(Co-Authored with Steven A. Cavaleri)
In our last post, we began a review and commentary on Steven Spear’s post on out learning and out racing the competition. we concluded that post by pointing out that Steve’s conclusion that high velocity organizations institutionalized inductive/deductive problem solving cycles and built new knowledge faster in this way than their competition was quite similar to our own view that open enterprises, or organizations with Open Problem Solving Patterns are more adaptive than others because they initiate and implement Knowledge Life Cycles (KLCs) more frequently that other types of organizations. Steve ended his post with a short discussion of the four capabilities of high velocity organizations, presented in much more detail in Chasing the Rabbit. We’ve already discussed these capabilities once here. In this blog, we want to add to our commentary on his four capabilities. Here’s his statement on Capability 1.
“The ‘rule’ of Capability 1 is: Specify in advance how work is expected to be done with tests built in to indicate immediately where and when a problem is occurring. A problem is either that something is being done or the results being generated are not what was predicted.
This is a nothing but a formalization of the deductive approach.
— Predict what behavior (design of work), in what context (your work site), will lead to what outcomes (the output you think you will generate).
— Identify anomalies–refutations of your existing ‘theory.’ This is the job of the built in tests.”
Capability 1 is a very important and necessary one for high velocity organizations. It’s particularly important because it implies that such organizations set themselves up to see deviations from expectations, and to view those deviations as refutations of the theories or mental models generating the expectations. But we wonder whether this capability, as stated, doesn’t overlap with Capability 2, whether or not this may not introduce a bit of confusion into the framework? We’ll consider this after we look at Steve’s statement of Capability 2.
“Capability 2–problem solving, continues the cycle first with inductive theory building and concluding with deductive theory testing since the disciplined problem solving of high velocity organizations requires:
— Describe the background/context
— ‘Grasp the current condition’–observer and describe what is actually occurring including the problems/symptoms/anomalies that are being experienced.
— Root cause analysis/diagnosis–investigate to find characteristics/features that might explain the anomalous behavior.
When you spin into countermeasure (treatment) construction you are now getting into a deductive mode.
— Based on what you think caused the problem, you construct a target condition.
— This is a prediction of what new outcomes will occur when new actions are taken (work done not as previously, but with the countermeasures in place).”
Look closely at Capability 2. The “deductive mode” referred to just above is said to include constructing a target condition of anticipated new outcomes resulting from the new actions or countermeasures. This is the design work referred to in Capability 1. So, is the Capability to do that work part of Capability 1 or Capability 2? From our point of view, it is part of Capability 2, which leaves in Capability 1 only the capability to use the design and associated expectation to evaluate whether there is a deviation or anomaly or not. And if there is, to decide to refute the current theory underlying the expectation, thereby seeing a problem that must be solved using Capability 2.
Another question about the statement of Capability 2, is whether its assumption that there is a “root cause” is applicable to finding a solution to all or nearly all problems. The idea of “root cause” is tied to the notion of an ‘efficient cause’, a force that brings a result into being. But there are different notions of ’cause’ that are important in human complex systems like organizations, such as ‘downward causation’ or ‘inter-level causality’; and also feedback or ‘self-causation’. Deep knowledge of causation therefore may not be limited to ‘root cause’ in the efficient cause sense of the term. Instead in many business process situations, we might be looking at some combination of efficient, inter-level, and self-causality. Since both of the last two types of cause are either propensity-like or operate dynamically, we may well be looking at probabilistic connections between actions and effects in revised business processes even if we successfully analyze patterns of complex causation.
The above comment is too abstract. Here’s an example. Assume that Toyota has Sales and Marketing Models that guide their efforts in various parts of the world. Those models, and the marketing and sales strategies arising from them may have performed very well until recently. Now, increasingly, predictions from the models may be deviating from expectations. What’s the “root cause”? It may not be that the causes conceptualized in the previous models are in need of modification by identification of some other “root causes.” Instead, it’s much more likely that phase changes in the world economic environment have interacted with feedback loops to reduce demand in ways that could not be predicted by Toyota’s original models.
This point, that ‘root cause’ analysis may not be enough to allow specific predictions in human systems experiencing substantial changes in their environments or unexpected feedback effects on their component agents, should not be taken as a criticism of the idea that problem solving should include specification of clear expectations as part of designing solutions. Instead, it makes such predictions even more necessary, since the more that behavior in our systems is subject to other than efficient root causes, the more we need good baseline predictions to allow us to assess and understand the impact of inter-level and self causality. Let’s now turn to Steve’s last two capabilities.
“As for Capability 3, that requires that when you discover a solution to a problem, you share what you have learned by making clear the inductive and deductive phases of the discovery, not just the solution itself.”
We think the emphasis here on making clear the full context of how the solution was formulated rather than just the solution itself is very important. Frequently, best practices and lessons learned systems fail because they lack that context. People who use context-less systems can’t grasp why the ‘solutions’ in them are ‘knowledge’ and not just ‘information.’ As a result they are much less likely to use the ‘shared knowledge’ than they would otherwise.
Sometimes the term “sharing” refers to any communication of solutions from one person to another. But we prefer a more specific categorization of such communication efforts, including: broadcasting knowledge and information to others asymmetrically; searching and retrieving previously developed codified knowledge and information; sharing codified knowledge and information among peers, teaching knowledge and information to others assumed to be subordinate in a knowledge exchange relationship. Based on our reading of Chasing the Rabbit, we assume that Steve means something similar by ‘sharing’.
“As for Capability 4, that requires that leaders develop these discovery capabilities in those for whom they are responsible.”
Leaders who develop Capabilities 1, 2, and 3, in others are engaging in activity intended to enhance the Problem Solving Pattern (PSP). As we said in Part 4:
“We might call the general process of enhancing an organization’s PSP, Problem Solving Pattern Management (or PSP Management). PSP Management activities include all initiatives directed at enhancing the abilities of an organization’s employees and collectives to perform:
— seeking, recognizing and formulating problems,
— solving problems by developing new solutions, and
— communicating solutions to people who may need them.”
In other words, Steve’s Capability 4 is the capability for high quality Problem Solving Pattern Management.
Finally, there are many views and definitions of Knowledge Management (KM). A recent survey found 62 definitions and excluded a good many we’ve run into ourselves, but won’t mention here. Right now there’s no consensus on the nature and scope of KM. Along with some colleagues, we have been developing a particular conception of KM we’ve called, perhaps immodestly, the New KM. According to that conception:
“Knowledge Management is an activity intended to enhance knowledge processing. Extending this simple statement a bit, it is activity intended to enhance the following behavioral processes:
— problem seeking, recognition, and formulation,
— knowledge production (creating or discovering new knowledge), and
— knowledge integration (making the new knowledge available to others in the system through broadcasting, searching/retrieving, teaching, and sharing).”
Comparing this view of Knowledge Management, with Capability 4, and with our statement on PSP Management, we think that Capability 4, the capability for high quality PSP management, and the capability for high quality KM, are one and the same.
To Be Continued