Home »
News & Trends »
Featured Articles

A Guide to Good Forecasting: Part 2

“Forecasting is something we do once a week. Quite simply, the forecast is the most up-to-date prediction of what our business or department’s performance results will be for a given period in the future. We usually forecast at least three to four weeks out; in some departments, six to eight, on key numbers. To use a sailing analogy, the annual plan is like charting an initial course before you leave port; the forecast is about checking current coordinates and then tacking effectively to end up at the agreed-upon destination.

What to Forecast
Ideally, every significant operational number in your business should be forecast. Start with a few key areas. Sales are certainly one. All large expense items are important—cost of goods and labor cost are somewhere from 50 to 70 percent of our costs, so we always forecast those. Each of our businesses forecasts cash on hand; most also forecast accounts payable and accounts receivable.

Five to ten critical areas are a good number to start with. Be careful not to make the work so overwhelming that no one can cope with it. And remember that deciding what items to measure and forecast is not something that is effectively delegated. It’s great to get input from others in the organization but, especially in the beginning, leaders must take an active role in determining what will be tracked.

Ultimately, we aim to forecast about 15 to 30 important items in each department or business on a weekly basis. Because we use what is referred to as a “balanced scorecard, we also forecast key service and food quality measures.

We forecast numbers that don’t show up on financial statements but are critical to success. At Zingerman’s Roadhouse, we forecast check average. At the Deli, we forecast purchases, inventory levels, etc. We forecast the number of proposals we’ll send to consulting clients at ZingTrain. We also forecast the number of new hires we’ll bring in. Again, the idea is to forecast any meaningful number that has significant impact on the performance of your business.

Who Should Forecast
Forecasting is done best by those closest to the work, not by the people who have historically had access to the numbers. In other words, sales forecasts should be done by someone who actually sells and talks to customers, not by an accountant. Labor costs would best be forecast by someone active in scheduling, not by the person who submits the payroll data. The delivery manager should address driving or delivery issues.

We assign responsibility for each forecasted item to an individual who then “owns that line—they report, forecast, explain variances between forecast and actual numbers, and are responsible for helping to get results. To work well, the forecast is not just a set of unattached numbers. A good forecast becomes a story, a story of what you believe—to the best of your ability—is going to happen.

How to Forecast
Front line people are not experienced at forecasting. At best, they have received reports on their numbers from office staff elsewhere in the organization. Almost everyone forecasting for the first time gets nervous and asks something like, “How can I possibly tell you what sales are going to be when I don’t know?

The answer is, “You’re right. You don’t know what’s going to happen, but you forecast anyway. Forecasting is not a science; even the best forecasters are off the mark sometimes. (Just watch those who forecast weather or movement in the stock market.) There are always uncontrollable factors that will impact the actual results and may or may not end up supporting the forecasts.

People’s natural reaction to that “out of control reality is not to forecast. “I can’t know what the sales will be, so why bother is a normal response. But forecasting, like anything else in life, is something you get good at only when you practice regularly. Anyone who works at it can become impressively good at forecasting. And the better we all grow to be, the better the organization runs.

5 Steps To Fine Forecasting
1) Get Data
• Check historical data by looking at performance over the past few years • Take note of macro trends in the world, product issues or other unusual events that might have impacted performance last year • Benchmark against others in your industry • Check current year-to-date and month-to-date data • Check recent trends in your business • Talk to front line staff to see what they think • Talk to customers • Check ratings from six months or so earlier • Keep up with the news locally and nationally • Check in with others on your team who are forecasting • Talk with industry peers • Check calendars for holidays, special events, school vacations, etc. (Note that holidays may fall at different times of the month—or with lunar-based holidays, different months altogether—from one year to the next.)

2) Check Your Math
All the various numbers being forecast need to tie out. So make sure that the numbers you are forecasting mesh. If you forecast sales of $10,000 a week with a labor cost of 20 percent but your cash flow forecast shows a labor expenditure of $2,400, something is not in synch.

Compare apples to apples. Is this year’s data comparable to last year? Or has a key product sales component or cost been moved from one line item to another? Be sure to contrast the most appropriate information—for instance, compare the third Monday in October 2004 to the third Monday in October 2003, not October 15, 2004 to October 15, 2003.

3) Check Reality
Do a quick reality check. Can the numbers you are forecasting happen? Divide the sales forecast by the number of sales you need to hit that forecast (based on average sale) and see if it makes sense. Can you produce the quantity of products required for such a big week of sales? Do you have enough staff to do the production work?

4) Run With It
Many people resist putting up a forecast on time because they’re worried it’s wrong. Remember: It’s a forecast. It is more important to forecast on time than to wait until you have all the numbers perfectly aligned.

5) Check Actual Results Against the Forecast
This is the key to getting good at forecasting. The week after you forecast, compare what you said was going to happen to what took place; when there’s a variance, you need to explain what caused it. Examine why you were higher, lower or on forecast. And, over time, watch the trends. If you forecast 10 percent high every week for four weeks . . . you get the idea.

Adjusting to Reality
At Zingerman’s, the annual plan is the peg that we’ve stuck in the ground to mark the place we committed to ending up at the end of the year. We don’t always make it but we don’t take that commitment lightly. If the weekly forecasts show we will be off target at the end of the year, we need to adjust to get back on plan. Forecasting allows us to adjust to the realities of week-to-week life in the trenches. But the idea is that usually—though there are exceptions—we’re giving it everything we’ve got to successfully meet or beat the annual plan numbers.

Ari Weinzweig is co-owner of Zingerman’s Delicatessen in Ann Arbor, Mich., and author of Zingerman’s Guide to Giving Good Service and other books. For information, call 734.930.1919.