We have to think of standard work in the context that it was invented.  Standard work is SPECIFICALLY there (according to Ohno) to give a basis for Kaizen.  His rules are clear:

1.        Standard work is a document of THE WAY THINGS ARE CURRENTLY DONE, not the way they should be done.

2.       Standard work is there only to enable Kaizen – not to tell people how to do their job.  You can’t do data-based improvement until you have a baseline measurement of the current way of doing things.  Period, end of discussion.  Standard work is a baseline for improvement, it is NOT a description of how to do the job, even if it appears to be.

Here is how I see standard work being abused:

1.       “We need documents so that anyone can do any job.”

Nonsense.  The idea of standard work is NOT to make people fungible.  It is to encourage people to do Kaizen. If it does not accomplish this purpose, it is being abused.

2.       “We need to use the same standard across the company.”

Why on earth?  Where ever did that idea come from?  What good are standard processes across the company?  All they do is squelch Kaizen, so they clearly are not a good thing.

Well, I can be flexible on that – standards across the company can make a good baseline for local organizations to start Kaizen efforts.  Some standards will have to span organizations.  But as a matter of course – “because everyone knows that a standard should be used across the whole company” – I don’t think that’s an adequate reason for corporate standards. I had a saying in my department when we were doing company-wide plant information systems:  “Find out what people in the plants don’t care about, and make that a standard. It will make their life easier. Find out what they do care about, and let them make their own choices on those things.”

I’m beginning to think that the idea of company-wide standards appears when you have monolithic organizational structures.  3M, Gore, Semler, and – I now discover – the Pennsylvania Railroad, all had/have small independent business units rather than a single monolithic business unit. The Pennsylvania Railroad was universally recognized as the best run railroad of the 1800’s – by far.  It had 3 separate divisions, each with less than 2000 miles of track to manage.   The person who designed this ‘decentralized’ structure for the Pennsylvania railroad theorized that it might have higher management costs, but it would result in better management decisions because managers were closer to and more involved in the business.  He was absolutely correct.

In such ‘decentralized’ organizations, there is no concept of standard processes across the company except in a very few centralized areas.  In 3M, the centralized areas were finance, personnel, purchasing, and capital equipment and facilities construction.  Everything else was local. Although processes were more or less similar across the company (good practices spread because people moved around), there was not much interest in standard processes across the company.  Certainly there was not a standard process (while I was there) in areas of core competence such as product commercialization.  And yet, all divisions had more or less similar product commercialization processes, with differences stemming from the type of business or product.

In monolithic/centralized organizations, I often see standard processes used as a substitute for good management decisions, because managers are too far from the work/customers.  I’m not sure there is a fix for this if decentralization is not an option.  But when used in that way, standard work has exactly the OPPOSITE effect from what Ohno – for example – intended.