My friend Steve Holt - a very brainy bloke who uses TOC to make aeroplanes (or something like that) - sent me the following note which, while it is about using Theory of Constraints at a strategic level, describes a fascinating analogy which I need to include in my boo. The analogy is that just as buildings can be designed to survive the severe jolts of earthquakes, so can businesses be designed to survive the severe jolts in their environments. This thinking applies to software development projects and systems too, at many levels.
Steve is much better with words than I, so I'll let his words speak for him:
I've recently been reading an interesting thesis out of the MIT Leaders for Manufacturing program. LFM is a dual engineering MS degree and MBA; the MS from MIT Engineering school and the MBA from the Sloan Management School. The students are all from industry and each has a 6 month assignment at a company, which forms the basis of their thesis. The program tends to attract some pretty high caliber talent and, consequently, some of the theses are quite excellent. Over the years there have been a number of great ones that touch on TOC; for example, one from a couple years back by Vik Sahney comparing and contrasting a Toyota-like kanban system, CCPM, and a dynamic optimization approach to building large complex products . His conclusion was that CCPM was the way to go.
But, the one that's recently caught my attention is by Ted Piepenbrock from 2004. The thesis is available from the Lean Aerospace Initiative web site (since Ted was a researcher there). The LAI link is: http://lean.mit.edu/index.php?option=com_docman&task=doc_details&gid=175&Itemid=88 ) The thesis is titled, "Enterprise Design for Dynamic Complexity: Architecting & Engineering Organizations using System & Structural Dynamics."
Ted is a civil engineer by training and the thesis makes the parallel throughout of building design and company design. Specifically, he uses the principles of earthquake design to draw analogies and conclusions with respect to company design. Earthquakes jolt a building with random waves of energy. Similarly, customers can "jolt" a company with waves of random demand. You can design a building to resist earthquakes by making it strong, so that the energy is transferred to the building and the building can hold up to it. Similarly, a company can be designed to adapt to and follow market demands. Imagine building a building and then needing to add on to it--you'd have to make it increasingly strong in order to be able to handle the energy. Likewise you'd have to build massive capacity to follow a growing market.
But, earthquakes are unpredictable. If your building is over designed it means you're using way more material than you'd need to. If it's undersigned it becomes rubble. Likewise, a company that tries to rapidly follow the market enters into boom and bust cycles. It is nearly impossible to accurately predict the "just enough" growth rate when times are good. Piepenbrock calls these Blue companies and they typically are companies that got into an industry at the very beginning, when it was all about growth and tie their future to market demands. His thesis talks about two industries, cars and aircraft. He says that example Blue companies are Ford, GM, and Boeing. All of them have had boom and bust cycles throughout their corporate histories typified by hiring and firing binges and buying and dumping assets as they alternatively over and under react to the market.
But, what do you do if you want to have a building survive earthquakes and you don't have a lot of extra material to use? You can build in a way to isolate it from the earthquake energy. The simplest way to do that is with the use of a base isolator, a means by which the building can transfer some of the earthquake energy into pendulum energy. It still shakes, but at a slower frequency and not as much. If you really need the ultimate in resistance you can build in an active damping system that senses and counteracts the earthquake energy. Since less of the earthquake energy gets transferred to the building, the building can be lighter and/or taller.
So, what's a way to create the intent of a base isolator in business? Piepenbrock says that you need to find a way to isolate yourself from the variation of the market so that you can always sell your product no matter what the market conditions and always provide a safe, no lay off, environment for your employees. Now, about here this should start sounding like the 3 necessary conditions for a successful strategy that are in the last chapters of "It's Not Luck." Namely:
NC 1: The company will make money now and in the future.
NC 2: The company will provide a safe and secure environment for its employees now and in the future.
NC 3: The company will provide a product/service valued in the marketplace now and in the future.
So, you could provide a business base isolator by keeping the constraint internal to your company (always keeping a demand for the product and growing slowly primarily by internal productivity improvements), ensuring that you have a no lay off practice so that employees are dedicated to improvements, and being tightly connected to what consumers want so that your products/services are in high demand.
And that's precisely what Piepenbrock says that successful companies that he calls Red Companies, do. He includes in the Red list: Toyota, Southwest Airlines, and Airbus (note that the thesis was released in 2004). Each of these companies was formed in mature, supposedly saturated, markets that had high start up capital hurdles. They simply didn't have the free cash or other resources to go after the (non-existent) unsatisfied market and, instead, had to compete by growing internally and niche marketing (i.e., segmenting the markets).
At this point you may be wondering, "Is Steve just showing that he can force fit any possible scenario into a discussion of TOC?" That's a valid question (and not without precedent), so I'll quote from Piepenbrock's thesis.
Page 259:
"Designers can take out the demand uncertainty by installing a base isolator under the building (i.e. at the interface with the customer). This limits and controls the amount of demand that can enter the enterprise, making it possible to take out wasted structure and capacity in the superstructure of the enterprise (i.e., define the weakest link and subordinate all others to it)."
OK, that part about weakest link and subordination does sound like TOC. (And you can come up with a good explanation of why the distribution solution works from reading the rest of this section and thinking about vibration dampers.)
Page 265: Section 9.8 Base Isolation and the Theory of Constraints
"I will argue later that a lean enterprise in fact utilizes the theory of constraints philosophy in order to achieve sustainable business results."
Hang on, did he just say that successful Lean companies got that way because of TOC? Yes, he did. (But it's not "active" or intentional TOC, it's more of an evolutionary understanding of what the constraint is and managing to it. Remember, I'm picking quotes out of a 341 page thesis. I'm leaving out a lot.)
Page 281:
"It is clear that uncommon corporate performance can be achieved through the counter-intuitive enterprise designs. Toyota and Southwest Airlines have both achieved stock market capitalizations greater than the sum of all of their major competitors in recent years through the adoption of lean enterprise architecture.
Their success lies in the counter-intuitive notions of extending the boundaries of their respective enterprises to include those stakeholders which can influence shareholder value, and then work to build long-term win-win relationships with them. As such, they have taken the counterintuitive steps to ensure no layoffs in the 50 and 30 year corporate histories of their companies amidst industry norms quite to the contrary. They have also demonstrated the courageous restraint by growing (largely) organically at the rate which can sustain learning within their core strategic assets, their people. This rate, which has been approximately constant at 10% over the lives of their corporations is substantially lower than the maximum possible growth rate that the markets demand. They have designed into their "integrated enterprise teams" (IETs) a base isolation fuse which mandates that they manage to the signal of the customer markets and not the noise of the capital markets. The irony is that they are highly valued by those same capital markets, that it is in their enterprise architecture to ignore. They recognize that maximization of shareholder value is achieved via management of stakeholder value."
That is, he's saying that Toyota and Southwest have been profitably rewarded by the stock markets basically because they ignore the stock markets and pay attention to their customers instead and by intentionally holding back growth to match capability. I particularly like "they manage to the signal of the customer markets and not the noise of the capital markets."
Page 286:
"9.20 Chapter Summary
In this chapter, I introduced the concept of "enterprise base isolation" as a potentially highly effective means to control the dynamics of an enterprise to its benefit. The physics of base isolation is rooted in the theory of constraints, whereby the enterprise leader explicitly defines and designs the bottleneck policy, to achieve stability in the enterprise. I then explain how to define this policy, which entails injecting flexibility and damping into the enterprise at its "substructure" which allows the "superstructure" of the enterprise to be made more efficient.
From this, I argue that this physics-based policy can be seen as a scientific definition of a "Lean Enterprise". Using the laws of enterprise physics, I describe "lean" as the eradication of waste, inflexibility and variability and I note that the way to implement lean is through stability, flow, takt and pull. I observe that world-class lean enterprises like Toyota and Southwest Airlines expertly utilize enterprise base isolation to their great benefit."
There you have it, he concludes that a company's ability to create systems capable of stability, flow, takt and pull (i.e., Lean) requires that first they must be operating to a management system based on the careful and intentional selection of an enterprise constraint.
Piepenbrock is currently in the doctoral program at MIT in the System Dynamics area. Both John Sterman and Peter Senge are on his committee. It'll be interesting to see what he comes out with.
Best regards,
Steve Holt
