New Product Forecasting with a Purchase Intention Survey

Abstract

Purchase surveys give good forecasts of new products... purchase probabilities are higher for longer than shorter periods... some modifications in the categories used in the survey may further improve accuracy. A sales forecast often plays an important role in marketing/ planning. It provides a basis for estimating profits associated with implementing a particular marketing strategy. A common forecasting method used by businesses to forecast sales of a new product is to ask potential customers the likelihood of buying the product that is being investigated. Consumer responses to probability statements like probably will buy or definitely will buy are often used by businesses to forecast sales of new or modified products. While this technique is widely used, there is little published research on the relationship between an individual's purchase intentions and actual purchase behavior. It would be useful to know what percent of those who say they are in the probably will buy group actually buy. In this article, we present a simple process for obtaining accurate sales forecasts from purchase intentions data. Along with this process, we report the results of an empirical study that links self-reported purchase intentions to actual purchase rates. FORECASTING PROCESS To obtain accurate sales forecasts of a new product from purchase intentions data we suggest following a three-step process: 1. Determine the demographic profile of the most likely target market. 2. Estimate the product's likelihood of purchase in the target market using a purchase survey over a specific time frame. 3. Forecast unit sales by combining likelihood of purchase with size of target market and a small set of important consumer attributes. TARGET MARKET Consider the problem of forecasting unit sales for a device that sounds an alarm when the wearer should get out of the sun based on skin type and the protection factor of the sunblock lotion used. (The authors were involved in producing a forecast of this product.) At the outset, one must determine the type of individual most likely to use such a product, or at least eliminate those individuals that will never use the product. Finding an appropriate demographic profile may prove to be a difficult task requiring considerable work with focus groups or survey research. For our example of the sun sensitive device, the company may only want to consider fair-skinned women from 25 to 55 years of age living in an urban area of the United States with a total household income greater than $25,000 as being the target market. This identifies the focus of marketing efforts and defines the population of interest to be used for the survey of purchase intentions. LIKELIHOOD OF PURCHASE The likelihood of purchase or estimated percentage of buyers for a particular product can be obtained by first asking consumers to indicate their likelihood of purchase using a multi-category purchase scale. A five-point scale is often used in purchase surveys, with the categories: * Definitely will buy * Probably will buy * Might/might not buy * Probably will not buy * Definitely will not buy Each of the multi-category purchase scale shown above is associated with a different purchase probability. As the first step in calculating the likelihood of purchase for a target market, the purchase probability associated with an intention category is multiplied by the percentage of individuals choosing that category. Next, results for each are summed across all categories to obtain the estimated percentage of buyers who will purchase the product. The procedure is demonstrated below for a five-category scale: Likelihood of purchase = Percentage of individuals choosing Definitely will buy x Purchase probability associated with Definitely will buy +. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 58 citations exist upstream.