Getting It Right the Second Time


By Gabriel Szulanski and Sidney Winter  

Harvard Business Review, January, 2002, p.62  

Copy exactly!" is Intel's catchy phraseology for the concept this article explains.  Another, more generic, description is "copy closely."  

This approach is an answer to the challenge of replicating successful but complex group systems within an organization.  Such systems can range from bank branches to Starbucks Enterprises to manufacturing processes.  

The authors noted that, in general, attempts to replicate successful approaches fail.  Usually, failure can be attributed to a combination of one or more of these typical mistakes:

1.   trusting in the "expert" at the successful operation to explain how it was developed and why it works

2.   only consulting documentation on the successful operation for information as to its components and functioning

3.   trying to "improve" the process before it has been successfully replicated  

While both asking the "expert" and consulting documentation can be used to gain critical knowledge, each suffers from substantial shortcomings.  In the first place, the expert, as a part of the system, generally lacks a full understanding of the key components of its success.  Some knowledge may be held only by workers within the system, and not shared with higher-ups, possibly because it makes the worker's job easier, or runs against stated policies. Some may be tacit -- learned on the job and well understood, but almost impossible to convey in any helpful way.  For example, workers may know the particular sound of fluid moving through pipes that signals a process is working well, but may be either unaware of the role this knowledge plays in daily operations or be unable to communicate it.  Finally, hidden contextual factors may contribute to successful process.  These can be anything from prevailing weather patterns to the design of particular machines.  

After considering these hurdles to understanding why a particular process is successful, it is easier to understand how attempting to "improve" a process as it is copied is likely to end up resulting in inferior performance.  For example, if the design of a particular machine is important to a system, and a “new and improved" model is substituted as the system is copied, then the system may well fail to achieve the level of success of the original, and the individuals participating in the copy, not having experience with all of the intricacies of operation, will be unable to pinpoint the reason for the discrepancy.  

Best Approach  

So, how do you copy a successful system?  

First, make sure you have something that can be copied and that is worth copying.  For example, success attributable to a great manager or to a great team that has worked together well for years cannot be copied.  On the other hand, those involved in a successful operation may attribute its success to their unique gifts, when much of the success is really due to the system.  Also, make sure the system's reputation is warranted.  Look for important systems with proven track records that can provide a detailed example of how to achieve results you would find acceptable.  

Second, have a working model to copy.  Such a model provides proof that success is possible, performance measurements to define that success, a tactical approach, and a reference point if the copy falls short.  Having a single model is important also; trying to combine best elements from various models falls into the "improve as we copy" fallacy.  The authors provide the example of efforts by Xerox in Europe several years ago to copy best sales practices.  The initial effort went very well as they identified particular countries as the "benchmark country business units" for particular practices.  For example, to learn to sell a particular type of color copier, organizations went to Switzerland .  To learn to how to retain major accounts, they went to Spain or Portugal .  This effort was very successful.  

However, in the second phase, the units identified as the best models for use of computer-assisted sales management were in countries which could not easily serve as models due to prevailing biases within the organization.  The practices were therefore extracted and presented as a combination of best practices from different countries.  Implementation failed because managers did not have a live, working model to observe and copy.  

Finally, copy as closely as you can.  Even if you think the system will need to be changed, and even if you think you already know how it should be changed, copy closely first. Live with the system and gain the knowledge and understanding that comes from direct experience, and allow your team members to do the same.  Then work together for improvement.  As one manager of Xerox said, “We lost a lot of best practices because people edited them before they were implemented.”  

Barriers to Success:  

Uncooperative Sources:  Those with the model don’t want to share  

Strained Personal Relationships:  copier and model managers dislike each other  

Internal Competition:  model and copier are engaged in internal competition  

Overemphasis on Innovation:  a culture of idealization of innovation and scoffing at copying  

Cranky Copiers:  managers of copying unit are resistant to new knowledge, afraid to change, or focused on preserving their status.  

Contextual Differences:  if the cultural or other contextual factors are sufficiently different, close copying may not work.  The article discusses copying across country lines as an example.  Of course, it’s always easy to dismiss superior performance with a, “Well we couldn’t do that because ….”  

Application to Schools and School Systems:  

Some aspects of successful schools seem difficult to copy. For example, the successful principal with his or her team of teachers that have been carefully selected and worked together for a number of years may be impossible to copy by a school whose principal has different strengths, whose faculty may not have been selected by the same criteria, and whose experience over time may not have created the same relationships and shared belief systems.  And the aspect copied may not be the basis for success.  For example, Barclay School in Baltimore made great gains after adopting the Calvert curriculum.  But how much of that was curriculum, and how much cultural changes in Barclay that may have resulted from the battle to be allowed to adopt the more challenging curriculum?  

On the other hand, this process might well have some applications:  

·        What if new teachers attempted to copy as closely as possible the most successful model of quality teaching in their school?

·        Some central office systems:  accounting functions, construction management possibly, maybe human resources, information systems, etc., should be candidates for close copying between systems.

As usual, studying the best thinking on business management techniques provides interesting and helpful insights, and raises significant and useful questions, but does not give easy answers.  Of course, if it were easy, it would have already been done!  (And we could copy it!)

July, 2003