SCDigest Says:

Judgment is frequently biased with over optimism, or allowing recent events to have unwarranted impact. Judgment is also clouded by personal or political agendas, where the forecast is used to represent what a person wants to happen, rather than what he or she honestly believes will happen. What Do You Say? Click Here to Send Us Your Comments Click Here to See Reader Feedback

The following article is made by special arrangement between Supply Chain Digest and the Institute of Business Forecasting & Planning; a full pdf of the original article from the Journal of Business Forecasting can be downloaded here: Forecasting New Products by Structured Analogy, by Michael Gilliland and Sam Guseman of SaS.

Please consider attending the upcoming Demand Planning & Forecasting: Best Practices Conference April 28-30, 2010. Early bird special still available at time of this article’s publication.

Forecasting demand for new products is extremely challenging, as most any demand planner will tell you, given the lack of history and the variety of pressures that can arise to make the forecast turn out to be something very close to the business plans created before the new product has been introduced to market.

Actually, there are several variants to the basic new product introduction (NPI) process and related forecasting efforts: (1) New to the world products (entirely new types of products), (2) new markets for existing products (such as expanding a regional brand nationwide or globally), and (3) refinements of existing products (such as “new and improved” versions, or package changes).

Number 1 is of course the most challenging scenario.

Whatever the specific type of NPI, there are a variety of approaches to forecasting the new product’s sales, some far from scientific in nature. These include:

Executive Opinion: Executive management provides a top line revenue number, and the forecasting staff comes up with a product mix to meet the top line target, including some guess as to sales for NPIs.

Sales Force Rollup: The sales force is surveyed to create a bottoms-up forecast by customer and item – but do they really know?

Delphi Method: A structured, formal process for gathering forecasts and building a consensus. Participants are surveyed, the results are shared (anonymously), and participants are allowed to make adjustments based on the forecasts of their peers. Effective – but impossible to scale.

Prediction Market: Anonymous wagering is used to gather group opinion. An asset (such as a new product’s first year total sales) is traded in a virtual marketplace (like the stock market), and the market price is interpreted as a forecast. Little used to date, also can’t scale well, it is believed.

Analogy: This based on the assumption that demand for a new product will be similar to demand for “like items” of the past. Like items are identified (based on common attributes or other criteria) and the forecast for the new item is based on a composite of the sales history of the like items. It is commonly used for NPI forecasting.

However, the traditional analogy approach is of necessity often heavily based on various judgments by the forecaster or others in the company that amend the composite analogy, since by definition there really is no true actual history on the product, only history analogies. This can lead to problems.

(Supply Chain Trends and Issues Article - Continued Below)

For example, judgment is frequently biased with over optimism, or allowing recent events to have unwarranted impact. Judgment is also clouded by personal or political agendas, where the forecast is used to represent what a person wants to happen, rather than what he or she honestly believes will happen. As an example, if a sales representative knows his forecast is going to be used to set the sales quota, there is a natural tendency to under-forecast to have a low quota that is easier to beat.

A more structure approach to the Analogy method can produce better results by reducing the impact of human error in applying judgment.

Structured Analogy Process

The structured analogy process for new product forecasting has six main steps:

Step 1 – Query:  Find a set of candidate products that have similar attributes to the new product.

Step 2 – Filter: Manually remove inappropriate or outlier products from the set of candidate products.

Step 3 – Cluster: Cluster the candidate products according to their sales pattern and use judgment to manually select the most appropriate cluster to serve as the surrogate products.

Step 4 – Model:  Select the most appropriate statistical model for the cluster of surrogate products.

Step 5 – Forecast: Use the statistical model to forecast the new product. 

Step 6 – Override: Where necessary make manual adjustments to the statistical model’s forecast.

Judgment is used in many places in this process, but in a consistent, structured way. Using a real example, when using this approach to forecast a new DVD release, after the third step (Cluster) three basic demand profiles were created from an analysis of all previous DVD releases that carried an R rating. Judgment was used to select which of the three clusters was most appropriate for the new title being released.

A structured analogy approach can be useful in many (but not all) new product forecasting situations. It augments human judgment by automating the historical data handling and extraction, incorporating statistical analysis, and providing visualization of the range of historical outcomes.

Judgment is always going to be a big part of new product forecasting. A computer will never be able to tell us whether Lime Green or Day-Glo Orange is going to be the hot new fashion color, but judgment needs assistance to keep it on track and as objective as possible. While the structured analogy approach can be used to generate new product forecasts, it is also of great value in assessing the reasonableness of forecasts that are provided from elsewhere in the organization. The role of structured analogy software is to do the heavy computational work and provide guidance – making the NPI forecasting process as automated, efficient, and objective as possible.

Do you have any experience with a structured approach like this to using the analogy method for forecasting NPIs? What approach in general have you found works best for NPIs? Let us know your thoughts at the Feedback button below.

SCDigest is Twittering!

Follow us now at https://twitter.com/scdigest