| Marketing Triangle |
 |
Information is Power, But Only if it is Used.
A Newsletter from Custom Decision Support Inc. & Lieb Associates Vol. 3 No. 1, Spring 1998
Good marketing strategy rests on a combination of market focus, new product development, appropriate reward policies and the exploitation of appropriate technology. In this issue we address a number of these strategic business issues. Much of this work has been motivated by our clients’ concerns and ongoing problems.
Businesses either grow or die. This is the new business imperative. New businesses of manufacturing firms rest on the skills and creativity of their technical staff. However, that staff often does not have the background and orientation toward marketing that allows full exploitation of their innovations. Bob Cooper (McMasters University) has shown that the lack of strong market orientation and marketing information are the leading causes of new product failures after launch.
We have been asked to assist research and technical groups in major technology driven firms in building marketing competence. The overall goal of these efforts is designed to facilitate a market oriented process in R&D for new business development. Our objectives in these programs are to: (1) facilitate the critical new product review processes, (2) strengthen development teams, (3) provide a common business and market analysis language, and (4) help integrate market analysis into the product development process.
We have found five useful principles:
|
Teaching scientists is very different than teaching any other group of professionals. |
|
This is a long term process. No single learning experience will result in this change. Deming calls this the need for a "constancy of purpose." |
|
The learnimgs must become relevant to real problems. It is critical that the program encourages the practice of the "art" to existing projects. Only when skills are actually used, do they become part of the portfolio of useful tools. |
|
We have found that using pre and post testing with workshops and seminars enhance the activity, provides guidance, and tracks progress. |
|
Providing "Web Ready" access to notes and planning guides extends these key resources for practical use by the technical professionals. |
For more information on our seminars and workshops, see our web site.
It is usually infeasible and unwise to try to approach the market on a customer by customer basis. Policies set for one customer are likely to conflict with others. For a business to function totally in response to each of its customer’s desires generates chaos. Similarly, market wide policies are too general and are usually ineffective. We end up providing services that only a few customers want and set prices below that what many customers are willing to pay. The purpose of identifying market segments is to target strategy in a coherent and effective manner. This is intended to take advantage of selected customer differences while maintaining consistent programs.
We have found it useful to position market segments based on what customers value. We group customer values into three marketing factors: (1) price, (2) product performance, and (3) service and brand. Segments and customers are positioned based on the relative importance of these three factors. Results are shown for hypothetical cases on the following figure. This is a triangular plot with 100% allocations at the vertices. Values are estimated using marketing research data or based on the opinions of sales and product management.. Note that there is a broad range of potential positions for industrial products.
| Marketing Triangle |
|
For analytical purposes we divide the map into four regions. three of these regions are dominated by one factor. The remaining, center region is a balance in which no single factor could dominate. Experience has indicated that the segments positioned in this center region are usually not well understood.
Each of the other regions suggest specific goals and strategies. For example, region II, the performance demanding segments, pricing should be based on value. Region III is dominated by the value of brand and services. Here we might wish to competitively price based on price premium. The suggested goals and strategies for the regions are shown on the following table
| Goals | Principle Strategy Pricing | Pricing | |
|---|---|---|---|
| I. Price Sensitive | Customer Share | Cost Reduction | Strategic Pricing |
| II. Performance Demanding | Long Term Profits | New Product Development | Value Pricing |
| III. Risk Reducing- Branding | Share and Earnings | Quality Improvement | Competitive Pricing |
Computing Bonus Qualifications: Using Customer Satisfaction
The Techno-Tip column consists of suggestions and comments for data analysis. It is intended to help analysts and managers directly involved in the analysis of business data.
It is critical to maintain customers. Edwards Deming asserts that 80 percent of new business comes from existing customers. As such, many firms have undertaken programs to reward personnel based on customer satisfaction. The reward is usually in the form of a bonus paid when the percentage of the relevant customers are satisfied. Ideally, bonuses would be given out when the percentage of dissatisfied customers fall below some threshold.
Beyond defining when customers are dissatisfied and the threshold below which bonuses would not be given; there are additional problems using customer satisfaction study data. The relevant number of customers that an individual is responsible for may vary. Furthermore, the sample size that is used to estimate dissatisfaction may differ from group to group.
These sampling issues introduce potential statistical bias that may influence the ability to have a fair policy. This is a statistical problem with a statistical solution. Confidence intervals around the percentage of dissatisfied customers allow us to determine within a given confidence level whether a person should be eligible for the bonus. The confidence interval captures variation in the number of customers and sample size. In essence, we compute whether a person should get a bonus given a target confidence level. In an earlier newsletter (Summer 1996) we showed how to construct conference intervals around a percentage. The confidence interval is determined from an appropriate sampling distribution. The following Microsoft EXCEL function computes the confidence interval.
Confidence Interval =
BETAINV(Level,[P · #],[# - (P · #)+1],0,# )/ #
where Level is the confidence level (i.e., 90%), P is the measured proportion of dissatisfied customers and # is the effective sample size. This formula would generate the confidence interval when the number of customers that a person is responsible for is large. To correct for small sample populations, the effective sample size, #, is given by the following expression:
# = n{(N-1)/(N-n)}
where n is the sample size and N is the number of relevant customers. The bonus policy, then, consists of the definition of a dissatisfied customer, the threshold level and the confidence level. The following chart shows the resulting policies in a health care area.
 |
The Software Review column captures our business data analysis experience with particular types of software packages. It is intended to help our clients in their selection process.
The more I used Microsoft EXCEL for data analysis and modeling the more impressed I have become with its power and ease of use. Even five years ago, we developed most of our models using standard procedural languages like BASIC or within a database such as dBase2. Today we use EXCEL almost exclusively for market and business models. The resulting models have the look and feel of standard Windows applications, without the expense and difficulty of traditional development. In addition to its ability to make a pretty system, EXCEL has a large number of functions and capabilities that make development easy and allows price optimization.
The major problem with EXCEL based simulation models has been its traditional slow speed and inefficient use of memory. However, the new generation of personal computers have given us speed and memory to spare. Recently, we developed an EXCEL spreadsheet market simulation model with over two thousand respondents, twenty variables and five competitors. Previously we would consider this well beyond the scope of spreadsheet systems. Today it is not only feasible but the resulting system is efficient. Moreover, development required only hours of effort rather than weeks.
A lot of things are happening in the world of software applicable to planning and marketing research. I recently got my hands on language analysis software (Diction 4.0) that could be useful in the analysis of verbatim data. I will report on this software after we test it on real marketing research data. Finally, we have started to use a dictation system (IBM ViaVoice Gold). This type of system requires a great deal of experience to get use to it. It’s going to take me a while. When I’m comfortable, I’ll report back.
"Empowerment" has become the mantra of Corporate America. Unfortunately, it has also become a cliché and grit for Scott Adams’ Dilbert. Basically, the problem lies in the conflict of needs. You can not have empowerment without control and direction.
Here, we might be able to learn from the US Navy. The Navy must give decisional power to its front line commanders. Ships are often thousands of miles from command centers and have awesome power. Over the centuries, the Navy has developed a concept of "doctrine," whereby commanders are indoctrinated into a culture by which their expected actions are predictable and aligned with the overall command. This concept may be a useful metaphor for control within an empowered organization. However, this approach does not come easily or cheaply. It involves heavy commitment to training, investment in people and the corporate culture. But, it is likely to provide the benefits that merely slogans can not yield.
I appreciate any comments and contributions from you. You can get in touch with me at Custom Decision Support, Inc., Phone (610) 793-3520. Fax (610) 793-2531 or E-Mail at genelieb@lieb.com
Gene Lieb (Editor and President)