Problem Solving Paper Mba 510

Voice of the Customer

Domestic US automotive manufacturers are faced with fierce competition from their Asian and European rivals, making mere survival a constant challenge. With high quality and fuel economy as the primary factors for deciding a vehicle purchase, consumers constantly raise the bar for automotive manufacturers, forcing them to revisit their design and manufacturing processes in an effort to meet and exceed expectations. Third party consumer agencies such as Consumer Magazine and JD Power play an important role in influencing consumer decisions by providing an unbiased evaluation of competitive vehicle sets. Successful corporations have come to recognize the importance of customer’s voice and implemented processes that integrate voice of the customer into product development cycle before inception product design. On the other hand car manufacturers are under extreme pressure to contain operating cost while delivering high quality products. This more than ever justifies the need for a market oriented research process that helps manufacturers understand how customers perceive their products and focus development efforts in areas that meets customer needs and adds value for the consumers. Statistical based market research provides car manufacturers with a valuable tool for aligning the functional objectives to customer needs and wants before investing time and money in the development of a new product. Implementation of statistical methods in everyday decision making is now more common than before with integration of Six-Sigma techniques now part of the vehicle development process, limiting total number of product defects to only 3.4 per million. This paper discusses potential application of descriptive and inferential statistics in making product strategy decisions early in development process through analysis of a set of real world secondary research data obtained through focus groups and consumer surveys. Through observation and subjective ride and drive clinics, car manufacturers can generate valuable data on customer preference and focus the design and development effort on features of the car that matter most to the customer and ultimately contain development overall cost.

While engineers and designers have traditionally used competitive benchmarking as part of the development process, the process in the past lack direct input from the end user of the product, the customer. This paper depicts an example of how customer input can form the basis for future cars by integrating the customer voice an integral part of the design process.

Business Problem Statement

Developing competitive and profitable world class vehicles requires US car manufacturers to focus design and development efforts on features and attributes that matter most to the customer through a research based competitive benchmarking process.

Research Based Benchmarking Process

As an initial step in the bench marking exercise, US vehicle manufacturers can identify customer preference and their top voices through parking lot interviews and through observing owners of the vehicles within the segment of interest. The outcome of this experiment is a list of wants and needs such as “I want a quiet car”. This process is followed by ride and drive clinics where participants are asked to rate several vehicles including the baseline vehicle being assessed for potential design enhancements and at least two competitors. Participants are asked to rate these vehicles in the categories established through customer voices such as interior fit and finish, quietness, cargo space, performance during acceleration and ease of controls. The experiment discussed in this paper requires participants to rate the vehicles being evaluated on the scale of 1-10 with 10 being excellent and 1 signifying major functional problems. The decision process involves inference from a set of ordinal ranked customer data using a nonparametric method known as Kruskal-Wallis Test or H Test.

Null and Alternative Hypothesis

The most crucial piece of information derived from customer data is to assess where the base vehicle stands against competition and identify weaknesses and strength. US car manufacturers can leverage the data to justify the need to improve the functionality of those vehicle attributes that are most important to the customer. The result of the benchmarking experiment is likely to suggest the need for a complete redesign if all key attributes fall short of the competition. Participants provide a subjective assessment of key vehicle attributes identified through top voices of potential customers. While the customer based bench marking exercise performs an overall evaluation of all three vehicles, this paper examines only a subset of the data that ranks the vehicles for their performance during acceleration. Here the hypothesis is that the domestic vehicle under investigation is in the same class as its foreign rivals and that its performance is comparable to the competition. “Hypothesis is a statement about a population parameter developed for the purpose of testing”, (Lind, et al, The McGraw Hill, 2004, P. 317). Similarly hypothesis testing refers to a statistics based procedure to determine whether the hypothesis statement is reasonable, (Lind, et al, 2004). The hypothesis testing adopted here is especially designed for ordinal-level ranked data referred to as H Test developed by Kruskal-Wallis in 1952. The test statistics does not require normal distribution and can be approximated to Chi-square. For the purpose of this work a significance level of is assumed.

There is no difference between the performance during acceleration of the domestic vehicle and the two competitor vehicles.

The competitor vehicles are superior in performance during acceleration compared to the domestic vehicle.

Data Source

The data is of secondary source and is derived from a ride and drive clinic conducted by research and marketing firm GFK to identify market trend and customer preferences within the small segment vehicle. The experiment is designed to allow a sample size of 29 participants perform a subjective evaluation of a domestic hatchback vehicle being considered for future enhancements against two similar size foreign rivals.

Analytical Approach and Tests Statistics

The test statistics adopted here is the Kruskal-Wallis test designated as H.

Kruskal-Wallis Test

, , are the sums of ranks