Management Is Prediction --DickesonMarch 2002
Here's what I believe top management needs to know at 9 a.m. every Monday for the preceding week on a rolling 13-week basis. Why for the immediately preceding 13 weeks? That's one-fourth of a 52-week year. It's current and provides needed context for converting weekly information into knowledge.
a) Sales invoiced
b) Materials consumed
d) Summary of period expenses paid
e) Earnings before depreciation and amortization
f) Summary of cash collections
g) Week ending cash position
h) Inventory summary
i) Days materials on hand (turnover)
j) Accounts receivable summary
k) Days sales on hand (turnover)
That's a spreadsheet with 11 columns and 13 data rows, plus some totals, weekly averages and a 52-week projection. Is it possible to have this weekly? Of course it is. And can we have simple graphs that picture the ups, downs and trends of the numbers?
If our system can't do this, we'll chuck it and get one that can. We're talking survival stuff here. Would this current information improve our predictions and make us better managers? What do you think?
Make Your Lists
That's a need-to-have information support list for top management. Now make similar lists of "need-to-knows" for each level of management accountability. Predictions are being made at all levels of management. Knowledge isn't restricted to top managers
Moving on from "need-to-have" reports, let's provide "drill-down" capability for the items. Put your cursor on a number and click to expose details. Increase knowledge depth to improve forecasts. Next, provide "moving range" charts for data points of the "need-to-have" information. With those XmR charts we know not only the average weekly experience of the past, but also the "range" of variations that can occur in our predictions. This builds decision courage and confidence. We must acknowledge, not contest, the limits of our projections.
Lay Down the Law
We know that routine variation is a law of economic life. Ninety-nine percent of the ups and downs are "chance" variations. Unless we substantially change the process there's a very high probability that deflections will be no greater than X and no worse than Y. Our predictive reliability is somewhere between X and Y. That's the best we can do. Look at the past and accept its variances as the measure of what's "normal."
Other industries are doing this in order to constantly improve quality and management by prediction. The fad of the moment is "Six Sigma" variation. Jack Welch, past CEO, decreed that General Electric would strive for this. But in printing we're still mired in the past, diddling with job costs based on the Farmer's Almanac. "What's a Sigma, Mr. Welch?" we ask. All we want is to be able to say, within limits, what we can expect a few days at a time...like the television weatherman on the local news.
If we feel lucky and want to bet that we can "squeeze a job in," and still deliver all the jobs on time, that's our prerogative as risk managers. But if we're smart managers we'd better know the risks of chance variation, hadn't we? We're not going to satisfy customers for the long pull unless we make dependable predictions of delivery time.
Or, you may get lucky and collect a big receivable sooner than "average." But then again, you may not. Want to bet the payroll on it? Are you a banker or a gambler? Your choice. How much confidence do you have in your guess...uh, I mean, prediction? Yes, managing is predicting, isn't it?
—Roger V. Dickeson
About the Author
Roger Dickeson is a printing productivity consultant based in Tucson, AZ. He can be reached by e-mail at email@example.com, by fax at (520) 903-2295, or on the Web at http://www.prem-associates.com.