What kind of innovation do you want?

Corporate strategist, defense investor and student of complex systems Pierre Chao understands the many shades of innovation, particularly how it differs—widely—from the government to the commercial sector. And he has specific ideas on what makes innovation happen, what impedes it and where innovation should not take place.

Army AL&T most often looks outside the realm of defense for the Critical Thinking column, but for this issue on innovation, we thought it important to stay closer to home. That’s because innovation in government and defense, and innovation in private industry—while not completely different—have significantly different incentive structures, an issue that’s important to Pierre Chao, who has been involved with the topic for many years. Chao has served on several Defense Science Board task forces, on a presidential commission and as a guest lecturer at both National Defense University and the Defense Acquisition University.

For five years, Chao was a senior fellow and director of defense-industrial initiatives at the Center for Strategic and International Studies, where he is still a senior associate, before moving on in 2008 to co-found Renaissance Strategic Advisors, based in Arlington, VA, and Enlightenment Capital, based in Chevy Chase, MD. Renaissance advises clients in four areas: corporate strategy, market analytics, mergers-and-acquisitions due diligence and transaction advisory. Those clients tend to be people on boards of directors or who have “chief” as the first word in their titles—chief executive officer, chief financial officer and so forth. Enlightenment Capital is an investor in small and medium-sized aerospace, defense and government services firms.

With an undergraduate degree in political science and management science from MIT, Chao describes himself as “a systems engineering-like thinker and pattern recognizer by genetics.” He has always been fascinated with the defense industry, he said when we spoke with him by telephone on Aug. 19. “It’s one of the few sectors that has a blend of technology, business, politics, finance, international relations; where they all come together and those complex intersections are relevant,” he said. Interested in complex systems since he was a child, he said that the defense industry is “eternally fascinating to me to the extent that all these factors always come into play. And I would argue that if you look at the problem from just one of those lenses, you’re going to miss a huge chunk of the issue.”

Add to that an interest in history and military technology and an affection for multi- and interdisciplinary thinking, and it’s easy to understand how Chao has participated in more than $12 billion in financial deals in the industry, including mergers and acquisitions as well as initial public offerings. We wanted to talk with Chao about innovation—what it is and what makes it happen, including the topic of acquisition reform. For Chao, you can’t really talk about innovation without talking about acquisition reform, something he’s intimately familiar with.

Army AL&T: Innovation is an odd thing in that it seems to mean different things to different people in different contexts. How would you define innovation?

Chao: Yeah, I agree. These days, innovation is a loosely bandied term and many times ill-defined because there are so many different types of innovation, and they’re all important. I think too often people default immediately to the version of innovation that describes a massive technological breakthrough, the disruptive kind of innovation. And yet the reality is that, more broadly speaking, there’s technological innovation, there’s process innovation, and you have business model innovation.

Business model innovation occurs when we invent a way to do things differently. In some cases, you can get a major revolution in how things are done or how markets will work just by changing business models—for example, by rethinking tasks that were inherently governmental and then outsourcing them as a service or a product, or capabilities that are undertaken as a service or solution rather than delivered as a piece of hardware. There are also plenty of dramatic commercial examples—consider what Amazon or the other online stores did to the brick-and-mortar retail industry. Or the idea of a credit card versus carrying cash—that’s a business model innovation.

Process innovation is more around the idea of, “How do you build things better, how do you manufacture, how do you improve?”—not necessarily changing the product, but finding ways to manufacture far more efficiently. Say a particular technology is well-established and well-understood—mature. It’s sometimes far more important to have process innovation to try to bring the costs of mature technologies down and make them more broadly available. It’s not necessarily about reinventing the technology or the product.

Finally, you have the kind of innovation that I think most people are familiar with, which is developing a new product. And there, you need to further subdivide it into two types. Disruptive innovation—I’m going to bring in a fundamentally new technology to completely change a market, which is a rare case. [Then] a more likely form of innovation, what people refer to as incremental innovation, where they’re steadily improving the product.

When somebody says, “I want innovation,” your first question back should be, “Well, what kind?”

Army AL&T: In an interview with Defense News at the recent Paris Air Show, you seemed to be saying that if you follow the money, you can find innovation.

Chao: The point was a little bit the reverse. As DOD or government is trying to encourage innovation, one of the things that in some ways make the government different than the commercial market is that, in the commercial market, you can create your own market. To the extent that you invent a product in the commercial world, it’s also possible to invent the market. There was no such thing as the iPad market until Apple invented it; we all discovered that we wanted or “needed” one, and voila, the market category is created. The government market doesn’t work that way: There needs to be a budget to buy an innovation and often, if a new technology is created but doesn’t match a budget bucket, it takes time to create one. The creation of that technology bucket is not purely market- or demand-driven: there’s politics, bureaucratic policies and other drivers.

And so there you need more of a demand signal from the customer in the defense world in order to stimulate innovation: “This set of capabilities is important to me,” or, “Solving this problem is important to us.”

The part that’s becoming very difficult for the industry is that, during the Cold War, you had a very large demand signal: “Beat the Soviet Union.” That allowed industry to self-organize around that grand strategic goal. As long as you were working to solve those kinds of problems, you were pointing in the right direction. That became extremely muddied at the end of the Cold War, where we lost that demand signal, and that was about the last time we had, I would argue, an innovation challenge that was so clearly defined at a grand strategic level.

And then the wars in Iraq and Afghanistan came along, and that, at least, provided some guidance, or some demand signals, and you saw people step up to the plate with innovations to solve the war’s problems—everything from MRAPs [mine-resistant, ambush-­protected vehicles] to different kinds of sensors or counter-IED [improvised explosive device] gear or UAVs [unmanned aerial vehicles]. People knew where to innovate because they had a clear capability demand signal.

Now, we’re back in this mode where industry has lost the demand signal—which priorities are important versus which ones are not. Is it to solve the high-end threat? Is it to solve the issues related to regional threats? Or is it still the terrorism challenge?

Army AL&T: You talked about mature technologies, and one of the things that comes to mind is enabling technologies. Tim Berners-Lee developed HTML and had really no idea where it would go, but knew that it could go somewhere. How important is that sort of thing—having those enabling technologies—to innovation?

Chao: Oh, it’s critical. That’s why throughout history innovation has not occurred in a linear fashion. There are bursts of activity. There’s a fundamental, underlying, enabling technology that gets discovered—electricity, the internal combustion engine or the microprocessor—that drives follow-on innovation. The modern world is still experiencing the aftereffects of the invention of the microprocessor and integrated circuit. We’ve been living in an era that’s been generally tied to Moore’s Law: processing power doubling every two years. And now it’s combined with the revolution in telecommunications, enabling the sets of technologies that wire the world together, providing the infrastructure for the Internet revolution—which then permits collaboration at a scale and level that’s never existed before, triggering social and business model revolutions.

So, where, once upon a time, somebody could invent something two continents away—and it would take time for that to proliferate—what occurs today is spreading in near-real time.

We have Moore’s Law that’s been clicking away, making processing power stronger and stronger, and storage is becoming cheaper and cheaper. It’s enabling all kinds of capabilities that we’ve always held out there as promising. Modeling and simulation [M&S], for example: People for a long time held out the promise and premise of what M&S can do. It’s only now, because we finally have the processing power to do the things we’ve dreamed of, that it’s finally coming into what people have been promising for 20 years in terms of capabilities.

Same thing with expert systems, AI [artificial intelligence]—whatever sets of terms you want to use around autonomy and more intelligent machines—we’re finally able to do some of the things that people were dreaming about 10, 15 years ago. Our autonomous vehicles and robots are getting better and better. We’re experimenting with self-driving cars, and it’s getting harder to distinguish between human and automated responses in call centers. In a few more Moore’s Law cycles, we’ll be able to put as many circuits on a computer chip as there are neurons in the brain (86 to 100 billion). That will be in our lifetime. Then who knows what machines will be able to do?

Army AL&T: You were talking earlier about process innovations. Things like continuous process improvement and Lean Six Sigma seem to have the power to move very large bureaucracies that by nature don’t want to move. Is it possible for the Army to make that a lot more pervasive part of the culture, even if you can’t have a flat military organization with very little hierarchy?

Chao: You hit on this topic with that one, right word, which is “culture.” In many cases, part of process innovation becomes a major culture issue. Historically you don’t get that kind of change in culture without a crisis of one form or another that forces fundamental relooks at how you organize, how you do things. That crisis for a military can be the beginning of conflict, and now all of your theory about how things should work goes out the window. You’re facing a true adversary and you’ve got to rethink how you do things. In some ways, I would argue, the Army has been forced to get a little bit flatter because we’ve been fighting a very flat, nonhierarchical adversary, one who certainly has a very different organizational structure. In some cases we’ve had to match that. We’ve learned lots of lessons on how to do the “light fight” in the last decade.

Coming out of a war, we have a different set of challenges—budgetary ones that will once again force a look at how we’re structured and how efficient the institution is, and how you function and maintain capability with far less budget and yet not much relief in terms of what’s expected of the Army from a capability standpoint. So those kinds of [things] usually drive the search for efficiencies or create the imperative for business model innovation.

There’s another element at play that relates to the topic of enabling technologies. The better you can measure the performance of an organization and the better you can analyze and understand the process of how your organization works, the more efficiently you can run it. It’s the equivalent of medieval architecture versus Renaissance architecture—as mathematics and the understanding of physics got better, you could reduce the tolerances and you didn’t have to build in as much margin because you could better analyze the structure. I think corporate America has been going through the same phenomenon. Operations research theory and tools have improved and we can better understand how it [corporate America] functions. So, therefore, you can more efficiently organize today than you could 20, 50, 100 years ago.

There is a limit, though, to how we design an organization that lives at the edges of tolerances and extremes of efficiency. Part of the inherent tension in any military organization is that you actually want to build in larger margins of error than in the commercial world because you face more extreme potential results in the case of failure or just bad luck. The consequences of burning through that margin are far more significant than in the commercial world—people die, battles are lost, countries fall.

So I would argue [that] it’s a falsehood to claim that you’d ever want a military organization to be as efficient as a commercial organization. The consequences of failure are so large that a certain amount of inefficiency should be accepted to create redundancy and “strategic reserves.” This is where the real subtleties come into play: If we want parts of our military organization to have inherent reserves, margin for error, inherent inefficiency relative to the optimum, then, in a tighter budget environment you would want—perhaps I should say need—all the noncritical functions to be extremely efficient to afford the cushion you want on the pointier end of the spear.

Army AL&T: People like Elon Musk, of Tesla and SpaceX, and others talk about how failure is not only an option but a must because if things aren’t failing, you are not innovating enough. Government is a different story. No one wants to fail the taxpayer.

Chao: Well, that goes back again to a cultural issue. That wasn’t always the case, and the tolerance for failure is something that we have lost, I would argue, over the course of the 1980s, 1990s, 2000s.

I think the willingness to tolerate risk has returned somewhat. The conflicts of the 2000s and 2010s created an imperative to take some level of increased risk. There’s a good historical case study to prove that as an institution, we need to accept a higher level of risk if we want innovation. For example, look at the Poseidon missile program and you will see old video footage of the test missiles coming up out of the ocean, spinning out of control and exploding. We tolerated those failures then. But, to your point, today, probably within 24 hours you’d have Congress all over you because those same videos would be on CNN a thousand times a day, and that creates that pressure not to have those types of mistakes. And yet it is exactly those kinds of mistakes that drive understanding and knowledge and innovation, frankly. And so the fear of failure that has crept into the system, into the culture, over the last 30 years—I believe—is a really dangerous thing. And this institutional fear of failure can only be beaten back, I would argue, by leadership and the willingness to protect those who take risks and fail, and celebrate [those risks] as opposed to knocking them [the people] down.

We certainly have institutions that have that fearlessness about failure, in places like DARPA [the Defense Advanced Research Projects Agency] or the NASA of the Apollo era, historically. So, that tells me that it’s not genetically impossible for the Pentagon, that it’s something cultural that’s crept in.

In a combat situation, the unwillingness to accept failure and a high level of caution can often cause strategic failure, failure on a grand scale. In peacetime environments, you don’t have something that creates a cost for being too cautious. And so, the system falls into, frankly, a bureaucratic stupor. I hate to use a controversial phrase.

Army AL&T: That’s a controversial phrase, bureaucratic stupor?

Chao: Yeah, where people check boxes and, because there’s no other way to distinguish between who is performing really well versus not, you pick away at every little failure as a way to differentiate between people.

It’s such a fundamental point. If you want innovation, you need to accept failure because you are asking the organization to push its boundaries. Which raises another point: You don’t necessarily want your entire organization to be focused on innovation—and to be clear, I’m talking about the extreme version of innovation centered on looking for major disruptive technology. Right? You only need a certain portion of your organization living on the edge, seeking extreme disruptive innovation. There are some parts of the organization that you frankly don’t want innovation in. I don’t want somebody experimenting around well-proven processes for managing the safety of a nuclear weapons stockpile, for example. It works, stick to the plan. Don’t go goofy. It goes back to the [earlier] topic: Recognize that there are different kinds of innovation and tailor different parts of the organization to focus on them.

Army AL&T: When you look at something like Better Buying Power, some of which seems to be more back-to-basics than revolutionary, do you think that it’s striking the right balance between technological and process innovation?

Chao: Acquisition reform is one of these perpetual activities. After having studied acquisition reform for a long time and participating in a lot of the acquisition reform studies and efforts over the last 20 years, I think what you end up observing is that—the knee-jerk reaction to date has often been, “Let’s rearrange the boxes or create czars or change the rules,” when, in the end, getting back to the basics is where you should want to be. What I find encouraging about the latest round of acquisition reform discussions, and what Better Buying Power 1, 2 and 3 are beginning to embed, is the topic of incentives.

It doesn’t matter what rules you set, it doesn’t matter how many boxes you change. If you set the incentives the wrong way, then you’re not going to get behavior change. This is why the topic that we just covered, willingness to take risk, is one of those elements that is disincentivized in the system.

If I’m a program manager in a company, and I take risks and I fail for legitimate reasons, not incompetence, and I’m punished for it—I lose my job, I lose my program, I lose funding, what kind of behavior change do you think that’s going to trigger throughout the entire organization? Or if a company spends its own money inventing a product and the government grabs its intellectual property or insists on having [that] intellectual property so it can subsequently do open bidding on the manufacturing to crush the margins down, what do you the reaction is going to be the next time around that they’re being asked to put money forward to invent something on their own?

So, ironically enough, acquisition reform is relatively simple if you get back to the basics, and I’m encouraged that [Undersecretary of Defense for Acquisition, Technology and Logistics] Frank Kendall, and many in the senior leadership in the Pentagon and the services, and certainly others like [Rep.] Mac Thornberry [R, TX] understand this. Congressman Thornberry’s acquisition reform efforts, I think, have begun to zoom in on this topic of making sure that the incentives are set right.

Army AL&T: Speaking of incentives, the Joint Light Tactical Vehicle program had three companies that were competing, and each built its own competitive prototype. And then it was a winner-take-all decision. Certainly those companies are getting paid for their work, but after scaling up, two companies weren’t going to win. They’re going to lose jobs. Should it be winner-take-all?

Chao: This is one of the huge tension issues that’s worth a deep look. We’re doing these big, joint programs to have large enough scale in the programs to entice people to pursue them, invest bid-and-proposal dollars, and undertake independent research and development, again, because they have sufficient scale to justify the investment required. On the other hand, it’s creating these winner-take-all situations in which—to your point—once you lose it, then we shouldn’t be surprised that one of those losers ends up exiting the business entirely because, well, now there’s nothing to go for until the next decade. For a midsize firm, it may cause them to exit that particular line of work or sell. For a larger, multibillion-dollar prime, it may cause them to exit that segment—or, if they see long-term potential, then hunker down.

The size of the budget will determine the health of programs. But it’s the number of points of competition that will determine the health of an industry. So, when you go to one Joint Strike Fighter, you shouldn’t be surprised that you’ll lose a whole bunch of people that were competing in the manned fighter business, and now sit there and say, “Well, there’s nothing to compete for [over] the next decade.”

So, that is a deep issue in terms of future acquisition strategies and how you want to be building much broader acquisition strategies because you have those two tensions—create programs big enough to attract investment, but not so large that once they are decided you lose an industry. Some of the acquisition reform efforts address this and say, “Maybe we should have more X programs or other things so that when someone loses one of the big ones, they don’t drop out; they’ll have other things to hold them in.”

I would argue if you want a broad, healthier industry, you want to have lots of smaller programs rather than huge ones. This goes back again to where you need to be sophisticated in your acquisition strategies. If it’s an older, more mature technology, and you think it’s going to be the last of something, then maybe you don’t mind that you’re collapsing the industrial base to a monopoly. But if it’s an area where there’s a lot of innovation, then you probably don’t want to be creating these big, mega-programs because you don’t want to be encouraging early exits by competitors, have them just give up because they don’t see another entry point.

Army AL&T: In terms of a big program and you have a product that’s supposed to last for decades, should there be some kind of mechanism where there’s some competitive hound snapping at the incumbent’s heels so they don’t get complacent because they’re locked in for decades?

Chao: Part of that depends on the broader technology cycles inside that industry. If the technology is not evolving that rapidly, or evolving at a 10-year or 20-year pace, then it’s probably OK to lock in a vendor for that period of time. If you’re seeing, however, as in parts of the commercial world, a technological evolution cycle of three years, then a decade or more without competition is probably too long. So, the acquisition strategy and the industrial-base policy strategy should not be divorced from an understanding of the technology development cycles. This is a sophisticated game, and one-size-fits-all policies will inevitably fail.

For example, one reform concept today is [that] we should make everything follow a spiral development program process. Spiral development, where you see how much you can get in a fixed time frame and then field it and move on to the next spiral, is very good for IT programs or other places where the technology is moving relatively quickly. It’s not necessarily correct for other programs where the technology is moving slower or needs a longer time to mature.

Army AL&T: Steve Jobs said, “You’ve got to start with the customer experience and work backward to the technology. You can’t start with the technology and try to figure out where you’re going to try to sell it.” With Soldiers as the customer, have we got the equation right?

Chao: That is a key point. In the end, the Soldiers are the customer. They are the user. And the disconnect between the customer and the buyers was one of those topics the acquisition reform efforts were trying to fix in the ’90s because they were beginning to drift apart. And then the war brought everybody back together. It dragged the “system” kicking and screaming into the current age because the [Soldiers] were, in frustration, becoming their own requirements generators and acquirers and going to REI to buy the gear on their own. I often made a point in the beginning of those war years: “Go take a picture of that Soldier’s gear, lay it all out on the floor, and draw a box around how much they were issued versus how much they were going to REI or some other store and buying.” [That] signaled that the acquisitions system had to catch up. And it did. We got multiple rapid-equipping types of mechanisms to change that.

As we get out of the war, [it’s] imperative that we don’t lose all of those hard lessons we learned. I think you’re seeing a growing sense that we do need to involve the users—whether it’s through the various mechanisms that the Army has established to get their input or through the COCOMs [combatant commands]—in terms of generating the requirements and needs. All of this calls for a better dialogue and interface between Soldier and developer. Systems just developed by the acquisition system, unhooked from users, unhooked from reality, will create a problem. On the flip side, users also don’t often know exactly what they want, either. Right? If you just follow the polls or the surveys, you will also be wrong. At times there’s interpretation—Jobs said start with the user experience, not necessarily the user. I’m sure a poll of the customer asking, “What do you want?” would never have come up with the iPod. It’s that creative tension between the two methods that usually generates the best results.

Having as much interaction as possible between the user, the buying community that’s trying to get things to you, the industry and the technological creators—that’s key. The growth of the use of IPTs [integrated product teams] is all a reaction, I think, to this fundamental truth that if you try to build a product in a disconnected fashion, you have a higher chance of failure.

Army AL&T: Engineers solve problems, but they also tend to focus on features over benefits. So the Soldier, the customer, might become secondary to the benefit because the engineers are so focused on solving feature problems.

Chao: This is related to the question we just discussed. Jobs said solve the user experience. The same could be said: Solve the Soldier’s problems. The best innovation comes from thinking about the problem as a whole—not by thinking about the technology or the feature. The benefit is derived from the fact that you are solving a real problem. I heard a story from a small startup company in the defense sector, founded by an ex-military person. His company was creating handheld devices to pull down fused intelligence data. He created the company because he was frustrated by what he had in the field—they were getting these big, clunky devices that had very nice touch screens developed by, I’m sure, very smart engineers. What they didn’t know or forgot or ignored, was that the Soldier was usually wearing gloves when operating these devices. Nice feature; useless in the field. Knowledge of the user experience allowed his firm to develop a better product.

Another example is the evolution of stealth. We didn’t start [out to] invent stealth out of whole cloth and without context because somebody said “I want to be invisible.” They were trying to solve a problem: “How do I penetrate an ever-better-developing Soviet air defense system?” It was costing a fortune to brute-force our way through it with electronic warfare and better-performing aircraft. We were starting to climb the steep part of the cost versus capability curve. And it was in trying to solve a problem that stealth came about—a very innovative answer that wasn’t an even-stronger electronic warfare box or ever-faster or higher aircraft, but a completely different way to solve an old problem. Wartime situations serve up your problems in a very stark and fast fashion.

The harder times are in peacetime situations where you’ve got to think through more, “What are the problems I’m trying to solve?” And that’s why we’re entering a period where the imperative has gone way up for leadership to send those demand signals and those priorities—“These are the problems that are really important to me as an institution that I would like to try to solve”—and then let the creative geniuses of the industry and everyone else try to solve it.

It goes wrong when you either muddy up what your priorities are or when you’re too prescriptive: “I want you to solve my problem by having you develop a better ‘black box.’ ” Really? Are you sure? How do you know that’s the best solution? What problem are you really trying to solve? Because maybe I’ve got a fundamentally different way than just a better black box.

Army AL&T: You said something almost 10 years ago about the shift from “what to buy” to “how to buy.” In your mind, what’s the difference?

Chao: I think a lot of the acquisition reform efforts of the last 30 years, which were the reactions to some of the procurement scandals of the ‘80s and ‘90s, overly focused the Acquisition Corps on how it was buying—placing all these layers of oversight and double-checking and rules and checking boxes every step of the way. The focus was on process reform: “Am I buying the things that I’m buying well?” as opposed to the strategic question of “Are we buying the right things?”

It doesn’t mean that you need to go willy-nilly and throw all the rules created over the last 30 years out. Some of those rules were put in place for a reason, but the original rationale may have gone away or the fix may be worse than the problem. The pendulum has swung really hard, I think, and probably overshot it, from the perspective of having the entire Acquisition Corps zoomed in on process—the “how” versus the strategic question of “what.”

Army AL&T: If you were king of acquisition and you could change or “reform” any aspect of acquisition as it exists today, what would you change?

Chao: I’d zoom in on the incentives structure we have for industry and for program managers and PEOs [program executive offices]. I think it’s so fundamental.

Today, we still have incentive structures that are misaligned. And, certainly, a lot of those misalignments are cultural. For example, if you have a well-performing program, usually what happens is that you don’t get rewarded for it. Your money gets taken away—if you’re under budget, you’re going to have your money taken away and given to an underperforming program. That’s not a good incentive for delivering good performance. Our overall profit philosophy is off. We still have a system that would rather pay $100 and 5 percent margins for something rather than $80 and 20 percent margins. That creates very perverse incentives.

King for a day? I’d spend a lot of brainpower and resources identifying those issues and trying to come up with offsets or fixes to those incentive structures. To be extreme and make a point, I can say it really has nothing to do with the rules or the FAR [Federal Acquisition Regulation]—which is actually pretty flexible at its heart; it has nothing to do with organizations and boxes and people and all that. It’s behavior that we’ve just sort of embedded in the system—which parts of the rules we’ve decided to emphasize.

Army AL&T: We’re talking culture again?

Chao: Yeah, we are, which is what makes this so hard. I have no illusions about how difficult these things are, because in the end, you’re talking about culture. And culture is very, very difficult to change.

---

This article was originally published in the October – December 2015 issue of Army AL&T magazine.

Subscribe to Army AL&T News, the premier online news source for the Acquisition, Logistics, and Technology (AL&T) Workforce. Subscribe

---