Forecasting with Qualitative Methods

The qualitative forecasting uses an approach and estimation methods based on the perception, evaluation and judgment of managers and experts within and outside the company. Thanks to their knowledge, expertise and experience, these people are able to build sales forecasts and other variables (market shares, new products, etc.) that fill gaps, inadequacies or the impossibility of quantitative methods.  

Qualitatives vs Quantitatives methods ?

The main difference between the two types of forecasting is that qualitative forecasting is subjective while quantitative forecasting is based on objective and precise calculations.

Quantitative forecasts are based on historical statistical data, both short and long. Forecast models use this data (exponential smoothing, simple regression, Box and Jenkins, etc.) to extrapolate and forecast sales for example for the coming year.

One of the advantages of quantitative techniques is that the forecast has, on the one hand, a solid database of real historical data and, on the other hand, very specific forecasting methods.

If the company has historical historical sales series or any other variables, a quantitative forecasting approach would be better suited, because the various statistical forecasting models will eventually find the one that minimizes the forecasting error. In addition, “exogenous” factors such as competition, consumer purchasing power and prices can be taken into account in modelling using econometric or multiple regression models.

Otherwise, if statistical data are poor or or not available, and there are far too many factors that impact sales, more qualitative forecasting methods would be better suited.

When you should use qualitative forecasting methods ?

  • Decision-makers and managers today have a wide range of quantitative and qualitative forecasting methods and approaches. The choice of methods depends on several criteria (see table below) the most important of which is whether or not historical data on the variables being predicted. In this case, the use of qualitative forecasting is the only alternative.
  • Take the case of the production of a brand new product for which the company that will produce it does not have historical data or even a “profile” of competing or similar products including in other countries. The qualitative approach can be used to construct forecasts with very specific assumptions.
  • To estimate their market for new aircraft, Boeing and Airbus use qualitative approaches to essentially determine the number of new aircraft to be produced in ten years or more. Because the production cycle of new aircraft, and even for models derived from existing aircraft (family of airbus A xxx) is rather long. For a new type of aircraft, the cycle is at least eight years while for a derivative model it is at least five years. In these cases, the methods used by these manufacturers is sometimes the combination of several qualitative methods (Delphi, intention to purchase airlines, …).
  • In some cases, such as the launch of a new product, companies develop forecasts using both qualitative and quantitative approaches. Let’s take the example of the launch of a new product that replaces a previous product whose demand has dropped sharply and which the company no longer wants to produce and market.
  • It is possible to use a quantitative approach such as the analog approach of modeling the profile of the previous product and adapting it to the new product. Growth and diffusion curves are used for this purpose.
  • It is also possible to use the qualitative approach to compare the two types of forecasts and can be filled in the gaps if there are any. This would allow business leaders to make more informed decisions.

Pros and Cons of qualitative methods

Qualitative methods are better suited to contexts where products and services sold by the company, for example, evolve rapidly or erratically.  It is therefore important to understand both the pros and cons of qualitative forecasting

Pros

  • Several qualitative forecasting methods can be done quickly and easily without having to collect massive and detailed statistics.
  • These methods offer a great deal of flexibility because they allow for a multitude of contrasting scenarios based on the assumptions adopted by the participants.
  • The skills, quality of judgment and experience of internal and/or external participants can greatly improve the company’s predictive capacity.
  • Improved forecast accuracy by taking into account factors that will impact future sales that quantitative forecasts may not explain, such as rapid market developments, competitors and customer purchasing behaviour.
  •  These methods can be very useful in situations where data is inadequate or simply missing and can help create a consistent vision of future sales developments.

Cons

  • Several qualitative forecasting methods can be done quickly and easily without having to collect massive and detailed statistics.
  • Most of these methods are subjective because they are based on opinions and judgment.
  • If the group of experts is not truly representative of the areas that require forecasts, the accuracy of these can suffer.
  • The opinion and judgment of some members of the group may influence other members.  This is the case when you have executives and line managers in a group representing several levels within the company.
  • When there are business objectives, participants may be overly optimistic or pessimistic about their sales forecasts.
  • Forecasts can be skewed by selective perception of the participants, especially when they ignore relevant, recent and structuring information that actually contributes to sales.

Some examples of qualitative forecasting methods

There are multiple qualitative methods of forecasting. Here is the classification most often found in the literature:

Experts Opinions

This method uses the opinions and expertise of those participating in the forecasting exercise.  It is among the easiest to implement and resources are not too costly.  Gathering the opinion of participants from the various departments of the company (marketing, production, sales, …) allows to reach a consensus on the forecast. This can be repeated periodically to update the forecast.

External consultants and experts can join the group formed at the company level to provide information that would be lacking (competition, regulation, etc.).

The Delphi Method

This method is well suited to long-term forecasts in a market or industry that expects significant changes or external events that will impact sales, as in the case of the aircraft manufacturers above.

In this method, a panel of experts fills out, without meeting, a questionnaire individually and anonymously with their comments. There is little chance that a consensus on the forecast will be reached as early as this first round.

We then move on to a second stage where these forecasts are aggregated by an external delphi coordinator and shared with the group, with answers always kept anonymous. The same group then completes the questionnaire again and has the option to modify its responses based on the answers provided. This sometimes requires three to four rounds before a final consensus on the forecast is reached.  

Consensus of Sales Team

This method is similar to that of “Experts Opinions” except that it integrates business and sales people of the company.  An undeniable advantage is that they are close to current customers and in contact with prospects.

Each person can give an opinion on sales forecasts and how customers will likely react to a different product or service. The answers can then be averaged, for example, to make the final forecasts.  Finally, as there are often issues in terms of trade objectives, participants may be overly optimistic or pessimistic in their estimates.

Customers surveys

Here, it’s about exploring customer expectations for launching a new product, for example. Customer surveys can help identify key trends and changes in the market, leading to sales forecasts.

Submit a comment