Standardization in knowledge work
Standard work is a common element of the Toyota Production System or Lean implementations. The general idea is to define the specifics for repeatable tasks, so that any trained operator can do the work at the same level and speed. Of course, good implementation also look for opportunities to improve the work, so "standard" is only a moment in time.
But what about work that isn't so repeatable - work that isn't putting together a car or doesn't have some observable component? Can the idea of standard work be applied in knowledge work?
The obvious reaction is No! Intellectual work cannot be encoded in the minute details of what one does from moment to moment. And I agree. It doesn't make sense to talk about standardizing the details of knowledge work.
It is the idea that applies. Rather than looking at the specifics of the tasks, look at the setup, management, and reporting of that work. Here are some things that can become standards:
- When people get ready to do this kind of work, what do they need to ensure success? Is there a policy on starting without this full-kit? (I recommend defining full-kit and the policy that prevents people from starting without it.)
- How can we see the status of ongoing work? What mechanism do we want to use to detect problems and opportunities before they get bad?
- How is the completion of the work to be communicated? (Report, Presentation, face-to-face handover) If the completion of knowledge work isn't reported in some fashion, then no one knows. Therefore it is NOT DONE.
- How does the team handle requests for help? Is that process different for help from within the team vs. outside the team?
This thinking has been percolating around my mind and those of many others. Atul Gawande's Checklist Manifesto gets at the idea of creating standards around knowledge work. And Dan Pink's A Whole New Mind takes the view that repeatable (often understood as standardized) work is easily gets outsourced or automated - and therefore standardization is the first step to job shifts. (It can also be taken as a call to standardize, so that people can do more interesting, intellectual work.) And several of the people I follow in the TOC / Lean / Continuous Improvement arena have said things about this of course: Joe Dager, Pascal Dennis, or for an interesting take on the topic applied to CEO's Dan Markovitz.
I specifically wanted to write about this today because I noticed reference to the danger of standardized work in today's post by Harold Jarche.
[Photo: "Standard Pumping" by Pete Zarria]
Previous entry: The New Social Learning
Next entry: Improving collaboration for what