"Without a doubt, your book [Data Sanity] is the most influential, insightful and educational book I have
ever read...My copy is full of dog-eared pages, highlights and underlines...You have forever changed the way I approach quality, teaching and leadership. Thank you, Davis. You gave me answers for which I had been searching for some time." – written evaluation after a plenary presentation
[Take 4 to 6 minutes to read over a break or lunch]
"Off to the Milky Way!" with all kinds of road maps all arriving at the same destination: Confusion, Conflict, Complexity, and Chaos
me recap the cancellation / no show data scenario of my last newsletter
, then take it a little further.
You're at a typical year-end review of performances versus goals. The 12 months' data are presented as a table including this year's average, last year's average, goals, and definitions of the of the red, yellow, green
In this case, there was additional concern about the year's average of 10.0 versus the previous year's average of 9.4 – Why was there a 0.6% increase?
Of course, the meeting proceeds to the usual additional time-wasting nonsense of leaders suggesting what they believe to be the "low-hanging fruit," then holding people personally accountable to another probably even "tougher" goal for the
What is the objective of these performance review meetings, then? Is the goal to meet goals or help people improve their processes (appropriately) to meet the goals?
The fear I've sensed and seen in these meetings pretty much evidences the former.
Such nonsense and waste have serious consequences that cause even more
- People spending time brainstorming and looking for (and no doubt finding) reasons with potential for action
- Deciding which solutions seem best and implementing them.
Vague solutions to vague problems yield vague results
Mark Graham Brown has estimated that 50 percent of time executives
spend in meetings involving data is waste.
"The most important figures that one needs for management are unknown or unknowable, but successful management must nevertheless take account of them."
– the estimable late Lloyd S. Nelson, a true statistical giant
Consider other glaring everyday examples: daily operational reviews, financial reviews, and the never-ending budget meetings. What is the cost of all this everyday data INsanity? – "Unknown
or unknowable," but who needs numbers?
The human tendency is to treat ALL variation from any goal as a special cause.
But what does a plot of the performance tell you?
Here is the chart that resulted from the 12 months of data.
(You would know how to do this in a boring meeting, right?)
Goal: < 10 percent
Given the chart of just these 12 months, one can now interpret the situation for appropriate action:
- It's all common
- Because its average was 10 percent, it has met the goal all year.
In other words, each data point is merely statistical variation on a process perfectly designed to consistently produce 10 percent cancellations / no shows.
Amused and curious, I requested more data and merged them with these 12 months to create a chart of 32 months' performances by adding the 12 months of the previous year and eight months into the then current year.
There is no signal that the literal difference of +0.6% between the first two years is a special cause. In fact, performance was stable for this entire 32 month period – and meeting the goal as well. "Meeting goal" should be determined by comparison with the
current stable process average.
I'm willing to bet there had been process changes: some additional confusion, conflict, complexity, and chaos may have been added due to well-meaning but misguided attempts by good, hard-working people to improve the process, no doubt including the obvious: "Call the patient one to two days before to confirm."
Speaking of which...
A great story sent to me by a deeply respected UK physician colleague after last newsletter
For those of you who like examples, it doesn't get any more real than this:
"The issue of did not attend (DNA) at clinic is one that I know well because I was clinical director for outpatients for three years and I had to deal with senior management's insistence on delusional ratios and arbitrary targets...and the monthly [Red, Yellow, Green] chart and pointless discussion that ensued.
"I presented the data back as an aggregate monthly run chart, just as you would, and showed that their ‘drive’ to
reduce the DNAs was ineffective.
"That of course just exposed the gap between the intent and the impact and just generated defensive behaviour. Completely counter-productive.
"Their Plan (which they had been told to do...by the Lean Team, so it was OK) was to set up a call reminder service … i.e. to call all 750,000 patients a year 2 days before their planned appointment to remind them. I was
"So I suggested a small, quick, focussed study to improve our understanding of why patients DNA … to phone a random sample of 100 recent DNAs (less than one days-worth) and ask them the reason. The team could not think of a good reason to refuse … so they did … and the results were not what they expected.
"They were assuming that patients had not received letters or had forgotten (hence the untested justification for another
expensive ‘fix’). That was only true in a small %.
"The majority said 'Something more important came up' and quite a few also said 'and I tried to phone up to re-arrange but the lines were always busy so I gave up'."
"Unknown or Unknowable"
Note how using a chart and small data collection made the problem go from vague to more focused.
The Lean Team's solution, issued by edict, was based on...?
True performance improvement vs. goal doesn't involve the goal!
As Lao Tzu (Tao te Ching) might have said: If all of your focus is on meeting the goal, you will not meet the goal.
Frightened people are
very clever. They will waste a lot of time and energy doing their best to meet an arbitrary numerical goal by any means possible – to the point of playing the game of working on the number instead of the process (other than distorting it to meet the goal). What do these games cost you? "Unknown or unknowable."
As in the example above, you could at least "plot the dots!" to see whether your interventions are
A better use of meeting time? What if these important indicators were presented using process behavior charts to assess their performance relative to the goal?
- If an indicator is not meeting its goal, what if the person took personal accountability to "own" their
result and answer the questions, in dialogue with leadership, What else is it going to take?
- It is then up to leadership to "own" the person's success (not necessarily suggest solutions) by dealing with appropriate issues, especially cultural resistance, interdepartmental barriers, and other such "cultural
Any goal has absolutely nothing to do with how one would go about improving a process's performance to achieve it.
This is only one example demonstrating how the everyday organizational use of data process is a staggering hidden opportunity, the extent of which is "unknown or unknowable."
If you can't right now think of at least a dozen similar examples
in your work, plot a number that makes you or your organization "sweat" over time. Begin to apply some of what I have talked about to your chart and watch how the conversations and the ways you think about data change.
Possible suggestion: What about looking at budgeting as a "spending process – for which you have history – to predict?
How many statistical tools did my colleague use? A chart and
simple data collection.
Which improvement process would you prefer? My colleague's or implementing a "known" solution demanded by the Lean Team (based on...)?
Just "Plot the dots!" please.
Chapter 6 of Data Sanity covers a process-oriented approach to statistics and "Plotting the
Data Sanity: A Quantum Leap to Unprecedented Results is a unique synthesis of the sane use of data, culture change, and leadership principles to create a road map for excellence.
One of its major goal is to create a common
organizational language for healthier dialogue about reducing ongoing confusion, conflict, complexity, and chaos.
Please know that I always have time for you and am never too busy.to answer a question, discuss opportunities for a leadership or staff retreat
, or public speaking
-- or just about any other reason!
hesitate to e-mail or phone me.
Please visit my LinkedIn page
to listen to a 10-minute podcast
or watch a 10-minute video
where I talk about data sanity.
Was this forwarded to you? Would you like to sign up?
If so, please visit my web site -- www.davisdatasanity.com
-- and fill out the box in the
left margin on the home page, then click on the link in the confirmation e-mail you will immediately receive.