March 1996 € Volume 6 € No. 3

Customer Connection

Forecasting 101


By Tom Wallace



In last month's column, I talked about the frequently adversarial relationship between marketing and manufacturing. Since a frequent contributor to this problem is sales forecasting, let's take a look at some of the issues.


Why bother with forecasting?
It's amazing, but I still hear this question in some of my travels. Often it comes from the same people who say, "You can't forecast this business."

My response: nuts. Of course the business can be forecast-perhaps not with great precision, but it certainly can be done. As a matter of fact, virtually all businesses do a significant degree of forecasting.

The problem in many companies is that people in manufacturing and materials do the "forecasting" by default, i.e., they order the long lead time materials and release the long lead time production items. Not having a forecast, they must guess. Unfortunately, most of these folks aren't close to the customers; they don't know about marketing's plans for promotions, pricing, sales force incentives and the like. Thus, they rely heavily on history rather than on the future outlook in the marketplace. When things get screwed up, it reinforces the beliefs of the people who are saying, "You can't forecast this business. See? I told you so."

Point 1: Forecasting is being done in virtually every company. The issues are who does it and at what level it's done. (We'll discuss this latter point-the best level at which to forecast-in a future column.)


Who owns the forecast?
In many companies, when I ask this question, crisp answers rarely come back. I hear things like "it all depends which forecast you're talking about" or "well, it's not very clear" or "it doesn't matter, because we use the forecast only for budgeting; we run the business on history and scientific (!) guesses.

That's too bad, because one of the best ways to increase customer service levels and reduce inventories simultaneously is to do a first-rate job of forecasting. And in order to do a first-rate job, we'd better have complete clarity on whose job it is.

The issue here is accountability, and the underlying principle is: The people responsible for developing the plan (in this case the forecast, the demand plan) should be the same ones who will be held accountable for executing the plan.

Thus, Point 2: The marketing and sales department "owns" the sales forecast. It's their job; they're experts on the demand side of the business, both planning and execution. People in other departments may support them, perhaps via operating the statistical forecasting system or otherwise generating basic data. But it's the job of marketing and sales people to review, update and modify the sales forecast; they own it.


How accurate should the forecast be?
When people ask me this question, I wince. I try never to use the word forecast and the word accurate in the same sentence. Why? Because it's a turnoff for the folks in marketing and sales who will be called upon to do the forecasting.

You people who routinely criticize the forecasters for their inaccuracy, ask yourselves a few basic questions. First, if the marketing people could predict the future with great accuracy, do you really think they'd be working for a living? Would they be knocking themselves out for 40 or 50 or more hours per week? Of course not.

If they could predict the future with great accuracy, where would they be? At the race track. And if the track were closed? They'd be at home on their PCs trading in stock options and speculating on pork belly futures.

Reality check: The forecast will always be inaccurate to some degree. The job of the forecasters is two-fold. First, get the forecast in the ballpark, good enough to enable operations people to do a proper job of initial procurement and production, capacity planning, etc.

The second major goal for forecasters is continuous improvement in reducing forecast error. In doing this, their goal is not to reach some enchanted nirvana of forecasting perfection. Rather it's to routinely produce forecasts that reflect what I call the "Four R's of Forecasting": Reasoned, Realistic, Reviewed frequently, and Reflect the total demand.

Forecasting is a process. It has inputs and outputs, just like a production operation. See Figure 1, which shows how the forecasting process accepts inputs and converts them into finished forecasts which adhere to the Four R's.

Point 3: Better processes yield better results and forecasting is no exception. When's the last time you did a major upgrade to your forecasting processes?

Although I don't talk about accurate forecasts, I do promote "good" forecasts. Good means that the forecasters are working the process, applying their knowledge of the customers, the marketplace, future marketing plans, and in general doing the best job they can. That works for me.

Point 4: Marketing owes "good" forecasts to their internal customers-the people in manufacturing, materials, and other departments who need this information to do their jobs well. Then these people can do a better job of developing and executing the production plan, thereby satisfying their internal customers-the folks in marketing and sales-as well as the external customers, the real ones, the ones who pay our salaries.
Tom Wallace is an independent consultant based in Cincinnati. He is the author of Customer Driven Strategy: Winning Through Operational Excellence (1992) and editor/author of The Instant Access Guide to World Class Manufacturing (1994). Tom is co-director and a Distinguished Fellow of the Ohio State University's Center for Excellence in Manufacturing Maintenance.

Commentary

Chief Executive magazine recently had an article about an executive I really admire: Stanley Gault, the CEO at Goodyear, and at Rubbermaid before that. If there's ever a management hall of fame, there has to be a place in it for Stan Gault. The interviewer asked: "Was there a single turning point that made you into the executive you are today?" Gault answered: "It was when I moved from distribution to product manufacturing. To go from distribution into the factory ... is a major career change."

Business Week recently had a cover article on Scott McNealy, the CEO at Sun Microsystems, a major player in workstations and also software for the Internet. His first job out of Harvard College was as a foreman at the Rockwell Plant making truck hoods. He left, got an MBA from Stanford, and went back into manufacturing with FMC Corp.

I can remember a few years ago visiting with Frank Merlotti, the CEO at Steelcase. Over the years, Frank was a major player in developing Steelcase's uncanny ability to deliver product to their customers on time all the time. His first job at that company was in the production control department.

I don't have any statistics, but it does seem that more corporate leaders have a significant operations background. Could this be one of the contributors to the profound rebirth of American industry?
Copyright © 1996 by the American Production and Inventory Control Society Inc. All rights reserved.
Click here to return to the table of contents.