Interview with Joshua Marcuse, Director of Strategic Initiatives, Google Public Sector

Recently, Marina Theodotou, who leads the DAU Innovate to Win initiative, and other DAU faculty members met with Joshua Marcuse, the Director of Strategic Initiatives at Google Public Sector and former Executive Director of the Defense Innovation Board from 2016 to 2020. He was asked about the state of play of scaling innovation at the DoD. The following are key excerpts from that conversation. 

Q. Can you describe your current and previous roles as backdrop to our discussion about the challenges facing the DoD in scaling innovation and engaging nontraditionals?

A. My last role in the DoD was running an advisory group called the Defense Innovation Board, which was such an honor. In that capacity, I worked really closely with organizations across OSD [Office of the Secretary of Defense], including A&S [Acquisition and Sustainment] and R&E [Research and Engineering], and especially with DIU [Defense Innovation Unit]. I had the chance to be involved in standing up a few organizations like DIU and JAIC [Joint Artificial Intelligence Center] and working with the Services, and a few program offices, too. So, I got a bunch of different lenses on acquisition, which was really a privilege. From that exposure, you could say I developed some strong opinions about why we were experiencing some of the modernization challenges that have been so well documented and so often discussed.

Then, two-and-half years ago, at the end of a decade-plus of government service, I came to Google to help stand up our public sector division. Now I see the very same issues from the vantage point of a company that’s very eager to work with the DoD, but also experiences some of that transactional friction in working with the government. Google’s origins may be from a Silicon Valley garage, but today we’re a trillion-dollar market cap company. Ironically, even with the resources Google can bring, we still experience many of the same challenges and pressures as today’s smaller venture-backed companies. I have empathy for all sides. And I have always believed—deeply believed—that DAU has a central role to play in driving forward the [acquisition reform] agenda for the Secretary, the Deputy Secretary, A&S, R&E, you name it. So, thanks for having me.

Q. Thank you for that powerful backdrop that tees up our discussion perfectly. So, from your vantage point, when it comes to innovation, what are the key challenges facing the DoD, and how can we tackle them?

A. Why don’t I start with that core thesis that I think you should be taking into consideration: Dr. Eric Schmidt—the first chair of the Defense Innovation Board and sort of my putative boss there—in his 2016 House Armed Services Committee testimony, said, “DoD doesn’t have an innovation problem, it has an innovation adoption problem.” And that framing became the bumper sticker for the next four years of our work.

Framing is so important because the tools that you would use to solve a problem that is fundamentally organizational, psychological, and behavioral are very different than the tools you would use to solve a problem that is basically an engineering challenge. I often see the words technology and innovation used interchangeably, and this stems from a flawed understanding of the problem. 

“DoD doesn’t have an innovation problem, it has an innovation adoption problem.”

Q. So, what does that mean for an institution of higher learning?

A. It means you’re the epicenter! You have an outsized influence on the behaviors that you are looking to inculcate and incentivize, and DoD would do well to invest in DAU and empower you. I know that was a priority for [former Under Secretary (A&S)] Ellen Lord, and I’m sure that’s a priority for current Under Secretary Dr. Bill LaPlante, as well. If you frame this as a challenge of “how do I get the organizational behaviors to evolve?” you’re led to different outcomes than if you say, “We have to change the FAR [Federal Acquisition Regulation]” or “we have to use more OTAs [Other Transaction Authorities].”

Q. Joshua, are you saying that we need to reframe the challenge from a technical lens to an organizational change centering on people and behaviors? Is that correct?

A. Yes. This is a reformulation of another trope, which is “DoD has all the authorities that it needs, but it doesn’t use them.” I think many members of the House and Senate Armed Services Committee would agree with that statement. We have to apply a bit of social psychology and behavioral economics to understanding why the DoD doesn’t use these authorities more. And that, I think, is where we should focus quite a bit of intervention. What if we shifted from “Those people are part of the frozen middle and should retire” to “What are the incentives to influence behavioral change; what are the levers, and who has these levers?” It requires an enormous amount of courage, grit, fortitude, creativity, and determination to do new things in a large system that is optimized for consistency and sometimes punishes deviation. So, not innovating can be a rational response, right?

Q. Can you expand on that?

A. Yes. There are two things that I’ve observed that are really hindering adoption.
As we know, since the 2018 NDS [National Defense Strategy], it has been the policy of the DoD to be faster and more agile with acquisition. I think we have made enormous gains in acquisition reform. I love Middle Tier acquisition. Use of non-FAR-based contracting, such as OTAs, or CSOs [Commercial Services Openings], etc., has skyrocketed. We’ve totally reimagined SBIR [Small Business Innovation Research] at AFWERX [the Air Force technology directorate]. I think for more than 35 years—35 years—since the Packard Commission we believed that, if we could fix acquisition policy, we would get the behavior that we want to see.

And yet, we’ve seen fewer results than we should. Why? Tragically, I think we solved the wrong problem.

While we have made a lot of progress on acquisition, we have made no progress on budget allocation. The formation of the congressional commission last year was really the first acknowledgement that PPBE [Planning Programming, Budgeting and Execution] is a problem—maybe the root of the problem. Budget reform is the third rail, but we’ve got to confront it.

The PPBE process is designed to stifle innovation. I call it the anti-innovation, the innovation killer. It is designed to stifle innovation because it optimizes for consensus, control, and predictability. It was a Cold War-era mechanism to maintain stability—a process that evolved for resource allocation that is so complicated and lengthy it’s resistant to shifting political winds or changing leaders and, as a byproduct, resistant to new ideas and new technologies. We designed the PPBE process to be stable; yet, stable is often the opposite of innovative, creative, or disruptive.

It requires an enormous amount of courage, grit, fortitude, creativity, and determination to do new things in a large system that is optimized for consistency and sometimes punishes deviation.

This system is not meant for ingenuity, agility, or velocity. It’s optimized for control. It reduces risks, but perhaps the wrong ones? Today’s PPBE system generates enormous risks: We’re accumulating technical debt over time and widening the gulf between what’s in the program, what the Warfighter needs, and what’s possible with commercial technology. Process takes so long relative to the pace of technological progress that requirements are obsolete by the time they reach industry.

Q. That is a powerful first point! As we know, the PPBE Commission is now underway to tackle some of the key issues you highlight here. I’m eager to hear your
second point.

A. This second area is funding. All the innovation organizations are basically broke.

We are in a situation now where the PEOs [Program Executive Offices] have nearly all the money, and the spending is all planned out to the last penny. They are building the platforms that the institution has approved over the years. Let’s call these Systems 1 organizations—Execution. But the innovation organizations—the small cells within the DoD— have comparatively fewer resources. These are Systems 2 organizations—Exploration.

Systems 1 is the main continent, a vast and dense concentration of land mass. By comparison, Systems 2 are poor, isolated island nations stretched out in a remote archipelago. So, this is a reformulation of the “Valley of Death” problem: how to move successes from Systems 2 culture to Systems 1 culture for scaling.

This situation fits Clayton Christensen’s “Innovators Dilemma” perfectly. It’s not unique to government. It fits every other corporate innovation model. Uncanny. It’s nearly impossible simultaneously to solve for Systems 1—quality, reliability, efficient execution, transparency—on the one hand and Systems 2 on the other—rapid experimentation, appropriate risk taking, and user-focused product mentality. These cultures battle each other, and one is a thousand times larger, so it wins.

What percent of the budget goes to Execution and what percent goes to Exploration? It’s got to be about 100 to 1 or 150 to 1? All the spending is happening in the Systems 1 rather than the Systems 2 organizations, right? We can’t get resources into the Systems 2 organizations because all the funding comes from the PPBE process, which pumps money into Systems 1. I think the solution is to reduce the impedance mismatch from Systems 2 to 1, and be intentional about where you want to optimize for stability and where you want to optimize for agility. To me, that looks like creating tech injection flywheels inside existing programs of record to pull new capabilities into existing systems. The PEOs need to be the center of gravity in our organizational change campaign as they’re the nexus.

When I talk to the founders and investors who are building a new generation of defense companies, they have been vocal advocates of wanting to win not only contracts but programs of record, specifically, as they see that as a litmus test. They say, “DoD isn’t really serious about innovation. Until there are programs of record awarded to companies that look like mine, you aren’t proving to the market that you’re serious.” I don’t agree at all. I know programs of record are the coin of the realm, because they are large and politically and bureaucratically stable, but I would argue that is backwards. Over time, the traits that make you successful at pursuing and winning programs of record conflict with the traits that made your company innovative at the start. Over time, that conflict sucks the innovation right out of your company culture.

Q. And why is that? 

A. Programs of record are the past. DoD should not want more of those. They are attractive [to companies] inasmuch as they are familiar and lucrative, but they are not an optimal construct for building the future. They are subjected to all of the barnacles and friction and added performative inspection and all the rigidity of all of our largest monolithic “too big to fail” programs. Instead, we want continuous and constant microinjections of technology into existing platforms.

Ideally, we would go from hundreds of programs down to dozens of capability portfolios, inside of which meaningful agility would be possible. The future of warfare is software-defined, so these systems will need nonstop computer upgrades, software upgrades, cyber patches, and algorithm refreshes.

We don’t want to try to move business from five big [legacy defense] companies to five newer companies we call nontraditionals. Over a few decades, they too will become optimized for placating requirements, complying with FAR-based cost and pricing rules. What we really need to do is figure out how to get those complementary companies to work together. How do we get a steady stream of capability moving from Systems 2 to Systems 1 and avoid altogether the sclerotic and glacial process?

Q. And how can we do that? How can we get the primes, the established industrial base companies, to work with the startups, the nontraditionals?

A. Yes, that’s the last point I’ll give you.

When I describe this using a white board, I draw a big triangle. At the top of the triangle is the Pentagon and the base of the triangle on one side are the traditional companies and on the other side are the nontraditional companies. You call it the Defense Industrial Base and the national security innovation base, if you will. We have spent 75 years on the leg of that triangle that connects the DoD to the Defense Industrial Base. We have an industrial policy office for that, as well as congressional committees, trade associations, and lobbying groups. We have created the military industrial complex for that purpose, and it’s an elegantly integrated system. These companies are not tech product companies, though. They operate more like intensely regulated state-owned utility companies. They are so essential to the country precisely because they are adapted to a particular set of strengths that are needed specifically to respond to DoD requirements.

The other leg of the triangle connects the Pentagon to new companies. Since about 2016, with former Secretary Ash Carter and [Deputy Secretary] Bob Work and the Third Offset, we’ve started to get interested in these, but the ecosystem is nascent. DAU doesn’t have much developed curriculum for that. We have arguably one association for that—the Silicon Valley Defense Group. It’s just a newer space.

But the two corners at the base of the pyramid are just operating in separate worlds. I hear a lot of pundits clamoring for DoD and Silicon Valley to fix their relationship. Sure, but that’s just half of it. That misses the greatest opportunity for America: collaboration at the base of the pyramid. That is the most important thing for us to study and reshape. Designing the acquisition process to drive relationships between the traditional and nontraditional companies is where opportunity lies.

Q. So how do you recommend we enhance that relationship? 

A. That is where I would love for you, DAU, to spend some time thinking about adapting the tools that the DoD has at its disposal for small business, like matching programs for big and small companies, and applying them to old and new companies. What if, in addition to a small business set-aside, we created a new business set-aside?

For example, we need to define and pull the levers to incentivize big hardware-centric companies with profound defense domain knowledge to work with newer software-centric companies that have a real, user-centered, agile product mentality. This is how to get the best of hardware and software blends that we need in the DoD.

Such partnerships are really promising. They facilitate a massive complementarity of cultures, organizations, systems, and workforces that have become optimized for different incentives over time and can be unified, if the DoD is clever about it. But more to the point, it’s the only pathway for the DoD to get back on the right side of the commercial cost curve and start using the massive scale of the consumer and enterprise market to its advantage. Without this, Big Defense continues to be isolated from where digital and other advances are happening fastest.

Q. How do you see the PEOs’ role shifting with this new paradigm?

A. The new PEOs’ role is coordination and deconfliction between the two marketplaces, or I might even say creating dynamism and harmony there. Their role today is centered on cost, schedule, and performance, and mainly risk reduction—or risk displacement—rather than capability maximization and speed. So, all of this is wrapped in this notion of great power competition, central to both the 2018 and the 2022 NDS. So, the question becomes what is the democratic response to civil-military fusion? It’s this.

Right now, we are in danger of falling behind in a technological competition with a formidable competitor. I think there’s real urgency around this because I don’t necessarily see that we’re doing the bold things we need to do to compete.

Q. Can you unpack that for us, Joshua?

A. Sure. The thing that really keeps me up at night is that there are enormous efficiencies to being an authoritarian, planned economy that doesn’t respect the rights that we embrace. In America and in the Free World, we believe that the efficiency of resource allocation, the efficiency of capital allocation, and talent flows in a free market will trump the benefits of efficiency that come from a command economy and a totalitarian regime. I’m totally certain our system is the most moral and just, but do we want to bet its survival on the notion that it’s also the most competitive? We need to prioritize winning this contest.

In China, they can be heavy-handed about directing teaming arrangements, mandating cooperation, abrogating rights. They have such an interventionist industrial policy that they are sidestepping a lot of the redundancy and duplication that we have in the free market system. And they are pouring in state resources in ways that are not true to our norms.

So, I think there are constraints in a democratic and free market that we would be naive to discount. We’re not entitled to win. We need to invent a democratic, rights-respecting, free market version of civil military fusion that is efficient and gets the benefits of American creativity and capital allocation.

Right now, we are in danger of falling behind in a technological competition with a formidable competitor. I think there’s real urgency around this because I don’t necessarily see that we’re doing the bold things we need to do to compete. If we create acquisition pathways that combine the might of our legendary Defense Industrial Base with the dynamism and velocity of our venture-backed ecosystem, that’s how we win.

Read the full issue of 

Defense Acquisition magazine

 

The views expressed in this article are those of the author alone and not the Department of Defense or Google. Reproduction or reposting of articles from Defense Acquisition magazine should credit the authors and the magazine.