Forecasting FAQs

Click on a link below to access that FAQ (frequently-asked question) topic.


Demand Drivers
Dropping a Product and Sales Forecasting
Forecasting Accuracy Calculation
Forecasting Hints/Tips
Forecasting Sales For a Double-Run
Forecasting Sales in a New Market Region
Forecasts and Active Products
Initial Forecasts
Sales Forecasts and Production Orders
Weird Results




Demand Drivers

“We're interested in understanding the key drivers of our business based on price, quality, service, and availability for our set-top box products. Do you have any suggestions on how we might approach this issue?”
Understanding how your "controllables" (price, product quality, service quality, and availability) influence your results (market share, volume) is central to sales forecasting and running your business. Please view this video for more details.

revised 05/04/2018
[000076.html]
listed under "Advice"
listed under "Forecasting"
listed under "Generate Demand"





Dropping a Product and Sales Forecasting

“When we drop a product from active distribution in a market, do we have to change its sales volume forecast to 0?”

Sales volume forecasts are only relevant for active products. When you drop a product, you can ignore the associated sales volume forecast and just leave it at whatever value it was before you dropped the product. Changing the forecast of a dropped product to 0 is unnecessary, but no harm will follow if you do so.

revised 11/21/2004
[000209.html]





Forecasting Accuracy Calculation

“There appears to be a mistake in our forecasting accuracy calculation. With a forecast sales of 10,275 and actual sales of 5,902, the Forecasting Accuracy Report shows a forecasting accuracy of 25.9%. But, my calculation yields a forecasting accuracy figure of 57.4%. What’s going on?”

As described in the LINKS participant’s manual, forecasting accuracy is defined as accuracy = 1 – (absolute difference)/actual, expressed in percentage terms. For your cited figures, accuracy = 1 – (4373/5902) = 1 – 0.741 = 0.259. This is the 25.9% forecasting accuracy figure reported in the Forecasting Accuracy Report.

You are apparently thinking about “(absolute difference)/forecast” rather than “(absolute different)/actual” in calculating forecasting accuracy to be 57.4% for forecast sales of 10,275 and actual sales of 5,902.

Actual sales is the denominator (the base) in the forecasting accuracy calculation because actual sales is the “target” of the forecast. In your case, a difference of 4,373 units from a target (actual sales) of 5,902 is a poor forecasting accuracy result, thus the resulting forecasting accuracy score of 25.9%.

revised 02/13/2007
[000117.html]





Forecasting Hints/Tips

“Do you have any hints or tips to improve our sales forecasting?”

Create a spreadsheet that tracks historical sales for each product, channel, and region. Graph the sales history for each product, channel, and region, paying special attention to the most recent couple of rounds of data. This provides a baseline forecasting capability (a simple extrapolation of past sales), assuming no changes in the market (e.g., no reconfigurations, no pricing changes, and no marketing spending changes). Then, modify your forecasts in light of your forecasts of potential changes in customer-facing initiatives for both you and for your competitors. Use the "Judgmental Sales Forecasting Worksheet" in the LINKS participant's manual to systematically organize your forecasting process.

revised 06/17/2004
[000184.html]
listed under "Advice"
listed under "Forecasting"





Forecasting Sales For a Double-Run

“How should we forecast sales for our last two rounds which will be a LINKS double-run?”

Forecast for the next round only, and that forecast will automatically carry-over unchanged into the second round of the double-run.

revised 04/13/2016
[000267.html]
listed under "Forecasting"





Forecasting Sales in a New Market Region

"How do we forecast sales volume in a new market region?“

It's obviously a major challenge to forecast initial sales volume in a new region. But, even more importantly, there's no a-priori knowledge about what kind of set top box customers prefer in a new region.

The largest uncertainly about a new region is customers' propensity to consume/purchase set top boxes. Even if customer demographics (numbers of potential customers) and economics (purchasing power) are comparable to pre-existing regions, new-region customers' interest (propensity to consume/purchase) in set top boxes could be higher or lower than in existing regions.

One approach is to wait and don't enter a new region immediately. Perform relevant research, especially about product configuration preferences of customers in a new region. And, continue to concentrate on existing regions. This delayed-entry strategy also allows initial demand for other firms' products launched into the new region to be revealed (again, by ordering appropriate research students for the new region).

Just plunging into a new region has lots of risks, but perhaps doing so with one product is viable ... with a modest launch effort (including modest commitment of marketing support spending) ... just to 'test the waters.'

revised 04/11/2016
[000266.html]
listed under "Forecasting"





Forecasts and Active Products

“We've input a forecast for a product in a market, but that product hasn't been activated. Why is that?”

All decisions in LINKS are independent of each other. A greater-than-zero forecast doesn't automatically activate a product. You must activate the product on the Generate Demand input web-screen by changing its "Active Status?" to "Yes".

revised 03/13/2007
[000135.html]





Initial Forecasts

“As we assume managerial control of our firm at the start of LINKS, there are forecasts already in place in our financial and operating reports. What is the source of these forecasts? What do they mean?”

These forecasts are sales realizations from the intial round in LINKS. The LINKS software made a forecast in round #1, when the simulation was initialized. It simply set the forecasts equal to current round #1 sales, thus leading to perfect forecasting accuracy in round #1. These round #1 forecasts have stayed constant as LINKS has advanced forward to the point at which you assume managerial control of your firm.

No sales forecasts are provided by the LINKS software, with the exception of the above-referenced "forecasts" associated with round #1. You must make your own sales forecasts throughout your LINKS exercise. The sales history at the bottom of the Forecasting Accuracy Report provides you with recent historical sales realizations, to provide some sense of possible trends. In addition, this sales history provides some information about the relative volatility of sales across markets. Of course, this sales history reflects "constant decisions" of all firms, since there were no changes in decisions of the firms since round #1.

revised 10/28/2004
[000196.html]





Sales Forecasts and Production Orders

“Do our sales forecasts take buffer stock into account in our production orders?”

Your sales forecasts are your best estimates of sales. Sales forecasts do not influence your production orders directly. You must make the mental linkages between your sales forecasts and your production orders. For example, suppose that you forecast sales of a product to be 10,000 units. If you have zero units of inventory on hand, what should you do for your production order of that product? 10,000 units is unlikely to be the correct production order in this situation. You'd probably want some buffer stock of production, to counteract demand variability. The bottom line should be clear: sales forecasts and production orders are separate and distinct decisions and you must mentally interrelate them in your supply management decision making.

revised 05/23/2001
[000123.html]





Weird Results

“The LINKS markets seem a lot more volatile than I’m used to in my own business experience. Is there something peculiar about the way LINKS behaves?”

The LINKS markets are what they are. Surely, the ultimate business principle is “listen to your customers” (markets). So, please do “listen” carefully to the various LINKS markets. You need to build your business practices around your markets, which may be different than markets in which you have personal business experience.

And please:

  1. Consider the term “markets” (with an “s”) … different markets may behave differently, so attend to such possible market segment differences.
  2. Don’t assume that your “Pacific” region is identical to the Asia-Pacific region with which you might have personal or business experience. The LINKS “Pacific” market could be different (different in size, in growth, in behavior, in sensitivity to various marketing instruments, etc.) than your personal Asia-Pacific region.
  3. Reflect on the possibility that markets can change through time (“drift” not “revolution”).

revised 02/05/2007
[000090.html]
listed under "Advice"
listed under "Financial and Operating Results"
listed under "Forecasting"
listed under "Strategy"




LINKS® is a registered trademark of Randall G Chapman.  Copyright © 2004-2018.  All rights reserved.