Showing posts with label Replication. Show all posts
Showing posts with label Replication. Show all posts

Friday, March 23, 2007

Replicating Organizational Knowledge: Principles or Templates?


Professor Sidney G. Winter
Professor Charles Baden - Fuller

Another paper by Professor Winter, this one co-authored with Professor Charles Baden - Fuller. It poses an interesting question with respect to the activities of the oil and gas industry. Should the homogenization of the administration occur? Will it lead to a lack of competitiveness in the industry? I assert that the producers competitive advantages are sought by their strategic land position, their management and the organizations ability to find and produce commercial quantities of oil and gas. The administrative burden has been codified to a large extent by the various associations and non-profit organizations that have standardized the data model (PPDM) defined the operating requirements (CAPL) asserted the acceptable accounting requirements (COPAS and PASC). A firms ability to differentiate their operations on the basis of the administrative requirements would be an abject failure at the word go. The industry has progressed very far in terms of the acceptable behaviours in the industry. The secret is to have a solid enough inventory of land, seismic and scientists pushing the edge of their sciences. Administration is an after thought, not a competitive factor.

This brings up the issue of how is the homogenization of the oil and gas industries going to occur. Simply its not, the administration will become less of a burden and the speed and innovation of the industry will no longer be organizationally constrained. Specifically, how are the optimum methods replicated from task to task, person to person and producer to producer? This is the topic of Professor Winter's et al paper. Using the two methods of "Principles" and "Templates" the authors come up with some surprisingly strong recommendations.

Abstract

"We address the challenges of providing operational measure for successful replication, and for comparing the efficacy of principles and templates. By using two longitudinal case studies of replication across the units of two multi-unity organizations, we support the central claim that in certain circumstances replication by principles can be as speedy and cost effective as replication with templates, and deliver results of comparable quality. The principle contingencies affecting the relative performance of the two methods are identified. We also point out that replication efforts can be a source or incubator, as well as an application area, for dynamic capabilities in an organization. We briefly suggest what the results may mean for theories of knowledge - based competition." pp. 2
Of the case examples one firm is in the retail distribution of oil products, with the other case example of the hotel chain in France, do not provide any direct relevance from an industry stand point. However from a large organizational point of view the case examples are effective in displaying the results of this study and their applicability to any and all organizations. The one key difference from my point of view is the capital intensive nature of the energy industry makes these recommendations more valid as the exposure to loss is higher then in a traditional labor intensive business. The bigger issue is the need and opportunity to have the standardized processes codified from an industry wide point of view. How these are replicated starts with a few quotations from Winter et al.
"In replication, an organization is intentionally reproducing or diffusing the success it has itself enjoyed in some limited setting or locale." pp. 3
"The value of replication is therefore the ability to diffuse faster than rivals can either imitate or innovate." pp. 3
"This paper seeks to clarify our understanding of the replication of organizational knowledge by introducing a distinction that has been little noticed. Our central thesis is that most organizations adopt some combination of two strategies or approaches, which we call "Principles" and "Templates". The guidance provided by "Principals" has the flavor "Let me explain why this works and the reasons why I do it is this and then try to make it work yourself - I will comment on any mistakes I see". The "Templates" approach is suggested by "Watch very carefully how I do this; then copy what I do and try hard to copy it exactly - but don't ask me why." The word why is clearly central to this distinction, being at the core of one approach while often considered a pitfall in the other." pp. 4
"Both approaches to replication can be supported by codification - by which we mean a "how to" manual recorded in the symbols of some appropriate, possibly technical, language." pp. 4
"Hence, it is rarely if ever the case that replication can be accomplished merely by supplying the manual to the recipient." pp. 5
The first question we should ask is how much tacit knowledge can be captured and codified in software? The answer to this question may vary depending on your IT capabilities and vision, I see significant benefit from the codification of the knowledge. For example, one of the most clarifying aspects of how the administrative issues are absorbed is with the use of schematics. Understanding of how the oil and gas industry operates is reflected in the physical makeup and physical flows of oil and gas. In terms of codification of the understanding of the "principles" is to have a schematic overlay appear in Google Maps. If that was possible with a dedicated industry related software developer, these schematics could populate the overlay with production flows, it becomes evident to the user how the energy related issues will play out. This is also as a result of the standardization of reporting processes primarily dictated to by the Alberta Government for its royalties, and the Energy Utilities Board here in Canada.
"In what follows, we further develop the contrast between the two approaches and seek to understand the circumstance in which each might be superior. The templates approach is understood, believed in and widely relied upon by managers in retailing and other sectors." pp. 5
"After a digression into history, the paper starts by exploring what is meant by replication, templates, and principles. It then proves the challenging problem of how we can determine whether replication has actually occurred. This sets the stage for the two in-depth case studies, which illustrate how replication by principles works; we finally discuss the factors affecting its success in the cases and in general." pp. 5
"... this opening section is a short digression into the field of economic history to show that using principles to recreate an existing success has a long documented history." pp. 6
"For example, he noted that Ford's moving production line was almost certainly "borrowed" from understanding the causal logic's behind the flour milling and other production line based industries whose "principles" were documented in contemporary magazines." pp. 6
"Hounshell points out that Colt did not gather directly the details of the mass production systems that had been developed to a fine art by the Federal armories at Harper's Ferry. Rather, inspired by their results, Colt set about designing a factory that use the spirit or principles of mass-production he had seen documented in contemporary accounts. Essentially, Colt recreated the instrumental logic for the American System without the necessity of observing the template. The Colt and Takaato experiences suggest that complex knowledge can be replicated without templates when the principles are evident and the copyist has good background knowledge and strong motivation." pp. 6
With these passages we can clearly see that the principles method of replication has been a key component of the success that business has achieved in the past 100 years. The traditional competitive forces creates the need to have some changes introduced to remain competitive. Although it may be argued that in today's market the competitive forces are diminished and the need to achieve an optimum operating focus is now driven through the lack of resources, and excessive demand, the overall objective remains the same. This objective of operating efficiency is the key motivation behind the replication across the energy industry.

What is replication?
"On the face of it, there does not seem to be much doubt that replication happens. The airport concourses and shopping malls of the great cities of the world provide ample (some would say depressing) testimony to the extent of replication activity in the arena of retailing and we know that it also occurs in other sectors. But the fact that the phenomenon is familiar belies the considerable challenge involved in defining it precisely. Indeed, the great philosopher Karl Popper warns us that defining replication requires judgement not absolutes." pp. 6
When we consider a software application that has the capability to replicate the knowledge of the users in most situations. We begin to rely on the knowledge of the user and an intuitive interface to enable the user to achieve what their objectives are. The software interface becomes a key component of achieving the work of the user and the method that the users attention is directed. Competing for the attention of the user is something that we have discussed in this blog before and is aggregated on the label attention economy. We have also discussed the division of labor and how Adam Smith was able to show that re-organizing around new divisions of labor enables economic growth. I have asserted the further division of labor will need to occur in order for the energy industry to expand its capacity. How the work is coordinated between those people and processes is the job of the the software application that I am discussing here. The interface is the manner in which the tasks and information are presented to the user in order for them to apply (replicate) what they know and understand.
"Replication of practices and routines cannot occur in an absolutely strict sense since the people in the organization change (whether on account of time or space) and the environment surrounding the organization is never entirely constant." pp. 7
"More precisely, we are concerned centrally with re-using knowledge of ways of doing things, i.e., it is essentially a matter of replication of organizational routines. Routines that respond effectively to differences in environment circumstance will produce different observable manifestations in different environment, even when replicated precisely." pp. 7
"Our conclusion on these puzzling question is that both process and outcome must matter in a fruitful definition of success in replication. Replication is successful when broadly equivalent outcomes are realized by similar means. On the outcome side, a positive return on the specific investment in replication sets a bare-minimum standard for "equivalence"." pp. 8
"Replication is about leveraging knowledge and is successful when "broadly equivalent" outcomes are realized by "similar means". In a specific context the words "broadly equivalent" acquire relatively precise meaning that are dependent on the replication intent. Likewise the works "by similar means" have more precise meanings that depend on the knowledge that is being replicated." pp. 8
I am going to boldly suggest that this is the primary reason that Users have been found to be critical to the success in the development of software. What has been described to this point is a massively difficult process to comprehend. The comprehension and implementation have to be defined and supported by the users. The iterative nature of the development triggers what the users want in the final product, and in some cases don't see what it is they are exactly after until the process is complete. To state that this process is ever complete is a bold assertion that fails on the belief and understanding that there is always a better way when dealing with the systemic innovation that is being set out to be achieved in this blog and in this software development project.
"We can learn much from the philosophers of science that have struggled with these questions. In psychology, according to Friedman, replication is paradoxical and difficult to define in absolute terms but none-the-less very clear in practice (Friedman, 1967: 149) In physics, Collins compares different kinds of replication including expert systems and comes to precisely the same conclusions." pp. 9

"So it is appropriate to recognize that in the business environment as in science, replication tests are matters of pragmatic truth, in which the understanding achieved in specific contexts carries a great deal of weight." pp. 9
What is software, Principle / Template. Professor Winter indicated that the two methods discussed in this paper were polar extremes. And it is with these two extremes we see the means at which we can replicate the knowledge of the industry to the user of the software that will be built here. The Principle method lets those that are experienced and understanding of the process to immediately run with it. Whereas the template can also help those that are less experienced in an oil and gas be productive and contribute as they learn. A hybrid approach is probably to a large extent necessary as we are dealing with something form a template point of view that is easy to implement and reduce to the lowest common denominator. But ideally that is not what we are after. The objective needs to be one that is supportive of the people who are innovating and expanding the knowledge and productive capacity of their firms and JOC's. In other words the user must be involved.

The point of the user is something that is lost on the SAP's and Oracle's. Their ability to deal with the energy industries needs have been atrocious. Their solutions are a point of view of understanding what best practice is, and implementing that concept within the software's interface. This is why the user is generally reduced to a cog in the wheel of the SAP / Oracle system. They are unable to represent what SAP defines as best practice to the activities of the firm. The JOC is not recognized or supported in these other applications and the ability therefore for the user to operate the system is constrained immediately. On these points Winter et al are clear, and I will leave it to them to explain their point of view in these following references.

How to Replicate
"What are the components of knowledge embedded in organizational processes? How are these components constructed and how do they get replicated? All methods seem to involve three key components: templates, principle's and background knowledge. Templates are working examples of the practices to be learned and principles are higher order causal understanding and rules. Background knowledge is what the recipient has to have to receive the knowledge." pp. 9
Templates
"While much of the knowledge in the template may be captured in codified form, in schematics, blueprints or manuals, the codified versions generally fall far short of capturing "all the knowledge." The key point about a template is that, notwithstanding any deficiency of the manuals, all of the knowledge must be there - in the "working example." The problem therefore is to find where precisely the knowledge resides, and to capture it for further use." pp. 9
"The use of templates typically requires the recipient to repeatedly observe the template in action. Repeated observation of the template in action is well known to assist in passing on craft skills." pp. 9
"Using templates often requires that the recipient organization borrows personnel from the donor to supervise the construction of the plant and oversee the start-up phases." pp. 9
What I am proposing in having the user involved in this process is fundamentally new, and something that has been proven in other solutions, (Google) that improve iteratively. If we sit down to approach the development of software in this manner. Will this be enough of a change in the process that we will be able to set out to attain what it is we are looking for? With success being the objective? Has software been developed by the user with industry wide replication explicitly stated? Not in oil and gas.
"In some cases, the template is a historical datum, originally created for its value as a business unit without regard to the possibility of replication." pp. 10
"The organization typically spends much time and energy when it engages in template construction, template refinement and codification of practice." pp. 10
"In sum: to use templates means that, one way or another an effective working example must exist, be observable and be actively used in the replication process." pp.10
"An obvious place where principles work better occurs when each potential recipients context differs so much that templates fail to capture the relevant information in a cost effective manner." pp.10
Principles.

This last argument of Winters is the ideal situation. "when each potential recipients context differs so much that templates fail to capture the relevant information." This is what Principles set out to achieve. Is this possible with software? With the level of education that is generally experienced in the energy industry, Bachelors being common and Masters in most disciplines, the users that this project appeals to are well educated. It is time to ensure that this talent is captured and codified within this software.
"Because the conveying of knowledge by principles is central to the process of teaching in universities, defining principles for an academic audience hardly seems necessary; we use them all the time. But further explication may be helpful for the context of organizational routines." pp. 10
"Success in conveying principles often depends, therefore, on supplementing them with more concrete examples, models, hints and sketches." pp. 10
"Armed with solid understanding of principles, the recipient organization can often find its own way to successful implementation. (This is also the premise of much academic instruction in the "principles" of this or that.) Of course, the freedom entails a risk that the implementation will be seriously deficient, as is well illustrated by what happened in many attempts to implement "quality management" principles (Zbaracki, 1998) and in the academic setting is too often revealed at examination time. Also, reliance on the principles of any particular routine in isolation risks the missing of the hazards and opportunities arising from interactions among the routines." pp. 10
"Greek and Roman armies used the theorem of two mean proportionals to build military machines of appropriate (large) scale on site, without references to individual templates and that these principles were more robust and more useful than templates had been." pp. 11
Both principles and templates require the recipient's of the knowledge to have skill and understanding, that is background knowledge. pp. 11
One of the areas that we easily forget is, to use the technology that is available. Sometimes what can be said in a variety of text, voice, image and video is worth so much more then we realize. Companies such as Sun Microsystem post a lot of customer information on YouTube. Although my competitors are not offering green screens to the user, they certainly have not been able to implement or take advantage of the rich media that is available today.

Finally Winter's et al have some things to say regarding management and the commitment to replication.

It is not always enough for top management to display commitment. It is often important also that people in the organization have reason to believe the claim that what is being proposed can actually be done and will have the intended effects. We see parallels here with the work of Garud and Nayyar (1994), who noted that many research-oriented firms had large stocks of dormant knowledge that had been kept "alive" and that a firm could access this knowledge and exploit it perhaps in new uses not originally conceived. Such dormant knowledge has a number of features that tend to reduce barriers to transfer.
Like most situations involving organization change, the replication context is shaped by considerations of resource availability on the one side and performance pressures on the other.
It may well be that the context of performance pressure had something to do both with the choice of principles over templates and with the favorable results of that choice. At the top management level, it produced a demand for prompt action and a willingness to accept risks of failure. Down through the organization, and at the working level in particular, acceptance of the burdens of substantial change may have been encouraged by the perception of a gathering threat to organizational viability, and hence to future employment prospects.
It is not that quality oriented routines and hierarchy necessarily stop innovation; it was that the particular application of these routines and hierarchy in these organizations did so. The conjunction of too many unnecessary levels with many stifling routines served to block rather than facilitate learning.
Thus replication by principles does not necessarily require the presence of pre-existing learning skills or dynamic capabilities; rather, experience with the successful use of principles can serve as an incubator for change and learning.

Conclusions.

As I indicated earlier, the authors were surprisingly clear in their conclusions. Recall also that these were large organizations that had deployed these changes in upwards of 1,000 employee groups. In the case of the oil distributor they were dealing with unionized truck drivers and clerical staff. Changing the makeup of positions that were previously handled by "other" groups. Conceptually introducing the new methods and replicating them through the organization would have been difficult. However the authors seem unconstrained in their support for the "Principles" method of replication.
"They used principles instead; we found that this achieved impressive results, as measured by tests of quality, speed and costs." pp. 28
"This further suggests that, however, that imitation (by principles) may be a much more significant as a threat to an innovative "first mover" than it is to the firm that has emerged as the winner in an extended competitive contest." pp. 29

Technorati Tags: , , , ,

Photo Courtesy David Sifry