The book How Google Works, by Eric Schmidt and Jonathan Rosenberg, offers many insights in a post-industrial (or post-corporate) world. I’ll discuss the book’s ideas and relate them to defense acquisition, but first a motivating comment.
I wonder why Eric Schmidt, chair of the Defense Innovation Board (DIB) Advisory Committee, has not translated his book’s insights into actionable reform. In some cases, DIB doctrine goes counter to the book’s ideas. For example, the DIB’s first recommendation under “People and Culture” is “Appoint a Chief Innovation Officer and Build Innovation Capacity into the Workforce.” However, in How Google Works, the authors ridicule this idea and call it a plan to “set up an innovation bureaucracy.” They continue:
A few years ago, a major consulting firm published a report advising all companies to appoint such a “Chief Innovation Officer.” Why? Allegedly, to establish a “uniformity of command” over all the innovation programs. We’re not sure what that means, but we’re pretty sure that “uniformity of command” and “innovation” don’t belong in the same sentence (unless it’s the one you’re reading right now)…
“Innovative people do not need to be told to do it, they need to be allowed to do it.”
Granted, the DIB’s recommendation added “The goal is not to centralize innovation activities, but to foster a culture of innovation and entrepreneurship across the workforce…” But still, it seemed that Schmidt and Rosenberg were painting us a picture of how organization and communication organically create this “innovation network” as opposed to managing it from above. The network is part of the CEOs job — and in defense acquisition that might point to Ellen Lord, USD(A&S).
I will dissect the book, How Google Works, for the insightful work that it truly is, and apply those lessons to the defense acquisition system. I will dive into the core of Schmidt and Rosenberg’s arguments. How does their view contrast with traditional views of management? How can we take what we’ve learned and reform defense?
Overview
Reviews of How Google Works seem to be mixed. There has been some praise (here and here):
Anyone managing technology-focused teams should read this book.
There has also been some pretty harsh criticism. For example, at the NY Times:
Most of these lessons have hardened into conventional wisdom and will not surprise anyone already steeped in Silicon Valley’s infectious dogma… Yet too many of the book’s supposed lessons are annoyingly trivial.
More deeply problematic are some of the book’s ideological inconsistencies.
Not having lived in the Silicon Valley management culture, the lessons are not trivial to me. And indeed, superficial inconsistencies are not reason to knock the ideas. Any book that’s asking big questions of our world should have some tension. RW Emerson reminds us that “A foolish consistency is the hobgoblin of little minds.” I guess logical systems have got to either be inconsistent or incomplete.
At its core, How Google Works effectively argues that we have entered a new “stage” of economic development. It seeks to define the new paradigm that we operate in. First, I’ll elaborate with historical narrative.
For Adam Smith, there were four stages of development: hunters; shepherds; agriculture; and commerce. For Karl Marx, the “contradiction” of commerce was that while it created progress, the plans of some businesses were in competition with, and therefore at odds with, the other business plans. The anarchy of competing plans resulted in continual waste that could have been avoided if the plans were rationally coordinated according to a single plan. Therefore, central economic planning was the next stage of economic development.
The rise of large multi-unit enterprises in the latter half of the 19th century seemed to prove, using market-tested means, that planning was more efficient than prices. Inside firms, management coordinates the specialized activities, not the price mechanism of markets. Alfred Chandler wrote in 1977 that “In many sectors of the economy the visible hand of management replaced what Adam Smith referred to as the invisible hand of market forces.”
The problem with Chandler’s work was that it did not discuss the limits to rational planning. Indeed, right around that time we started to see the demise of the large industrial firm and the rise of firms like Microsoft and Apple, then Google and Amazon. Clayton Christensen wrote a sweeping book on how large firms fail to capture disruptive innovations. Here is Schmidt and Rosenberg:
After the industrial revolution, the definitive twentieth-century institution became The Corporation. Think General Motors, an automobile company where mass production was happening at plants… Meanwhile, both union members on factory floors and white-collar workers in headquarters enjoyed safe careers and comfortable middle-class lifestyles.
In the twenty-first century, The Corporation as a hub of economic activity is being challenged by The Platform… A corporation’s relationship with consumers is one-way. GM decides how to design, manufacture and market a new product to its consumers, and sells it through a network of dealerships. In contrast, a platform has a back-and-forth relationship with consumers and suppliers.
I will discuss platforms and “knowledge workers” later. But at the heart of this move from the Industrial Era of corporations to the Information Era of platforms are two changes. First, the internet has caused transaction costs to plunge, making it more efficient to outsource functions and work with a “bigger and more diverse network of partners.” Second:
… the cost of experimentation and failure has dropped significantly. You see this most dramatically in high-tech industries, where a small team of engineers, developers, and designers can create fabulous products and distribute them for free. It’s ridiculously easy to imagine and create a new product…
But experimentation costs are lower for manufactured goods as well. One can model prototypes digitally, build them with a 3-D printer, market test them online, adjust their design based on resulting data, and even raise production funds online.
This change in the scale of investment to “have a go” has major consequences for styles of project, and more importantly, people management.
I think this is the structural narrative underlying all the advice coming from the book. I also view some tension here. For Adam Smith, commerce was coordinated by prices and little treatment was given to management. It was understood that giving people freedom allowed for the gears to turn in some black box called a firm. For Alfred Chandler, industry was coordinated by scientific management and little treatment was given to its limits or how it changes over time.
Now, for Schmidt and Rosenberg, we have The Platform. It seems that we have these platforms, which can scale and make small companies large very quickly. Google had grown possibly faster than any in history (even though Google may not have done platforms well). But the point of platforms is to be open, share information, and try not to keep too much of the value. So that leaves room for other players, which I assume allows for a dynamic and competitive process where you have disruptive start-ups and large incumbents like Google co-existing. Here are the authors again:
This is the difference between twenty-first and twentieth-century economies. Whereas the twentieth century was dominated by monolithic, closed networks, the twenty-first will be driven by global, open ones. There are platform opportunities all around us. The successful leaders are the ones who discover them.
Successful leaders will create large corporations, I presume. While the fate of these corporations it isn’t clear from the book, it was clear that industrial era management seems to have been totally left behind. In the firm, we are left with a slew of intangibles (software, databases, designs, processes, training, culture), the value of which is not accurately represented by accounting data.
The workings of firms with high intangible investments look something more Smithian. It focuses on the sympathies between decision-makers, and providing maximum freedom when there is interpersonal trust. It focuses on conflicting perspectives, finding propriety in workplace actions, and what lines cannot be crossed. I therefore see How Google Works as part of this emerging literature that brings bottom-up processes and culture into the “visible hand” of project decisions that drive innovation. Like The Theory of Moral Sentiments, these accounts focus on relationships, aspirations, and personal conduct.
When filling in this hole of how innovation is managed, and not simply assume it comes naturally from free people under law, property, and contract, industrial-era thinkers rallied around the policy-administration dichotomy. The policy-makers could set rational policy according to a careful calculation of the costs and benefits of alternative courses of action. Then, all that is left if for the administrators to execute that strategy. Any new information would flow up the chain with decisions flowing back down. In this framework, the people doing the work aren’t really creative, they are working to standing orders.
But How Google Works shows that under uncertainty, most decisions to take action should come from where the knowledge is, which is in the day-to-day operations and not from a distant manager operating behind PERT and Gantt charts. It creates a culture of “Yes”, whereas the policy-administration view was a culture of “No” or at best “Ask First, Then Wait and See”. This all means that there is a premium on goal coordination, which comes through again and again in the book.
When you are foregoing before-the-fact control from the top to coordinate system-wide goals, then you need the decentralized decision-makers to be steered in a common direction so that results appears to have been coordinated after-the-fact. This means employees can’t just be specialists, they need to be generalists.
Rosenberg and Schmidt tell us that praise, and getting employees self-motivated to solve company problems, is one of the secret sauces of management. This is much like Herbert Simon’s 1973 paper on Applying IT to Organizational Design:
The principal normative concern here was to create organizational environments in which employees would be motivated to join the organization, to remain in it, and to contribute vigorously and effectively to its goals… there should be a deliberate sacrifice of effectiveness and efficiency in order to make work itself a rewarding and enjoyable part of human life.
The thing Simon perhaps didn’t appreciate at the time was that taking liberal measures within the firm actually can boost effectiveness. If defense policymakers take this new economic “stage” seriously, and Eric Schmidt is a person with some clout on the DIB, then they should be considering radical change of defense institutions which were designed with the mindset of the 20th century industrial corporation.
This post has taken broad brush strokes to paint a picture that places the ideas of How Google Work. In future posts, I will go into a more detailed comparison of (1) project decisions, (2) culture, and (3) platforms as they are treated in Google and in the Defense Acquisition System. The purpose will be to continually reconnect what is said in the book to what occurs in defense acquisition.
Everything in the defense acquisition process, down to its “PERT and Gantt charts”, was implemented under assumptions that are clearly rejected by Schmidt and Rosenberg. The most important aspect of the DOD’s rational planning system is the Planning-Programming-Budgeting System (PPBS, now PPBE), which requires detailed plans to be explicitly justified long before the fact and generates a “valley of death.” The stage-gate milestone acquisition process and the JCIDS requirements process are secondary to the overarching patterns of funding organizations. The PPBS and its 5-year plans are an anachronism of so-called scientific management carried out to a far more exacting degree than General Motors (the inventor of the PPBS in 1924) or the Soviet Union ever had.
If we take How Google Works to heart, then the patterns of funding should focus on people and organizations first, not programs and requirements. Most of the time the ideas that give you 10x or 1,000x gains require scrapping the old way and trying something new. That requires tacit knowledge; it cannot be based on articulated “business cases” that seek to have answered all the important questions before experimentation, or require the approval of specified requirements, costs, and schedules. The authors promote a “hands-on, trial-and-error approach.” To this effect, an organizational budget would go a long way to increasing speed and flexibility in project decision-making.
This overview has provided a framework for understanding the broad themes of How Google Works. It has also identified the central impediment to defense reform based on its philosophy, namely, the PPBS. None of the DIB recommendations address the elephant in the room, an elephant that will stamp out the intent of its initiatives and leave behind only more processes to follow and boxes to check.
Leave a Reply