J. Scott Armstrong
· Faculty Page
· Return to Educational Materials


People
· Faculty
· Former Faculty
· PhD Students
· Phd Alumni
· MBA Students
· Undergraduates
· Staff

Home

Techniques and Concepts for Marketing Consulting

Below, I describe techniques that are useful to those in marketing consulting. Words that are italicized and underlined are defined elsewhere in this list. For ambiguities or mistakes on the thumbnail sketches, send e-mail to Scott Armstrong. Revised August 2004.

Action steps: An integral part of any presentation. The idea is to conclude the presentation with a decisive plan of action. In other words, outline the steps that should be taken to implement the stated objective. To do this, it is important to keep in mind the "foot-in-the-door" technique by which one puts forth small steps with a time deadline in the near future; these steps should be ones that can be easily applied by the client.

Bootstrapping (or Judgmental bootstrapping): [As used in forecasting and decision making, not in statistics and computer science.] The development of a model of an expert by inferring the experts’ rules. This is analogous to conjoint analysis except that a small number of experts is used, rather than a large number of potential customers. The expert is presented with a set of situations, say 20, and is asked to make predictions (or decisions) for each. Regression analysis is then used to infer the rules. Judgmental bootstrapping can be used to forecast sales for products having alternative designs and alternative marketing mixes. Few consulting firms offer this approach, and few organizations use it (the Dallas Cowboys being one that did).

Brainstorming: Paradoxically, this is a highly structured set of rules to enhance creativity. Groups without rules are not as creative as they could be, according to the research. The rules have one major guideline: to reduce evaluation. To make it work, you need to appoint a facilitator and use a checklist. A useful alternative for small groups is "brainwriting,” where you take a period (say eight minutes) and ask people to write down all their ideas. This is typically used to generate creative solutions, but the problem can be expanded first by brainstorming problems. Groups seem to improve after using brainstorming a number of times. This technique will save much time when working in groups.

Conjoint analysis: A two-step procedure where you first prepare a set of alternative designs (using principles of experimental design that have cleverly been imbedded in this program); then you try to estimate the probability of purchase for various designs. Conjoint programs are popular in business. Design your own experiments and analyze them using regression (all this can be done with a good spreadsheet program). Conjoint software programs are available.

Core benefit proposition: What is the primary benefit that the product/service provides to the consumer?

Decomposition of judgment: One of the basic strategies of management, science; Break a complex problem into pieces, solve each piece, than reassemble. Use for forecasting sales and for decision making (especially for group decision making). Checklist rating scales are one application. Decomposition can be used for both brainstorming and evaluation.

Delphi: An iterative survey of experts. Use this for evaluation. Delphi provides substantial improvements over ratings by unstructured groups. Freeware provides guidelines and terms to help in the use of Delphi.

Demand and cost pricing: When you initially structure the pricing problem, define the feasible range. The lowest price is defined by your variable cost (assuming that you are not the Government, in which case the lowest price is determined politically).

Design for assembly: Design the product such that it will be easy to assemble. IBM did this for one of their printers and found that they could greatly reduce the number of parts, thus lowering assembly costs and improving reliability.

Design for use: Design the product so that it will be easy for the customer to use and to repair (relies on ergonomics). This step is often poorly done, as you may have noticed.

Devil’s advocate: One person in the group is assigned the task of trying to tear down a proposal - a useful procedure for evaluating your venture proposal prior to presenting it to the client. The key is that this be done as a structured approach with group support and that it be used only for a short period (e.g., ½ hour). If you adopt this role without getting permission of the group, you can become unpopular quickly. Note that it is the opposite of the "Build” philosophy.

Ergonomics: The study of how people interact with machines and products (and services). Protocols are useful for learning how people use products (see protocols).

Expectations: Ask people whether they expect to use a product. This is more general than intentions.

Experimentation: To learn about the optimal price to charge for a frequently purchased product, it is necessary to depart from what one believes to be the optimal procedure and to experiment.

Expert opinion surveys: In contrast to consumer surveys, expert surveys can be done with small samples. The technology is similar (see surveys). 

Expert systems : Explicit rules derived from experts and from previous studies. Often developed as a set of conditional statements (“If X then Y”).

Focus group: Use the same procedure as for a nondirective interview, but apply it on a group basis. A popular and overused technique that costs more than $2,000 per session, but you can do a decent job (with friends) at a low cost. Used early in project to generate ideas (e.g., about target market needs). It is only useful in rare situations. May be useful for ideas about advertising products that are highly visible and when a buyer’s decision is based on what others think of him (like have you ever heard of a liberal professor who dives a Cadillac?) One cannot draw statistical inferences from focus groups. Interestingly, the skill requirements to run a focus group are not high. In general, non-directive interviewing with individuals (group depth interviews) are more effective. For more, see "When Should You Use Focus Groups?".

Foot-in the-door: Select a small operational step and set a time deadline. The foot-in-the door technique is useful for gaining commitment, such as in presentations to get seed money or venture capital. Use this on yourself to experiment with new techniques.

Forcefield Analysis: List the pros and cons for the attitudes that you would like to change with the ad. Put the pros on one side of the page and the cons on the other. Then represent each by a line that approximates the strength of the attitude from the perspective of the target audience. When designing the advertising campaign, consider not only how to increase the pro items, but also how to reduce the cons. You can also use this technique for group decision making.

Formal planning: Improves group performance only when the group is small enough to reach commitment. Has four stages: (1) set objectives (2) generate alternative strategies, (3) evaluate strategies, and (4) monitor results. For example, new product introductions that do each step explicitly are more likely to succeed. In practice, few firms do this. See Systems Approach, Marketing Planning, and Planning Process Checklist.

Gatekeepers : Who is it that makes the purchase decision? This consideration is often important for industrial products. Consider that you may have to target advertising to the gatekeepers.

Intentions surveys: Ask people whether they intend to purchase your product. This can provide a useful way to forecast, especially for new products. Also consider asking people about their expectations. (see expectations.)

Lewin’s change process: For important attitude or behavioral changes, it is helpful to get those involved to go through an “unfreezing” phase before attempting the “change” which is followed by a “refreezing” phase.

Marginal analysis: When setting the media budget, the optimal budget should have equal marginal returns. Thus, if you can get a larger return by spending an additional $100 on radio advertising, rather than on newspaper ads, you should continue to shift money into radio until the marginal returns are equal. Also, of course, you want to ensure that the marginal returns are profitable.

Marketing concept: Product design (as well as the rest of the marketing mix) should start with the customer and it should look at the customers’ needs first at the highest conceptual level. In practice, this is typically an expensive and difficult approach.

Marketing planning framework: This framework first assesses the marketing objectives as well as the current and future status of the environment (both internal and external). The next step is to find opportunities/problems that exist within the market and to translate these into action steps involving the marketing mix. After a marketing program is designed, it is then carried out and monitored. The monitoring process brings us back to square one where we reevaluate the objectives, environmental status, and implementation of the plan, and then judge the effectiveness of the marketing plan. This framework enables one to keep abreast of the changing needs of the consumer, the environment, and the four P’s. For more, see Marketing Planning.

Monitoring: A formal process to review the performance against the plan.

Non-directive interviewing: The listener follows a set of rules to help suspend judgment and to focus the interviewer’s attention on listening.

NPV (Net present value): New products often require early expenditures in hopes of later returns. Thus, you should determine a cost-of-capital (given the level of risk) and apply this to the profit stream (or, more accurately, to the cash flow stream). In addition to an expected NPV, you should also estimate the confidence intervals to reflect the risk involved in a venture.

Objectives: Organizations often confuse objectives (where they want to be) with means (how to get there). Studies in organizational behavior show that groups can improve their effectiveness by setting objectives that are (1) explicit, (2) measurable, (3) relevant, (4), ambitious yet achievable, and (5) operational.

Parallel processing: Divide your group into two or more subunits and have each unit independently solve the same problem. Then compare the results to select the best solution (or to modify it). It is useful for increasing creativity. It also helps to guard against mistakes in analyses (separate analysts are unlikely to make the same mistakes).

Price elasticity: The ratio of the percentage change in units demanded over the percentage change in price. This is useful in marketing to determine whether a product price is too high or too low.

Product life cycle The marketing plan will differ depending on what stage the product is in: Introduction, growth, maturity, or decline.

Product/market opportunity matrix: A guide to searching for new business opportunities according to changes in new versus old products, and new versus old markets. The sources of information vary according to which of the four resulting categories one is examining, and strategies would also vary.

Protocols: An examination of people as they use a product. People are asked to talk as they learn to use a product. This provides ideas for better product design and for product instruction booklets and warnings.

Regression analysis: A widely misused method of analyzing data. Fits the line by minimizing the least square errors (but there are also programs to minimize mean absolute error). For two variables, people can often do as well with a ruler. So the big gain comes from a systematic analysis (people cannot solve such problems in their heads). It is especially useful when analyzing more than two variables where the independent variables are correlated. Estimate price elasticities with non-experimental data – providing prices differ in the data.

Role playing: See simulated interaction.

Scenarios: Written stories that describe the future. They should be written like a short story and should be in the past tense. Typically one uses a benchmark scenario (trends continue) an ideal scenario (start with the ideal and work backwards). Scenarios are misleading when it comes to forecasting, but they are highly effective in gaining commitment to future courses of action. Especially useful when large changes are contemplated (see How to Gain Acceptance for Change).

Second solution: Assume that your preferred solution is not feasible. Develop a new solution. Now compare the new solution with the old and decide which is best.

Simulated interaction: A type of role-playing in which subjects reach a decision by going through the same typed of interactions as might occur in the real situation (e.g., in meetings or by exchanging messages). You can have them "pre-enact" the situation to see what happens when various strategies are employed. For example, Lockheed Aircraft acted out how its customers (airlines) would react to alternative aircraft designs. Simulated interaction has been found to be especially useful for testing alternative approaches in situations involving conflicts among groups. See paper by Green (2002).

Stakeholder analysis: The success of a firm depends upon the cooperative efforts of various groups that make contribution (e.g., stockholders, employees, creditors, customers, suppliers). When making large changes, such as proposing a new product, be sure to consider the opinions from each group and decide what would be a satisfactory reward for each group.

Structured analogies:  Experts are asked to think of situations similar to a current situation. The decisions made in those analogous situations are summarized and used to forecast for the current situation. See Green and Armstrong working paper.

Survey research: Useful for assessing customer needs and also to assess expert opinion about how customers will react to an ad. Extensive research has led to effective procedures - the best summary of these is in Dillman, Don, Mail and Internet Surveys. It is easy to look at a survey and tell whether the creators have used the research in designing it. For more, see surveys.

SWOT (Strenths, Weaknesses, Opportunities, and Threats). Do not use SWOT. There is no evidence that it works. To the contrary, it has detrimental side effects.

Systems approach: Look first at objectives, then at alternative strategies. In each case, start at the highest conceptual level, then make the objectives and strategies more operational. It sounds easy, but it requires much time and discipline. Enables people to go outside of current solutions and to produce rational yet often very surprising solutions. I sometimes get calls from former students telling me how they used this procedure to come up with unusual and effective solutions that differed greatly from the existing thinking in their company. For details, see Systems Approach.

Time line: The process of allocating one's time in accordance with the tasks they must accomplish. This method allows one to break down their jobs into small tasks. One of the essential components of a time budget is slack time. With slack time, people can deal with the inevitable setbacks without upsetting their schedules. PERT (Program Evaluation and Review Technique) can be used to create a time budget. CPM, the Critical Path Method, is another method. Do the time line, show relationships, and estimate times. Then figure what path is determining the time to completion. You might then consider changes to this critical path to shorten the time or to introduce slack. (See Time Lines.)

Virtual group: A group in which members work on a common problem, yet they do not meet face-to-face. They might interact in other ways, such as via telephone, email, or websites - or they might not interact at all. Delphi is a form of a virtual group that allows for some interaction. Markets are another form, but they do not provide for interaction among group members. Virtual groups avoid many of the problems of groups, such as "group-think." Thus, they use information more effectively and they save time. In face-to-face groups, influence depends on such irrelevant things as how much people talk (people who talk more do not necessarily know more) and gender (males talk much more than females). In virtual groups, influence depends more on performance. Finally, because of the need to write, virtual groups leave a paper trail that aids in communication.