We harp a lot around here about the differences between the
underlying data and the headlines; and about the associated dilemma between today's data, underlying or not, and the context. What does it really mean in terms of trends, timeframe, pattern and change points ? What we've found over and over again is that people tend to look at today's unfiltered data, extrapolate it into the future in their heads rather than based on some deeper analysis and then be surprised when the world turns out differently than they expected. What's needed is some way to collect, filter and present that data in a user-friendly, easily grasped and accurate format. A dashboard in other words that captures the essence of a particular problem. Take another look at yesterday's post on the most recent economic data and take a look at the charts...if those aren't "dials" about the health of the economy we don't know what is. The trick to a dashboard, which historically didn't come about by accident and wasn't perfected overnight - more like decades if not a century - is that it's based on understanding the underlying "machine" in question, sampling the right data and presenting it in the right way.
The immediate trigger for these thoughts was yesterday's post on the real data behind the curtain but the real trigger was an exchange with Tim Walker of Hoover's expressing concern about whether or not people were spending too much time worrying about the economy instead of doing their jobs. Now on the surface you'd think that we, with our constant Economy-Industry-Company mantra, would be far apart on this; and we are to some extent. The point being that one can NOT ignore the economy nor the industry as it will swamp you best efforts in a blindside. On the other hand if you spend all your time obsessing about headlines, especially distortionate ones, at the expense of performance that's equally bad. Perhaps worse. Tim and I didn't finally resolve our little discussion so this is a continuation.
Have you ever gotten off the highway in a strange/new city ? And been overwhelmed looking at the street signs, power/telephone lines, traffic patterns, etc. ? Too much data, no filters and no information. But as you get familiar with the street and the patterns you can start looking for the key data - your brain (& this is biological btw) creates a filter that focuses on the key & changing elements in the overall pattern. The same way a lifeguard scans for splashes not every swimmer. The catch is that building the right filters has to be suited to the job. Too simple and you don't get enough of the right information. Too complex and it takes a lot of time to learn the data and acquire the interpretive and decision-making skills required to "fly the plane". The complex, busy and massive cockpit dashboard may be heard to learn - but is it required for the problem ? As well of course as being right, minimal, and well-designed ?
And when you need a Dashboard that covers all the relevant information across a variety of key areas and indicators now you're scaling from simple problem dashboard to Moonshot operations control center. Yet, as GE/Immelt learned and demonstrated with their last quarterly
announcements, lack of the proper control is deadly dangerous. Think about it for a minute - GE is a well-run, famously controlled company led by a CEO whom Warren Buffett describes as one of the best. And they got blindsided by economic and credit forces they didn't anticipate or position for. And yet many of those forces were visible and publicly analyzed for a long-time, e.g. real estate as CalculatedRisk handles it. Obviously GE didn't have the right kind of super-dashboard or war room for the size and complexity of their organization. Read the Wiki description of the flight control center and all the myriad functions embodied in it to get an idea of what it takes to fly a shuttle mission. And then consider the analogy/metaphor/model to a global business.
It's one thing to design and build the right kind of control room but entirely another to use it correctly. As a bit of a stretched analogy bear in mind that ALL of the functions in NASA's "little" room are required just to fly orbital missions. For the scifi spaceships of our dreams that entire team, the data, monitors and computing power have got to show up onboard the trip thruout the journey. Contrast the "panel" in MS Space Simulator to the reality of NASA's FCC and ask yourself what's reasonable ?
We've been having a little fun with the dashboard analogy and the pictures but hopefully it's clear this is a really serious business. The sub-text beyond just understanding the problem well enough to design and build the instruments for the dashboard(s) and the right dashboard(s) for the war room is what then ? You, the operations director, have to be able to interpret that information flow, make decisions, evaluate changes and then decide & act again. In a continuous and on-going loop. That's a matter of training, experience, skill and attitude. Especially attitude.
Perhaps the hardest thing in all this is finding the right decision-making patterns and training you mind to them and then sticking with that discipline. After the break you'll find an interesting collection of readings on the importance and impacts of failing to come to grips with these problems to explain why this is important. Followed by some interesting stuff on how our minds work and how filter/dashboard building is so difficult mentally along with some complementary stuff on self-development and sustainable discipline. And ending with a suggestion of some hard-won rules of thumb to get you started. We particularly like the selections on the psychologies of misjudgement and how practice is required to implement change.
You might want to reconsider yesterday's post: Behind the Misperception Veil: What's that Data Behind the Curtain ? as dashboarding exercise ! :) Take a look at the set of charts again and think of them as speedometers, power status indicators and the like. How's your/our spaceship doing btw ?
Continue reading "Data, Dilemmas, Dashboards and Decisions" »
