How to Make a Good Forecast
You need a good forecast because you have to make decisions today that will affect the outcome of your business in the future. The need for a good forecast could be for the next few days, few weeks or months or a few years from now.
Forecasting is like peering into a crystal ball. You’ll be looking for trends that suggest that you invest your resources in one area instead of another. Your lead time for decision making could range from several years (for the case of capital investments) to a few days or hours (for transportation or production schedules) to a few seconds (for telecommunication routing or electrical utility loading).
Preparing a good forecast can be quite a challenge. A good forecast needs to accurate yet easy to make and adjust. It should be easy to understand and use.
Types of forecast
Most managers have to make a forecast at some time during the business cycle. The most common type is probably the annual sales forecast. Here you are trying to predict the future outcome of the various business decisions you are about to make. Perhaps, you will be tasked to forecast the future demand and sales of our company’s product, product line or services. The types of forecast you could be making are:
- a short term forecast or
- a long term forecast.
Management may ask you to make a forecast on the following:
- How much will you sell this year?
- How much demand will there be for your product or service?
- How much will it cost to produce the product or offer the service?
- Where should you invest so that you can make your targets?
- How much inventory do you need and when will you need them?
- Are sales coming in to forecast?
While even a good forecast can never be completely accurate, it is better to have one than have none at all. Many of us start the year by presenting our sales or revenue forecasts for the year. As the year progresses you can compare your actual performance to the forecast. The comparison may give you a better insight on the steps you need to take in order to hit your performance targets.
General Characteristics of a Good Forecast
- Forecasts are nearly always wrong. Instead of trying to compute a very accurate single value for a forecast, work on developing a system that develops a range where the desired value is likely to fall. Create systems that can adjust to a range of possible values. For inventory forecast, you can use the forecast error to develop buffer or safety stocks to cope with the uncertainty of demand.
- Forecasts are more accurate for groups of items or families of items rather than individual items. Your forecast for “filled donuts” is likely to be more accurate than for “Bavarian Donuts.”
- Forecasts are more accurate for shorter time periods compared to longer time horizons. The longer the time horizon, the more error will enter the forecast.
- Every forecast should include an error estimate. Since we now accept that there is an inevitable error in the forecast, your forecast procedure should give you an estimate of the error. The error can be expressed as Confidence or Prediction Intervals.
- If the demand is of the dependent type, calculating demand is more accurate than forecasting demand. Carefully examine the type of forecast you are making. Is the forecast for an Independent Demand or for a Dependent Demand? You need to understand the difference.
Seven Steps to Make a Good Forecast
- Determine the use of the forecast
- Select the items to forecast
- Determine the time horizon of the forecast
- Select the forecasting model(s)
- Gather the data needed to make the forecast
- Make the forecast
- Validate and implement results
Time Series Forecast Methods
These are Statistical methods based on historical data
- Naïve Forecast
- Moving Average Forecast
- Exponential Smoothing Forecast
- Seasonal Decomposition
- Trend Projection
- Linear Regression
- ARIMA Forecasting Techniques
Picking a Time Series Forecast
- First, look at the data. Software like Ms Excel can quickly generate a graph to tell you if your demand is stationary (level demand), has a trend up or down, seasonal or cyclical.
- Second, forecast using one or more time series techniques
- Third, evaluate the accuracy of the of the different forecasting techniques and
- Last, pick the best one and implement.
Preparing a forecast can be quite a challenge for many new managers or supervisors especially if they have has no prior training in forecasting techniques. RMP Consultancy regularly conducts seminars that can help you understand the concepts of forecasting as well as how to use a spreadsheet program to prepare the forecast. These seminars are very practical and hands-on (you work on forecast templates prepared by the Presenter) where you learn by actually doing the work.
Raffy Pefianco of RMP Consultancy has presented this seminar to many of the largest corporations in the Philippines and in Saudi Arabia. Learn more. Give us a call.
Attend our Seminars that Include Training on How to Make a Good Forecast
Upcoming Business Process Improvement Seminars
5 Sep 2019, Thursday, Business Analytics Essentials (One Day),
5 Sep 2019, Thursday, Statistical Process Control (SPC) Training - 2 Days,
16 Sep 2019, Monday, TQM: Creating a Culture of Continuous Improvement (Two Days),
1 Oct 2019, Tuesday, Business Analytics Essentials (One Day),
28 Oct 2019, Monday, TQM: Creating a Culture of Continuous Improvement (Two Days),
18 Nov 2019, Monday, TQM: Creating a Culture of Continuous Improvement (Two Days),
21 Nov 2019, Thursday, Focused Improvement: The 7 Quality Control Tools,
19 Dec 2019, Thursday, TQM: Creating a Culture of Continuous Improvement (Two Days),
Register for one of the Seminars Above
Register for the Seminar
RMP Consultancy Seminars are available in the following formats:
- Public Seminar at our training center in Quezon City Philippines
- In-House (at your site) Workshop Format for a Max of 25 pax (Recommended)
- Special Exclusive Run
- One-on-one Coaching
Please Call Marco or Claire:
Landlines: Tel (632) 697-0767, Tel (632) 579-0233
Mobile (Globe) 0917-882-5578, Mobile (Smart) 0920-918-7358
Comments are closed