When I was younger and stupid, I used to spend lots of passion, energy, and time around the development of terrifically insightful analyses. "Look at what the numbers tell us," I would
proclaim, waving a batch of color printouts ($1.50 per page at the time) of line, bar, and pie charts. (
see a few of my very favorites here).
After days or weeks of combing through the hairiest of data sets, I'd emerge with new strategies for rationalizing the sales force; reallocating marketing spending; moving R&D dollars from
one project to another; or providing better service to more valuable customers. All ideas that would have made or saved the company lots of money.
Relatively few of them ever got
implemented, however. Somehow, people always managed to get in the way. Seems that every great idea had at least one key assumption in it which couldn't be 100% assured by the facts. Also seems
that what I once thought of as "facts" were just numbers. And people are much smarter than numbers. Especially when protecting their budget, turf, powerbase, or even their jobs.
You've heard the adage that "you can torture the data until it confesses"? Well I could make it confess, but time and again I would still fail to get a clear verdict from a hung jury
of managers.
Now that I'm older (and still stupid, but in more enlightened ways), I find myself thinking backwards more often. I start by thinking about the possible implications
of a proposed analysis path -- e.g., "What question am I trying to answer and what would I do if the answer were X or Y?"
This tends to focus me on the things that may matter most
to the "jury" of people most directly affected by the analysis, and to then engage them in the process of defining the question and conducting the analysis, so it's clear that A) my
goals are altruistic and transparent; B) the process is flawed, but still credible; and C) I care to understand the issue from their perspectives so I see ALL the possible variables and implications.
I'm then free to approach the analysis from a more productive angle right from the start.
I find this has three major benefits:
1) I spend MUCH less
time on analyses that don't ultimately have their recommendations adopted;
2) I tend to get at metrics that
aren't just insightful, but actionable; and
3) I get asked to solve much bigger, more strategic problems.
Most interestingly, the difference seems
imperceptible to most people. They just think I've done a more thorough analysis. In a way, I suppose they're right.
.em dna uoy neewteb taht peek s'tel tuB