Monday, August 16, 2010

What is Lean Management?

If the word "excellence" is to be applicable in the future, it requires wholesale redefinition. Perhaps: "Excellent firms don't believe in excellence--only in constant improvement and constant change." That is, excellent firms of tomorrow will cherish impermanence --and thrive on chaos. --Tom Peters

An ongoing discussion theme on the Value-Networks list at Google Groups has been the intersection of “formal process and informal networks.” A new posting on the website provides a graphic output of some of their latest Value Network Analysis (VNA) thinking, along with the claim:
“Value Network Insights™ fills the important white space in business performance – aligning the power of human interactions with processes to effectively manage complex business activity. Our more human-centric, role-based network view delivers the needed analytics. It brings a sharper business focus to collaboration platforms and reveals hidden network patterns that drive business performance.”
I first encountered this graphic in a posting which included the far more succinct and provocative question “Where does Lean Management fit in?”

To an IT Manager or a Project Manager, the diagram makes a lot of sense. Clearly someone has applied VNA to several companies that are heavily engaged in an IT effort for a complex supply chain. Details are not available, yet the graphic would be useful for insurance, e-Business (external focus) or even ERP (in a multinational).

Replace “Supply Chain” with your business and “BPM/BI” with your technology stack and it remains coherent, for it requires a tight focus on the internal activities of the company in order to effectively use the IT systems. And to run them as well. A tight focus on your external customers is by definition the nature of all business.
To begin to find out why this is such an interesting picture, we need to break Lean Management down a bit. To borrow an idea from top chefs, we need to deconstruct the conventional IT management recipe.

Set Boundaries (Plan the menu)

In a Cynefin based map the process and the product will reside in different quadrants. The Lean Manager rightly considers both to be equally important, and getting both the right product and the right process is an essential measure of quality for the management service. This is the practice of management and the praxis of better management.

The approach I am taking is grounded in the latest thinking in Knowledge Management as described by Dave Snowden at Cognitive Edge. The Cynefin Framework is the critical tool; it will be used to supplement the Lean Management graphic. VNA and Cynefin complement each other -- VNA is a discovery tool and Cynefin is used for organizing and classification in these assessments.

Knowledge Map (Assemble Ingredients)

The ingredients of Lean Management can be found as we examine the question “What do we manage?” Within the boundaries we have set the answers are startlingly easy:

• The supply chain is managed by contracts. All contracts are quite clear on deliverables. With rare exceptions all IT shops have service-level agreements (SLA); these are measured contract deliverables.
• All IT systems are managed by requirements. Traditional and agile development methodologies each deal with collections of requirements, even though they intentionally use different terminologies to distinguish one methodology from another.

Collaboration is managed continuously as we manage people. We manage through employment agreements and performance appraisals. VNA analyzes Roles as a first-class object.

• Last and most important is the Social Network. It is fundamentally unmanageable!

The knowledge map at this stage can be used to supplement the VNA data collection effort. A knowledge audit using the techniques from the Knowledge Management experts at Straights Knowledge/Green Chameleon could be used. Their video series is great. I also recommend consideration of the anecdote circle as practiced by

Value Map (Mise-en-place)

A remarkable strength of the VNA approach is how it articulates the tangible and intangible values of the extended organization.

Collectively, Roles, Deliverables, Transactions, Exchanges, Tangibles and Intangibles are first class objects of a VNA. To use our cooking analogy, these are the characteristics of the original dish which will be retained in the new creation.

As will be shown in a moment, Roles and Deliverables are firmly anchored in Cynefin based knowledge maps. Notably they are anchored in opposing quadrants separated by Disorder. This demonstrates that these two attributes are truly orthogonal. This characteristic can be used as a integrity test of your efforts; if you find these two in adjacent quadrants then you have not found the right level of analysis. Most likely your boundary conditions are not strong enough or your VNA has been forced to a premature and disordered conclusion.

The Cynefin framework is also an important tool to avoid over-analyzing your data. Especially with regard to social networks, I offer a caution that attempts to investigate the internal or external social network using these techniques for Lean Management will be doomed to failure. An excellent resource on the many reasons why social networks are unmanageable is The Relationship

When using Cynefin we need to remember the Patterns as well as the labels. Simple and complicated form the Ordered set where knowledge can be discerned. Complex and chaos form an Unordered set where knowledge emerges retrospectively. Probe and act are not interchangeable, and in the chaos quadrant I dare say that sensemaking is found from reactions to your desired action; a pattern of Act-Respond-Sense.

The Cynefin framework has been rigorously validated -- you can’t change from act-sense-respond to sense-analyze-respond and discover reliable knowledge from the results because, in doing so, you have changed the boundaries of the social system.

Knowledge Re-Map (A cooking method)

We are finally ready to begin the cooking, or in our case, to actually classify the Value Network Analysis methodology and analysis that we have started.

The first step is to transform our Lean Management graphic into the prescribed Cynefin orientation. (Trust me on this; because there is so much to remember, you will be hopelessly confused if you don’t place chaos in its proper place).

We can test this transformation by applying the highly refined definition of Collaboration that Anecdote has discovered. Can we see the ascending scale of coordination, cooperation and collaboration in the work we are studying?

Our classification actions occur as we place the VNA first-class objects into this map. For illustration purposes we’ll start with a blank frame. Remember that we are looking at the completed VNA study with concepts that have meaning in your business or social network as your context. For example, Manage people becomes Managing Field Office X.

Because people are managed (whether we enjoy it or not) and because the people who care about whatever is the substance of an intangible will act on it’s behalf, we can reasonably say that intangibles are the product of probe-sense-respond (complex) as opposed to act-sense-respond (chaos). On the other hand, many exchanges occur without reciprocation or expectation. Thus a set of many named exchanges is clearly chaotic.

The same logic applies for tangibles, transactions and deliverables. An occasional surprise on this set isn’t worrisome nor an indication of “mistakes.” A sudden shock may indicate the serendipitous discovery of something interesting, a “black swan” event if you will.

Having built out your VNA analysis to this stage, you can see quite clearly that in using only first class objects we have an unbalanced knowledge ontology. What we have done is to classify the Ordered set of your business objects. These items are what the collected group of participants have agreed to. It is the common knowledge that they share!   

To fill the gaps we need to find a way to go back through the collected mass of material. Because we will now be working predominantly in the Unordered set, I can’t suggest a particular method that should be used. By definition, what makes sense to your group will emerge from the discussions you hold.

I suspect that the conventional alternative to the team discussion, either locking a couple of key people in a room, or leaving this step to your internal/external consultant, will produce results that are confused, incomplete, or judged inacceptable. Comparing the conventional and VNA approaches provides an example of how fragmented knowledge -- fail-safe and safe-fail -- emerges out of chaos.

It will only take a few items to bind the gaps. As a metaphor for bind, think of baling wire and string. We are simply preparing a package and not building a bridge. Adding too many items is a sign of clutter and a poor choice for the level of analysis.

For our IT Management case I will introduce some concepts from a career in a very large Government organization.

Across the bottom we will find concepts that are simple in principle and potentially complex in action:

• Everyone in the private sector is aware of their companies P/L Statement. In Government, the Program Assessment Rating Tool (PART) is a reasonable facsimile. Our ‘awareness’ is simple knowledge readily obtained . Conversely our ability to affect the “bottom line” is far more complex in a behavioral sense. The direction of the arrow signifies the relationship.

• Audits clearly drive work behaviors. While the change in behavior may be complex, in actuality it is the unintended consequences that are most often significant. This is the definition of unordered knowledge, in retrospect the ‘Unintended’ aspect is simple.

Our analysis here is quite different than the methods used for determining cause and effect in Lean organizations. In this enhanced VNA methodology we are working with an Unordered set of intangible objects. In Lean, techniques such as 5 Whys and Fishbone diagrams are used to investigate Ordered sets of tangible objects, the work on the factory or office floor.

At this stage of our analysis we can safely entertain questions of leadership. We can examine the question of “fair outputs or fair outcomes” that was excluded in the boundary setting stage. It is safe because we now have at hand an adequate collection of data that describes the true environment, rather than the privileged, preferred or artificial explanations of the status quo that are espoused.

Across the top we must place concepts that are more complex. We are now binding with string whereas the bottom is bound by baling wire. As a test, concepts along the bottom can quickly generate consensus. The concepts on the top will engender arguments, votes and minority reports.

IT departments are under constant pressure to manage many things. They are tasked with keeping the lights on, fixing bugs, upgrading the infrastructure and being prepared. They also provide new features, support new customers, assess and implement new technologies, and aid in the development of brand new lines of business. Add in reuse and repurposing (one is copying, the other adapting) and one can see how complex managing an IT department is.

IT shops I’ve known tend to organize their staff according to these functions - Development lead by Project Managers, an Operations unit (or units) and the Help Desk or Call Center. Top management, the Chief Information Officer, is most likely situated in the Chaotic frame. The area of significant conflict is across the top, where it boils down to the most difficult CIO decision (after staffing of course): “develop something new or keep plugging with the old.” The new disciplines of Enterprise Architecture and Systems Operations are simply the post-modern form of build and manage.

What has changed in the last decade is the complexity of the infrastructure. To draw an analogy, we are no longer managing a boxing match as was the case in the mainframe era, nor the football team from the PC and LAN age. It now takes small armies to support the extended infrastructure across the web. 

Gartner predicts a hyperconnected and spontaneous workplace in the next decade as the new supply chain (Gartner Says the World of Work Will Witness 10 Changes During the Next 10 Years). If half of what is predicted is realized, I am afraid that “Cloud Computing” will require we grow an entire marketplace within the castle walls.

It will be quite informative to observe the efforts and activities of the US Government (and other governments) in these three areas.

Presentation (Service a la ruse)

I can now reveal the big picture and explain the semantics of our arrows:

The arrows represent channels. The paired set of arrows is a manifestation of the complex adaptive system or social system that spans the two quadrants. Directionality indicates the bias or scale, and to be more precise, that the simple named concept has a complicated, complex or chaotic impact.
There is no fixed relationship between the two channels: they may be independent, amplifying or conflicting. They will co-evolve because the labels are descriptive of a metastable state. To dig any deeper will break the gossamer web and doom your attempt to disorder.

We can finally test the original claim that “Value Network Insights™ fills the important white space in business performance.” My test is the following hypothesis for knowledge management as theory and practice :
• With knowledge in hand, both process and structure are mutable. Without knowledge, we are left with the rituals of disorder.

• Using Cynefin we are constantly reminded of the rituals we may not see. Value Network Analysis spans that void, giving us in turn a tool that can indeed fill-in the white space on our cognitive map
I can’t say I have found Lean Management yet. What I have found is a recipe I can use to build Lean Management into my organization. I encourage you to give this recipe a try and please let me know where your journey takes you.

No comments:

Post a Comment