I'm Jeff Zients.
I'm the Deputy Director for Management at OMB and the
Federal Chief Performance Officer.
I want to welcome all of you to the chilly White House today and
thank you for coming here.
It's an important day.
My job is to help make government more efficient and
effective and work as well as the American people expect and deserve.
As we all know, a productivity boom has transformed private
sector performance over the past two decades.
These productivity gains are familiar to all of us.
They've transformed entire industries.
The market leaders driving these productivity gains have
increased output, lowered prices, and boosted customer
satisfaction, all at the same time.
Unfortunately, the federal government has almost entirely
missed out on these gains.
We need to catch up and catch up fast.
We need to make government more productive, efficient, and responsive.
And IT, information technology, is the key.
Technology has been at the center of those productivity
gains I just mentioned, both efficiency gains and service
quality improvements but despite spending more than $600 billion
on information technology across the last decade, the federal
government has failed to realize the potential of information technology.
Too often federal IT projects run over budget,
behind schedule, and fail to ever deliver their promised functionality.
Fixing IT is our top priority, and we've been focused on this
since the begin of the administration.
Let me give you a quick update on our progress, and then I'm
going to hand it over to Vivek for the rest of the session.
Last year we launched the IT dashboard, where you can monitor
every dollar the government spends on large technology projects.
We're using the information on the dashboard to power what we
call Tech Stat sessions, where we get all the stakeholders in a
project, both technology and business line owners, together
in the same room to diagnose problems and agree on what's
necessary to fix a project.
This summer we targeted high priority projects for review,
projects that were either over budget, behind schedule, or at a
high risk for the kinds of problems that too often plague
our IT efforts.
We identified 26 high priority projects.
We completed the review of two of those projects in October,
and today we're pleased to announce the results of the next 16.
Here's the breakdown of the 16 projects: One project has been
terminated all together as we were clearly throwing good money after bad.
For seven projects, we've significantly accelerated
delivery of meaningful functionality, insisting on new
functionality every few quarters rather than every couple of years.
For eight projects, we've reduced budgets and scaled the
projects back to increase the likelihood of success.
So across all 16 projects, we've cut delivery time by more than
50% and reduced life cycle budgets by $1.3 billion.
Overall, we've now reviewed more than 50 major IT projects,
reducing life cycle budgets by $3 billion.
So we now know that we can improve our IT performance on a
project-by-project basis, saving money, and accelerating delivery.
But just as importantly, we're using the learning from this
work to identify the structural changes required to drive
sustainable improvements across all of government IT.
The federal CIO council has been very active in our work.
We've gotten input from folks on the Hill, and we've been working
very closely with the private sector.
The president brought 50 CEO's to the White House to advise us
on what works and what doesn't in managing IT.
And industry groups have put forward strong, thoughtful recommendations.
As a result, we now have a clear understanding of the structural
changes we need to make to achieve sustained performance
and improvement.
We all believe that IT, government IT, needs to be more
agile and responsive to evolving technologies and more
accountable and focused on delivering results.
But to get this done, we need to remove the barriers that get in
the way of consistent execution.
We are making five structural changes.
We are, first, adopting light technologies and shared services.
Second, aligning the budget and acquisition process with the
technology cycle.
Third, strengthening program management.
Fourth, streamlining governance and increasing accountability.
And last, we are increasing engagement with the IT community.
When I announced these changes just before Thanksgiving, I
noted this is one of those rare occasions in life, rarer still
here in Washington, where all the parties agree on the root
cause problems and share a common understanding of what
needs to be done.
This isn't the stuff of controversial frameworks or
radical new approaches.
The bottom line is that it's all about execution.
To that end, we're moving forward with our execution plan.
Vivek will review today the specific actions we're taking.
For each action we've identified clear owners and deadlines, and
today we're looking forward to getting your feedback and help
in refining our execution plan.
I want to thank you for all your help to date and ask that you
work with us and hold us accountable for executing on our
plan with discipline and urgency.
With that, let me hand it over to Vivek to moderate today's session.
Thank you.
Mr. Kundra: Good morning.
How's everyone doing?
Audience: Good.
Vivek Kundra: Excellent.
So, what I'd like to do today is walk you through the 25-point
implementation plan.
We just posted it on cio.gov.
You can find all the details in terms of the reforms that we're
going to be advancing.
I'd like to begin by talking about the results that have been
achieved so far when it comes to the financial reviews and also
the tech stats on the 26 high priority IT systems.
If you look at this data, as we reviewed 38 projects across the
federal government, what we've been able to do by driving
accountability by making sure that we were looking at what
works, what doesn't work, by making sure that we're making
the tough decisions that are required to turn around some of
the worst performing IT projects, we've been able to
reduce the budget of these IT projects by $3 billion.
More important than reducing the budget by $3 billion is the fact
that we've been able to accelerate delivery by 50%, over 50%.
Think about some of these IT projects that we've talked about.
They've had deliverables out there that are five, ten years
out with nothing delivered.
In a lot of cases what we've seen is seven years later what
the government got was nothing more than an architectural document.
That is why these fundamental reforms are going to change the
way we manage information technology moving forward.
There are two major areas that we're focused on where the
United States Government spends billions of dollars.
The first area is around commodity IT where we spend
approximately $24 billion every year.
The second area is around how we actually manage large scale IT programs.
These are all the programs where we're modernizing the United States Government.
This represents approximately $50 billion in IT spending, and
this is an area where we've had persistent problems that go back
decades of major IT projects that just haven't delivered for
the American people.
What I'd like to do right now is just give you the highlights of
how are we actually going to make these changes stick, what
are the fundamental differences between the previous efforts
that have been tried.
Well, I want to begin by saying that this is not about some
radical new framework or breakthrough thinking, but you
can go back to the Clinger-Cohen Act, the E-Gov Act, you can
actually look at the FAR.
All along, we've actually talked about how the U.S.
government needs to move towards a model where we're
simplifying the way we buy technology.
At the same time we need to make sure we're focused on execution.
It's not about this grand design, but it's about actually executing.
So, number one, what we're committing to is that we're
going to turn around or terminate at least one-third of
the under performing projects within the next 18 months.
Number two, we're going to shift to a "Cloud First" policy, and
what agencies are going to do in the next three months is
identify minimum of three systems that can move to the
Cloud and execute within the next one year.
Number three, we're committing to reducing the number of data
centers by at least 800 by 2015.
Number four, OMB will only approve funding for major
programs that actually have a dedicated program manager.
This sounds really basic, but you'll be shocked by the number
of major programs that don't have a dedicated program manager.
We will only approve funding if agencies are using a modular
approach with deliverables that are meaningful, deliverables
that are not five, six, seven years out, but actually breaking
them down into six-month chunks.
And finally, we're going to make sure that they're specialized IT
professionals who are working on these major programs.
Because buying complex, large scale IT programs is not the
same as buying pencils.
We're also going to work closely with Congress to make sure that
we're consolidating commodity funding when it comes to
departments and agencies under the agency CIO.
A department like HUD, which has over 200 data centers -- or
sorry -- Department of Interior has over 200 data centers.
That doesn't make sense.
And the reason is because the way the department is funded,
it's bureau by bureau, program by program, which leads to great
inefficiencies and wasteful, duplicative spending across the
federal government.
We're also going to work with Congress to make sure that we
actually give agencies and CIO's the budget flexibilities
actually necessary to make sure that we can deploy these
projects in a modular, agile fashion.
I've been working right now late at night on the 2012 budget, and
it's not fair to hold agencies accountable two years out in
terms of what they're going to spend their IT dollars on, and
then when they're not meeting those targets, you know, OMB
drags them and says, why did you fail to meet, you know, these targets?
You committed to making sure that you would hit these 29
feature sets.
They're Congressional hearings.
It's unfair.
It's a structural problem.
And, frankly, this model won't work, because the technology
cycles moves way too fast for the budget cycle.
And lastly, what we're going to do is we're going to launch an
interactive, online platform to make sure that the industry is
engaged in the pre-RFP phase so that you know what the U.S.
government is looking to buy, not after a procurement hits the
street, but before.
We want to be able to benefit from the ideas and the thinking out there.
We've seen programs that move aggressively in terms of looking
at technologies that are out there.
They end up making better decisions.
Unfortunately, there's a chilling effect across the
government where some people believe that, you know, you're
going to go to jail if you start talking to the vendors before
you put out the RFP, or after you put out the RFP,
and there are a whole host of myths that exist out there that
need to be demystified.
So this platform will be critical in making sure that
we're getting the best thinking around how we can procure
technology intelligently.
Now, why is adopting light technologies and shared services important?
The reason is, we've seen the number of data centers grow explosively.
We went from 432,000 data centers to 2,094 data centers
this year, and we're still counting as we continue to do
our inventory.
And what we're committed to is making sure within the next six
months every agency will identify a dedicated program
manager for the data center consolidation effort, and we're
going to launch a task force so that when we're optimizing our
assets, it's not just within a department, but across the
federal government.
Think about an open table for data centers.
Right now, you can make reservations to your favorite
restaurant online.
Why don't we have the same system for data center
capabilities across the federal government so if somebody at HUD
or EPA needs space, what if they could find it in another agency
that's about to spend half a billion dollars?
We're going to launch that within the next 12 to 18 months.
We've talked about shifting to Cloud First policy, but what
that really means in terms of execution is that within the
next six months we're actually going to put in place
infrastructure as a service contract that's certified
government wide from a security perspective so we can start
leveraging that.
But more importantly, a fundamental shift to Cloud, if
you think about e-mail systems, if you think about financial
systems, we want to be able to put in place government-wide
vehicles so that we can stop wasting money.
GSA and USDA have all done an amazing job by moving to the Cloud.
GSA was able to cut its cost by 50% and save $15 million as it
engages in its move to the Cloud.
The USDA is actually going to be saving approximately $6 million
a year with its shift to the Cloud.
But more importantly, it's freeing up resources from the
CIO's to do work that's higher yield and adds value to every
day Americans.
We're also going to be rethinking our entire shared
services strategy.
The last time this was visited was almost a decade ago, and the
technology landscape was very, very different.
Today, we may be able to deploy better services at lower costs
in a federated web services API driven world that's going to
allow us to deploy some of these technologies much faster at a
fraction of the cost that we've deployed it in.
I want to share with you this story at the Department of
Interior and the great work that's happened with the
leadership from Bernie and Andrew.
Now, Department of Interior, here's an example of why lights
technologies are important.
This is one of the projects that were reviewed as part of the
high priority projects.
The Department of Interior was able to, by applying light
technologies and shared services, reduce the life cycle
costs of the infrastructure investments by $500 million and
is committing to go from 210 data centers down to 115 data
centers by 2015.
What you're seeing here is this $36 million savings over the
next five years as a result of moving to the Cloud.
But more importantly, what you're seeing is this massive
reduction, 55% reduction, in data centers, and that's not
just in managing those assets.
What that does is it actually focuses the CIO's in the agency (inaudible).
So I want to recognize the great leadership that Interior is
providing when it comes to adopting light way technologies.
Another area for us that's really, really important is to
make sure that we're strengthening program management.
Unfortunately, program management is not even a career
track within the federal government.
Yet, program managers are the ones who are spending hundreds
of millions and billions of dollars every year on
information technology.
And what's happened is we've unfairly plucked people from
various jobs and said, today you're going to be responsible
for a $100 million project without arming them with the
appropriate training, without making sure that there's a
community where they can share best practices, without making
sure that their role is lifted up so they're not buried ten
layers deep within the bureaucracy.
That is why working with OPM in the next six months we're going
to be designing a formal IT program management career path
with direct higher authority for agencies.
As we were reviewing some of these projects to the tech stat
accountability sessions, in one example, at the Department of
Homeland Security, the CIO was so frustrated, it took him five
months to bring on board a program manager for a critical project.
The direct hire authority is vital.
And then we want to be able to scale that program management
career path across the federal government, but first we're
going to pilot it with a number of agencies.
We also want to be able to make sure that we have integrated
project teams as a precursor of launching any major project with
the leadership from the acquisition community, the CFO,
the business side, and the technology world, the
architects, the engineers who are looking at these solutions.
Unfortunately, the world we live in today, what ends up happening
is, people are hiding behind their e-mails going back and forth.
And the incentives are not aligned.
The procurement community has its incentives.
The program management community has its incentives.
The architects have their own incentives.
By integrating these teams, the shared incentives to make sure
that we're advancing the delivery of this project in a
meaningful way at low cost is going to be much, much easier
when you actually have these team that are collocated and are
precursors to major programs.
The federal CIO council is also going to launch a best practices portal.
And initially, this portal is going to be focused very much
around program management.
If we look at what's happening across the federal government,
the colossal failures that we've seen across the federal
government, but more importantly the great projects within the
federal government where things have worked really, really well,
there isn't a central repository of that knowledge.
That information is not shared across the government.
It's within departments or bureaus.
And frankly, what we need to be able to do is share those best practices.
Everybody can learn from successes and failure within
the federal IT community.
We're also going to be launching a technology fellows program
modeled after the Presidential Management Fellows program.
This is really exciting because I get countless people who
approach me when I'm traveling all over country on how they can
give back to their country.
And we don't have a formal mechanism to engage them.
We don't have a formal mechanism to bring in fellows who can come
in for a one year, six month stints and help us change the
way we manage IT or introduce us to some of the new technologies
that are out there.
But more importantly, the ability to actually build a
pipeline of talent for the next generation of innovations that
we need to drive across the federal government.
We also need to be able to go back to the e-Gov Act where
an authority was built around making sure that the industry
and the government could actually begin rotations, where
you could have somebody from the industry come into the
government and work on a set of projects and somebody from the
government could go to the industry.
The reason it didn't work was because I think there were a lot
of conflicts of interest.
And I think the challenge is going to be, how do you do that
without conflict of interest.
Well, you don't do it with technology companies.
They're plenty of enterprises out there that have huge
operating units, you know, from Starbucks, which is a global
enterprise, to looking at companies like Alcoa [phonetic]
and figuring out, well, how do we get some of the folks who are
working on large scale IT projects to rotate between the
federal government and the private sector?
Now, here is an example of what happens when you don't have good
program management.
Richard Spires [phonetic] and I reviewed this program as one of
the high priority projects.
This is a project where the government spent $40 million in
seven years.
And it got a system that didn't work, code that was buggy.
And unfortunately, when we looked at it, we dug deeper, we
realized well, they really didn't have a dedicated program
manager, there were no performance metrics, they
weren't providing the appropriate oversight of contractors.
That is why we made the decision to terminate this project.
And that is going to be the future of projects that don't deliver.
We want to make sure that we do not continue to throw good money
after bad money.
And the old path here would have been to continue to spend this
money until this project became a national headline.
And then it would have been cancelled.
But we want to be proactive.
We want to crack down on wasteful spending across the
federal government.
Now, one of the biggest challenges that confronts the IT
community is the fact that we haven't really aligned our
acquisition processes with technology cycle or our budget
processes with the technology cycle.
Dan Gordon and I are going to be working very closely on making
sure, when it comes to aligning the acquisition process with
technology cycle, we actually have meaningful reform.
One of the things we're going to be doing within the next six
months is designing and developing a cadre of
specialized IT acquisition professionals.
We also want to be able to make sure that we're identifying some
of the best practices that already exist, similar to the
program management.
Within the acquisition community, the challenge has
been that for years we've known that we should move towards a
modular agile methodology.
As a matter of fact, Dan and I looked at the FAR language, and
it's clearly spelled out there.
The question is, well, why isn't it happening?
The reason is, because historically, we haven't been
very execution-focused because agencies want to move towards
agile modular, but they don't know how to do it or we haven't
provided the right tools.
How do we put in place the appropriate templates that
support modular development?
How do we make sure that we have model language that it can plug
into their contracts?
How do we make sure that these integrated teams are coming
together as they're deciding the future of some of these IT projects?
That is critical in terms of what we're going to be doing next.
Well, one of the areas that I'm very, very excited about is item
number 16 here, reducing the barriers to entry for small
innovative technology firms.
And the reason that's really important is that the
government, unfortunately, ends up spending millions of dollars
sometimes when they could find an innovative solution that
could cost pennies on a dollar.
And the reason is because a lot of the start-ups that I talk to,
their view is, you know, we don't really want to deal with
the federal government because your procurement processes are
too complicated, we don't even know how it works, by the time,
you know, we get on the schedules, it's too late.
I was talking to the founder of Hotmail who told me that look,
you know, from the moment he started Hotmail, to the moment
he sold it for over $400 million and millions of users, the