Lean Improvements and Data

I write a weekly RELENLESS HANSEI topic on my desk whiteboard and was looking through Liker’s The Toyota Way for a possible topic.  On page 261 I came across a powerful message:

“We discovered the top management in the companies with vital programs had process orientation, while the unsuccessful companies had results-oriented managers.  The results-orientated managers immediately wanted to measure bottom-line results of the continuous improvement program.  The process-oriented managers were more patient, believing that an investment in the people and the process would lead to the results they desired.”

This passage came from the section on the role of metrics.  This made me reflect on the harmonies of organizational culture.  If you speak about investment in people or lean improvements but measure hard results, the potential to be out of sync is high. 

When organizations are results-orientated, the potential for bad data exists also.  I have seen where correlation and causality get confused to be able to report results.  Results-oriented metrics can get murky for medical publications as well.

What are your thoughts?

Subscribe to Improve With Me via: RSS | Google Reader

Follow me on Twitter

Advertisement

My Reflections From Tufte

I learned a lot when I saw Edward Tufte last week.  I have increased ^ my knowledge of data presentation (the little carrot in this sentence is an example of mixing symbols directly into narrative).   Here are some more take-aways:

• We get so much into our graphs, it is easy to lose the power of a simple table (people read them all of the time in sports and market papers – why do we think people need fancy picture-graphs).
• Sort data by performance.  We often see data sorted by alphabetical order which doesn’t allow people to see trends.
• Make content rich and give a data dump to allow the viewer to make their own conclusions.  Viewers will find a point that hooks them, we do not need to direct them (unless you are in sales). 
• Rich content = More than 1:1 comparisons.  For instance, instead of just showing process cycle times – add data about time available, visit type, patient acuity, etc.  Separate tables and graphs do not allow the viewer to see how every element is connected.  Multivariate data should coexist and not be separate!

I was frustrted with one thing from his course.  The majority of the graphs he loves need a drawing tool and not possible with Excel.  Some of his advice boils down to “nice to know but I don’t have the tools to apply it yet”.

The course was worth the time and money and I encourage anybody to see him.  You get all of his books which pretty much make up for the price.

Subscribe to Pass The Buck via: RSS | Google Reader

Last Week’s Event

The team did an outstanding job last week.  A couple of their targets were to increase on-time starts for lab draws in the cancer clinic and create visual systems to see demand.  They met their targets and really caught the spirit of improvement.

One factor that made the workshop a success was having a patient’s parent involved.  The parent was able to come in and explain their experience and perspective.  Their participation created a lot of value for the team.

This event demonstrated the power of going to gemba.  A team-member commented how it always felt chaotic but could never pinpoint where the chaos was coming from until they observed and broke it up into pieces.

Last but now least, I saw how important data is for an event.  I am a data consultant so this was impactful to me.  Data in other events seemed to stop after establishing the current state of a process.  This team was looking at cycle times and percentage scheduled during time of day all week.  This helped them figure out specifically where to make improvement adjustments.

I had a great time and learned a ton!

Subscribe to Pass The Buck via: RSS | Google Reader