Executive Summary:
Optimal Configuration of Product Features: Using Online Auctions
There may be a better way to determine which products and features to offer the marketplace that hits the mark on both consumer preferences and firm profitability. Methods that firms use to configure and price products may not truly be the most profit maximizing or particularly most in synch with consumers’ real preferences. Research by Eli Snir and Marion Sobol of SMU Cox offers an innovative way—by the use of online auctions—to select and price features for products that would be the most efficient products to sell, at profit maximizing prices, especially for somewhat standardized or bundled products. Snir commented, “I don’t think anyone has used prices generated in online auctions to extrapolate the underlying structure of demand. And, I can’t recall that it has even been suggested as a generic tool with respect to what we are proposing.”
A Deep Question Answered
So, how does a firm choose which configurations or product features to offer the market? In the past, market researchers have relied on focus groups, conjoint analysis, and a number of other types of tools to dissect consumer preferences or utility. This research proposes a different methodology for measuring what consumers’ value. The emergence of active auction markets for a wide variety of products and services provides a venue for calibrating customer preferences. Coupled with cost data, this information facilitates the identification of efficient configurations for a firm to produce and offer to the market. As Dr. Sobol commented, “Online auctions would be beneficial for standardized products. The use of auctions, particularly online auctions, has not been proposed to determine optimal product bundling. In the past, auctions have been used for pricing single items but not for bundling.”
A firm may infinitely customize their products, but profitability often dictates choosing only a finite few to produce. “This method replaces the traditional focus group or market research approach where firms identify respondents’ trade offs between this and that feature—their willingness to buy,” Snir explained. “Prior methods can determine consumers’ stated preferences, i.e., what they would buy versus actual prices consumers’ will pay given certain configurations of features. Online auctions can facilitate what consumers’ actual (revealed) preferences are compared to stated preferences.” The applications for using online auctions are numerous and varied, and can be used on an ongoing basis to determine how consumer preferences are changing over time for different features and configurations. “The use of online auctions in this manner could effectively help firms reduce and beat obsolescence over time,” Snir stated.
Holding an online auction might take about one week, and data analysis could take a few days or be done in real time. Since consumer preferences are relatively stable over weeks and months, that value wouldn’t change so rapidly, therefore it would not negate the data or findings. Another suitable bundled product to ‘configure’ would be travel packages. From holding an online auction, a firm could determine how including different features would weigh in with consumers. The optimal bundle, for example, hotel with the best meal selections and a direct flight to the Bahamas in February/March, may be one bundle falling into the most efficient offering based on value to consumer and firm cost.
The Efficient Frontier, a Reality Check
Snir and Sobol developed an efficient frontier in this study, which serves as the optimal configurations or bundles to offer the market. It is the set of features for a product that could not be improved on in terms of optimal features and cost. In fact, the production of these sets of product configurations prove to maximize social welfare—the best offerings with maximum consumer value, at the least producer cost, and with less waste by producing products that are of greatest demand. In a complex product, with sufficient data, the final selling prices at auctions can be a good indicator of the demand for certain features using appropriate analysis to compare products' sales prices. Thus, online auctions can be used as a way to elicit consumer preferences so that sellers can raise their profit levels above those that would be obtained by simple pricing methods including cost plus pricing.
In this study, used Dell laptop computer prices (offered on their website) were compared with auction prices for the same configurations of Dell used laptop computers. These prices were observed in the same time period. Of the 800 different feasible bundles or features combinations that could have been offered (based on combinatorial statistics), only nine fall on the efficient frontier, i.e., nine products offerings were the most desirable by consumers at the least cost for the firm. “There are many ways to think about what would be an efficient product,” Snir explained. “If an omniscient central planner were to determine what to produce given that the market drives prices, what would he/she do? This set of configurations on the efficient frontier are driven by consumer value and manufacturer’s cost—what consumers want most and simultaneously how firms may best profit,” Snir recounted.
Because the online auction mimics consumers’ real behavior, this methodology provides a reality check on market conditions. Sobol commented, “There are many methods that marketers try, but we think this more realistically demonstrates the consumer’s behavior. Most of the things tried have been somewhat unrealistic, for example, simulation experiments with bidding where money is given to the respondent.”
Conclusion
Optimal product configuration is a complex task and companies require new tools to effectively introduce products. Consumer electronics manufacturers can benefit greatly from accurately forecasting which configurations offer the highest margin. Modularization and outsourcing enable companies to choose from a plethora of potential bundles of attributes; shorter product life cycles increase the opportunity cost of incorrect configurations. Snir and Sobol also see this research serving as a platform to analyze trends in customer value, therefore providing an estimate of obsolescence per feature.
With value to the consumer being ‘notoriously difficult to elicit,’ using online auctions for a wide variety of bundled products is a win-win for consumers, producers, and their distribution channels. This would also prove important in products where margins were thinning as well as due to product life cycle and competitive environment factors. This line of research helps validate for firms “what the market will bear” which, coupled with transaction data, can maximize profit.
|
