Subscribe to the FREE Business Rules Journal Newletter


 

 

 

     ZACHMAN ARCHIVES ...
untitled

Life Is a Series of Trade-Offs and Change Is Accelerating!

by John A. Zachman

This column originally appeared in the Jan./Feb. and Mar./Apr. 1999 issues of the DataToKnowledge Newsletter.

The older I get, the more I am aware that life is a series of trade-offs and as much as we would like to "have our cake and eat it too," that is a figment of our imagination.  It is simply not realistic.  As soon as we start thinking we can have it both ways we are positioning ourselves for the inevitable 'rude awakening'!  We would be far better served to clearly understand the choices we are making, set our expectations correctly, and mitigate the downstream impacts of the choices we make.

Life's trade-off in its most simplistic form is between the short term ('immediate gratification') and the long term (what's 'good for us').  The way this trade-off reads out in the context of Enterprise Architecture is:

short-term options

vs. long-term options

implementation

vs. integration

point-in-time solutions

vs. infrastructure solutions

expense-based approaches

vs. asset-based approaches

implementation optimization
(at the expense of the Enterprise)

vs. Enterprise optimization
(at the expense of the implementation)

pay me now or pay me later

vs. takes too long and costs too much

etc., etc.

Some times, and for some aspects of the Enterprise, you may want to take the short-term option, and other times, for other aspects, you may want the long.  When you want the short, you want to mitigate the downstream impacts of the suboptimal approach when you pick the long, you want to find clever ways to satisfy short-term demand without compromising enterprise integration.

The Framework for Enterprise Architecture (the 'Zachman Framework') is a useful analytical tool to help you think about the trade-off, correctly set your expectations, and devise strategies to mitigate the effects of your short- or long-term choices.

The Framework is a classification scheme for descriptive representations of any complex object in our case, an Enterprise.  It identifies a total, comprehensive set of descriptive representations (that is, 'models') that are relevant for describing the Enterprise.

Some people seem to think that the Framework prescribes that all the models have to be made explicit before any implementations can take place.  That could not be farther from the truth!!  If that is your impression, you need to return to the basic concepts of the Framework.  It is only your (short-term or long-term) methodological choices that determine which cells (or 'slivers' of cells) are produced over the process of implementing systems, automated or manual.  The Framework is totally neutral relative to methodologies.  In fact, ANY methodology and EVERY methodology whatever actual work you are doing can be mapped against the Framework to determine what your choices and underlying assumptions are, simply by identifying which cells (or 'slivers' of cells) you are producing and managing, in what sequence.

This misperception about the Framework being 'all or nothing' probably stems from the fact that the Framework does, in fact, identify the total, comprehensive set of relevant models for the Enterprise, even though it presupposes nothing about which models (or 'slivers' of models) you take the time and spend the money to make explicit.

Actually, if the Enterprise exists, all of the models already, by definition, exist.  If you do not take the time and spend the money to make any one or more of them explicit does not mean that the models go away.  It just means you didn't make them explicit.  They are implicit.  That is, you are making assumptions about them.  In fact, if no models have been made explicit, someone is making or has made assumptions about every one of them, and the assumptions are embedded in the reality of the operating Enterprise, that is, in the glob of spaghetti, the implemented 'systems' -- automated and/or manual -- and it's very difficult for anyone (including the programmer, or methods analyst, or Industrial Engineer or whomever made the assumptions) to remember or to figure out what assumptions have been made, or whether they are any longer relevant, or whether they are coherent, or consistent, or integrated across the scope of the Enterprise, or integrated horizontally (with other types of models), or integrated vertically (with other models of the same type), and so on.  In fact, it is especially difficult to figure out what assumptions have been made months or years after the fact.

It is obvious that in a changing environment, the ultimate, long-term best interests of the Enterprise would be served if "all of the models were made explicit, Enterprise-wide, horizontally and vertically integrated, at excruciating levels of detail."  In this case, anyone who was authorized could tell immediately what assumptions have been made, as the assumptions would be explicitly described in the models.  It would be very easy to evaluate the assumptions in the light of changes taking place in the marketplace or the environment external to the Enterprise and to determine what very specific components internal to the Enterprise have to be changed.  You could know precisely what has to be changed and what cannot be changed, what other of the models would be impacted by the change, how much it would cost to change, how fast it could be changed, etc., etc. If the external environment starts changing more and more rapidly, it will become more and more imperative to have the models already made explicit so you can very quickly assess the impact of the changes and decide what parts of what models have to be changed and then actually effect the change.

This brings me to another criticism I have recently heard, that the 'Zachman Framework' "is rooted in the slow change environment of the 80's.  It has little value to companies with high change environments."  Obviously that could not be farther from the truth either!!  If this is your perception, you need to do some research into the fundamentals of change management.  If you are going to manage change, it simply means that you are going to be driven to take long (or longer) term approaches to Enterprise Architecture.  You are going to have to build (that is, make explicit) the models or 'slivers' of models that you are going to want to change.

For example, let's assume the worst for a moment that you are in a high change environment and that you do not have any explicit models, that you have been optimizing the implementations and not building or maintaining any models for the last 50 years, and that all the assumptions about all the implicit models of your Enterprise are embedded in the implemented systems (automated and/or manual) and that suddenly you have to change.

Basically, you have two options.

First, you could build new systems.  In this case you would then have to try to figure out how to turn off the old systems when you get the new ones built or else ... at worst, have conflicting implemented systems, or at best, redundant systems.  And, without any models, it would likely be rather challenging to figure out what systems to turn off and how to turn them off.  Also, the smaller the new systems you build and the faster you can implement them, the more discontinuity and redundancy you are likely to build into the overall set of Enterprise systems ... because you wouldn't know how the new systems relate to any of the existing systems since no models exist that describe the old ones.

Clearly, this is the approach we have been using for the last 50 years.  It is the legacy creation approach, which I doubt that I have to explain, has some severe shortcomings, particularly when it comes to change!!!

Your second option would be to try to recreate the models first, reverse engineering the assumptions out of the running (manual or automated) systems up to whatever row or column or cell you want to change so that you could assess what is impacted by the change ... then reanalyze, redesign, reconstruct, retest, and rerun.  If the Enterprise has any complexity at all, and if the changes you are trying to make are extensive or interrelated, this might take a lot of time. In fact, you may not be able to stay in business long enough to reverse engineer and get them changed!

Take General Motors for example. I have never been to General Motors, but I understand GM is around 3 years into a 5-year project in which they are attempting to take something like 28 months (over half of the cycle time) out of the current product cycle massive, wrenching changes to their business!  And I suspect, there are no models and I further suspect that there is no one around anymore who knows the assumptions that are embedded in the actual operating Enterprise (manual and/or automated).  That is, probably, no one knows how the business works at least, no one knows how it all works together as a contiguous business.  I understand that currently GM is sitting at about 48 months from 'concept to customer' whereas Toyota is at about 28 months and Chrysler is down to around 18 months.  GM's market share has declined in the last 13 years from around 49% to about 32% today.  I have a feeling that the big question is, are they going to get the models reversed engineered so they can figure out how to get the business changed in time to sustain their asset base and stay in business?  If my suspicions are accurate, I could argue that it would have been a lot easier and faster and maybe cheaper and certainly safer if they already had "all those models, Enterprise-wide, horizontally and vertically integrated at excruciating levels of detail etc., etc."  I suspect that somebody is saying something like, "why in the world didn't we take longer-term approaches to our Enterprise Architecture?" or worse, "why didn't YOU take ?" (etc., etc.)

Anyone who does thoughtful research into change management will find, in all of history, older disciplines that have had to deal with change to complex objects, for example, Architecture and Construction, Manufacturing, and Engineering they all start with the descriptive representations of the object (the 'architecture' -- the drawings, functional specs, the bills of materials, etc., etc.).  No architect or contractor in their right mind would start making changes to a 100 story building (or ANY building for that matter) without first going to the 'Architect's Plans' -- the descriptive representations of the building, the architecture, the 'models.'

My friend Bernie Boar, who recently has written a book, Constructing IT Blueprints, said at a conference that we recently ran under the auspices of ZIFA, "When the Titanic hit the iceberg, the Captain did not call for the architectural principles; he called for the architectural blueprints."  At the point in time when you hit the iceberg, you have to be able to look at the architectural representations (the 'models') to understand what has happened and to decide what you can do about it.  I would observe that, when the Enterprise gets into 'extremis' (hits the iceberg) and has to change quickly, if the models are not already explicit there is going to be a real question whether you can build from scratch or whether you can reverse engineer and change what you are before you run out of time and the thing sinks!  And the faster the changes start coming, the worse the problem is going to get!

If the world was a perfect place, in a changing environment, it is clear once again that the best of all possible worlds would be to have "all of the models, Enterprise-wide, horizontally and vertically integrated at excruciating levels of detail."  Therefore, this would represent the best, long-term architectural strategy for the Enterprise.

On the other hand, it is also clear that the world is not a perfect place and you are going to be tempted to compromise the long-term strategy every time you want to implement something because it will take some time and cost some money to preserve the long-term architectural continuity.  And the pressure for immediate gratification -- because the stress levels are out-of-sight -- is enormous!

The question you are faced with for every implementation is, "do you want the long-term approach or the short-term approach?"  The problem is, "you cannot have your cake and eat it too!"  If you want quick implementations, you take the 'you start writing the code and I'll go find out what the users have in mind' approach.  If you do this exclusively for any length of time, you will find you have built something that is impossible to change -- a 'legacy.'

On the other hand, if you want to build something that will accommodate change, you will have to build and manage the models that will constitute the base-line for managing change.

I would argue that it would be very good to know (be conscious about) which approach you are choosing for every implementation and to understand what its implications are and to make sure everybody else in the Enterprise who is significantly interested is conscious about it as well, so that the expectations are correctly established.  Otherwise you are very vulnerable to a lot of hate and discontent when the stark reality of the expectation gap -- with whichever choice you make -- inevitably presents itself!!

The Framework is helpful for analyzing the options -- to understand what they are, what the implications are, and to communicate and have objective dialog about what they are.

For example, which cell or cells or 'slivers' of cells are we going to make explicit?  Are the cells (models) we make explicit going to be Enterprise-wide (integrated)?  And if not, do we care about the fact that they are not?  (Note:  the Enterprise will care whether some cells, or some 'slivers' of some cells, are 'integrated' or not.  Other cells or 'slivers' of cells they will not care whether they are 'integrated' or not.)  In the cases where we do care, what (modest amount of) additional work could we do to mitigate the impact of not producing the entire Enterprise-wide model, or would we be better served by a longer-term, more integrated approach?

Further, are there cells (or 'slivers' of cells) above the one (or ones) we are working on that will not be made explicit?  And if so, are we willing to assume the risk of making assumptions about them?  And again, are there other cells in the same row whose natural integrity we will compromise if they are not made explicit and related to the chosen cells?  And, in both of these cases, are there things we can do to mitigate the impact of making the short-term trade-off with a modest amount of additional work, or do we have to consider a longer-term approach?

If you are going to compromise the long-term strategy, you probably should have a pretty good reason for doing it, making sure the implications are exercised, expectations correctly set, and mitigating actions taken appropriately.

I have a number of friends (notably Doug Erickson of Data, Applications, Technology Associates) who declare that it is actually cheaper and faster to take Enterprise-wide, top-down architectural approaches to systems implementations because the cost reductions of removing redundancy and the cycle-time reductions of leveraging reusability vastly reduce development cost and time to market.  This is the lesson that has been painfully realized by Western manufacturing over the last 20 years -- that it is actually cheaper and faster to build products that have no defects (the long-term approach) rather than to build defects into the products and then attempt to remove the defects after the products are built (the short-term approach).  This is the 'quality is free' idea and, as the argument goes, it applies to Enterprises as well as to automobiles, microprocessors, airplanes, bicycles, machine tools, or whatever.

In summary, the Framework defines the total, comprehensive set of models relevant for describing the Enterprise and, by so doing, it defines the optimum, long-term architectural strategy.  However, the Framework is inert relative to your methodology.  It only helps you evaluate the methodology in terms of which models you will produce and how they relate to the optimum, total set.  In this fashion, you can make a reasoned judgment whether to take longer-term approaches (building and managing some models), or shorter-term approaches (just getting the system implemented), and/or to attempt to balance the two by finding a kind-of middle ground, selected-cell, 'sliver-based' compromise.

If someone proposed to me to do some implementation work in my Enterprise without using the Framework, here are the kinds of things I would ask them:

  1. What models are you going to build?  (Then, I would map those models to the Framework cells or 'slivers' of cells to see how I felt about it.)


  2. How are you going to deal with quality making sure that what you produce meets the requirements of the Enterprise?  (Then, I would see if the logic takes me straight up the column(s) from the cells being produced to row 2.)


  3. How are you going to deal with integration?  (I would look at how they intend to deal with 'scope integration,' that is, Enterprise-wide integration, as well as 'horizontal integration' -- that is, integration with other models in the same row(s) they are building.  Are they going to optimize one model at the expense of another model?  Or, are they going to create redundancy?)


  4. How are you going to deal with change?  (Are they going to build models and keep the models they build as a baseline for managing change?  Also, I would expect them to leave the models and the model of the models (the 'meta-model') with me because these are key for me to manage change when the model-builders/system-implementers are long gone.)


  5. If they are not going to build any models at all, then I would ask myself, "Do I really want these guys to help me or not?" because I happen to be convinced that change is a big problem and that longer-term approaches are imperative.


  6. Last, I would ask myself whether I am happy with the short-term or long-term choices that they are making explicitly or implicitly as a result of their methodology.

Life is a series of trade-offs between immediate gratification and what's best in the long term.  The Framework is helpful as an analytical tool for having a meaningful, objective dialog about these choices and for making these choices or for finding a right balance between the choices for the Enterprise.  At the heart of managing change is Architecture -- Enterprise Architecture.  If you are going to manage change, you are going to take longer-term approaches to Enterprise Architecture.  The Framework for Enterprise Architecture could be extremely valuable as a management tool, central to making the long-term/short-term trade-offs and managing any complex Enterprise in an environment of increasing complexity and increasing rate of change.  The Framework clearly is not 'all or nothing.'  At the same time, explicit, architectural representations (models) clearly are fundamental to managing change.  In fact, I can find no examples in other, older disciplines where change is managed in complex objects without explicit models.

Even though the Framework is totally neutral to whether you are building and managing models or not, if you find yourself in an increasingly complex and changing world, "someday, you are going to wish you had all those models, enterprise-wide, horizontally and vertically integrated at excruciating levels of detail!"  The key lies in figuring out how to find the balance between the short term versus the long term, between implementation versus integration, between point-in-time solutions versus infrastructure ... how to satisfy short-term demand and at the same time build something coherent over time.  The Framework may be your only hope for figuring that out.

Copyright (c) 1998.  Zachman International.


standard citation for this article:
John A. Zachman, " Life Is a Series of Trade-Offs and Change Is Accelerating!" DataToKnowledge Newsletter, Vol. 27, Nos. 1 and 2 (Jan./Feb., Mar./Apr. 1999), URL:  http://www.BRCommunity.com/a1999/a409.html.

January 2017
The Issue Is THE ENTERPRISE
By John A. Zachman


October 2016
Strategy Spectrum for Enterprise Engineering and Manufacturing
By John A. Zachman


July 2016
The New EA Paradigm
(4) The Assemble-to-Order Pattern

By John A. Zachman


June 2016
The New EA Paradigm
(3) The Provide-from-Stock Pattern

By John A. Zachman


May 2016
The New EA Paradigm
(2) The Make-to-Order Pattern

By John A. Zachman


April 2016
The New EA Paradigm
(1) Expenses and Assets

By John A. Zachman


March 2016
The Information Age: (3) Powershift
By John A. Zachman


February 2016
The Information Age: (2) The Third Wave
By John A. Zachman


January 2016
The Information Age: (1) Future Shock
By John A. Zachman


December 2015
Defining Enterprise Architecture: Economics and the Role of I.T.
By John A. Zachman


November 2015
Enterprise Physics 101
By John A. Zachman


September 2015
A Historical Look at Enterprise Architecture with John Zachman
By John A. Zachman


August 2015
Cloud Computing and Enterprise Architecture
By John A. Zachman


June 2015
The Zachman Framework Evolution (Part 2)
Special Guest: John P. Zachman


May 2015
The Zachman Framework Evolution (Part 1)
Special Guest: John P. Zachman


April 2015
Architecture is Architecture is Architecture
By John A. Zachman


April 2013
John Zachman's Concise Definition of The Zachman Framework
By John A. Zachman


November 2004
The Zachman Framework and Observations on Methodologies

 

November 2003

Framework Fundamentals: Frameworks, Reference Models, and Matrices

 

August 2003

Framework Fundamentals: A Dialog With John Zachman

 

June 2003

Framework Fundamentals: Miscellaneous Enterprise Engineering Concepts

 

April 2003

Framework Fundamentals: Framework Fundamentals: Level of Detail is a Function of a CELL

 

February 2003

Framework Fundamentals: Responding to Questions from the OMG

 

May 2002

Enterprise Quantum Mechanics (Part 2)

 

March 2002

Enterprise Quantum Mechanics (Part 1)

 

January 2002

"What" Versus "What"

 

November 2001

Security And The "Zachman Framework"

 

September 2001

Fatal Distractions (Part 2)

 

July 2001

Fatal Distractions (Part 1)

 

May 2001

You Can't "Cost-Justify" Architecture

 

March 2001

Conceptual, Logical, Physical: It Is Simple (Part 2 of 2)

 

January 2001

Conceptual, Logical, Physical: It Is Simple (Part 1 of 2)

 

September 2000

Building The Enterprise - An Infusion Of Honesty

 

July 2000

All the Reasons Why You Can't Do Architecture or ("We Has Met the Enemy and He Is Us")

 

May 2000

Enterprise Architecture Artifacts vs Application Development Artifacts (Part 2)

 

March 2000

Enterprise Architecture Artifacts vs Application Development Artifacts (Part 1)

 

November/December 1999 & January/February 2000

Enterprise Architecture: Issues, Ingibitors, and Incentives


July/August & September/October 1999

Packages Don't Let You Off The Hook

By John A. Zachman


January/February & March/April 1999

Life Is a Series of Trade-Offs and Change Is Accelerating!


November/December 1998

"Yes Virginia, There IS an Enterprise Architecture"


July/August 1998

Enterprise Architecture: Looking Back and Looking Ahead


January/February 1998

The Framework for Enterprise Architecture (The 'Zachman Framework') and the Search for the Owner's View of Business Rules

 

 

 about . . .

 JOHN A. ZACHMAN

 

John A. Zachman is the originator of the “Framework for Enterprise Architecture” (The Zachman Framework™) which has received broad acceptance around the world as an integrative framework, an ontology for descriptive representations for Enterprises. Mr. Zachman is not only known for this work on Enterprise Architecture, but is also known for his early contributions to IBM’s Information Strategy methodology (Business Systems Planning) as well as to their Executive team planning techniques (Intensive Planning).

Mr. Zachman retired from IBM in 1990, having served them for 26 years. He is Chief Executive Officer of his own education and consulting business, Zachman International®.

Mr. Zachman serves on the Executive Council for Information Management and Technology (ECIMT) of the United States Government Accountability Office (GAO) and on the Advisory Board of the Data Administration Management Association International (DAMA-I) from whom he was awarded the 2002 Lifetime Achievement Award. He was awarded the 2009 Enterprise Architecture Professional Lifetime Achievement Award from the Center for Advancement of the Enterprise Architecture Profession as well as the 2004 Oakland University, Applied Technology in Business (ATIB), Award for IS Excellence and Innovation.  In August 2011,  he was awarded the Gen. Colin Powell Public Sector Image Award by the Armed Services Alliance Program.

Mr. Zachman has been focusing on Enterprise Architecture since 1970 and has written extensively on the subject. He has facilitated innumerable executive team planning sessions. He travels nationally and internationally, teaching and consulting, and is a popular conference speaker, known for his motivating messages on Enterprise Architecture issues. He has spoken to many thousands of enterprise managers and information professionals on every continent.

In addition to his professional activities, Mr. Zachman serves on the Elder Council of the Church on the Way (First Foursquare Church of Van Nuys, California), the Board of Directors of Living Way Ministries, a radio and television ministry of the Church on the Way, the President’s Cabinet of the King’s College University, the Board of Directors of the Los Angeles Citywide Children’s Christian Choir, the Board of Directors of Heavenworks, an international ministry to the French-speaking world and on the Board of Directors of Native Hope International, a Los Angeles-based ministry to the Native American people.

Prior to joining IBM, Mr. Zachman served as a line officer in the United States Navy and is a retired Commander in the U. S. Naval Reserve. He chaired a panel on "Planning, Development and Maintenance Tools and Methods Integration"  for the U. S. National Institute of Standards and Technology. He holds a degree in Chemistry from Northwestern University, has taught at Tufts University, has served on the Board of Councilors for the School of Library and Information Management at the University of Southern California, as a Special Advisor to the School of Library and Information Managementat Emporia State University, on the Advisory Council to the School of Library and Information Managementat Dominican University and on the Advisory Board for the Data Resource Management Programat the University of Washington. He has been a Fellow for the College of Business Administration of the University of North Texas and currently is listed in Cambridge Who’s Who.

 

 

 

[ Home ] [ Staff ] [ About BRC Publications ] [ Editorial Feedback ] [ About BRCommunity ]
[ Contributor's Guidelines ] [ Privacy Policy ]