LO 8-1 Identify the reason for conducting marketing research. 100 words
LO 8-2 Describe the five-step marketing research approach that leads to marketing actions. 100 words

LO 8-3 Explain how marketing uses secondary and primary data. 100 words
LO 8-4 Discuss the uses of observations, questionnaires, panels, experiments, and newer data collection methods. 100 words
LO 8-5 Explain how information technology and data mining lead to marketing actions. 100 words
LO 8-6 Describe three approaches to developing a company’s sales forecast. 100 words

The Hunger Games, Star Trek, and The Hobbit are movies that rewarded their studios with huge profits.1 Unfortunately, not every movie has such favorable results. So what can these studios do to try to reduce the costly risk that a movie will be a box-office flop?
What’s in a Movie Name?
Fixing bad names for movies—like The Surrogate, Rope Burns, and Shoeless Joe—can turn potential disasters into hugely successful blockbusters. Don’t remember seeing these movies? Well, test screenings—a form of marketing research—found that moviegoers like you had problems with these original titles. Here’s what happened:
•The 2012 movie The Surrogate, which garnered Helen Hunt an Academy Award nomination, became The Sessions because a rival studio controlled the rights to the name The Surrogate.
Rope Burns became Million Dollar Baby because audiences didn’t like the original name. The movie won the 2005 Academy Award for Best Picture and starred Hilary Swank as a woman boxer and Clint Eastwood as her trainer.
•Shown frequently on television now, Shoeless Joe became the baseball classic Field of Dreams because audiences thought Kevin Costner was playing a homeless person.
To reduce risks to the studio, filmmakers want movie titles that are concise, grab attention, capture the essence of the film, and have no legal restrictions—the same factors that make a good brand name.

QR 8-1
The Hunger Games Movie Trailer

The Risks of Today’s (and Tomorrow’s) Blockbuster Movies
Bad titles, poor scripts, temperamental stars, costly special effects, and several blockbuster movies released at the same time are just a few of the nightmares studio executives face. They try to reduce their risks by developing appealing sequels, like Iron Man 3, Transformers 4, and Despicable Me 2. And, you guessed it: The Hunger Games will have sequels in 2013 (Catching Fire), 2014 (Mockingjay, Part 1), and 2015 (Mockingjay, Part 2).

Page 193

With a typical film costing over $100 million to produce and market, studios also try to reduce their risks by:
Conducting test screenings. In test screenings, 300 to 400 prospective moviegoers are recruited to attend a “sneak preview” of a film before its release. After viewing the movie, the audience fills out an exhaustive survey to critique its title, plot, characters, music, and ending to identify improvements to make in the final edit.
Using tracking studies. Immediately before an upcoming film’s release, studios will ask prospective moviegoers in the target audience three key questions: (1) Are you aware of the film? (2) Are you interested in seeing the film? and (3) Will you see the film this weekend? Studios use these data to forecast the movie’s opening weekend box-office sales and, if necessary, run last-minute ads to promote the film.

Page 194

Converting Marketing Research Results into Actions
Sometimes marketing research can mean the difference between having a blockbuster hit and having no movie at all. Consider the fact that director James Cameron had to produce a short “mini” test screening of a 3D segment of Avatar to convince four Twentieth Century Fox executives to fund his movie. The studio made the right choice; Avatar became #1 in all-time worldwide gross ticket sales.

The text describes how its “mini” test screening brought Avatar to life.

Test screenings resulted in the now-classic Fatal Attraction movie having one of the most successful “ending switches” of all time. In these sneak previews, audiences liked everything but the ending, which had Alex (Glenn Close) committing suicide and framing Dan (Michael Douglas) as her murderer. The studio shot $1.3 million of scenes for a new ending, which led to Fatal Attraction’s box-office success.
These examples show how marketing research leads to decisive marketing actions, the main topic of this chapter. Also, marketing research is often used to help a firm develop its sales forecasts, the final topic of this chapter.
Let’s (1) look at what marketing research is, (2) identify some difficulties with it, and (3) describe the five steps marketers use to conduct it.

LO 8-1Identify the reason for conducting marketing research.

What Is Marketing Research?
Marketing research is the process of defining a marketing problem and opportunity, systematically collecting and analyzing information, and recommending actions. Although imperfect, marketers conduct marketing research to reduce the risk of and thereby improve marketing decisions.
The Challenges in Doing Good Marketing Research
Whatever the marketing issue involved—whether discovering consumer tastes or setting the right price—good marketing research is challenging. For example:
•Suppose your firm is developing a new product never before seen by consumers. Will consumers really know whether they are likely to buy a product that they have never thought about before?
•Imagine if you, as a consumer, were asked about your personal hygiene habits. Even though you know the answers, will you reveal them? When personal or status questions are involved, will people give honest answers?
•Will consumers’ actual purchase behaviors match their stated interests or intentions? Will they buy the same brand they say they will?
Marketing research must overcome these difficulties and obtain the information needed so that marketers can assess what consumers want and will buy.

Page 195


Five-step marketing research approach leading to marketing actions. Lessons learned from past research mistakes are fed back to improve each of the steps.


Five-Step Marketing Research Approach
decision is a conscious choice from among two or more alternatives. All of us make many such decisions daily. At work we choose from alternative ways to accomplish an assigned task. At college we choose from alternative courses. As consumers we choose from alternative brands. No magic formula guarantees correct decisions.
Managers and researchers have tried to improve the outcomes of decisions by using more formal, structured approaches to decision making, the act of consciously choosing from among alternatives. The systematic marketing research approach used to collect information to improve marketing decisions and actions described in this chapter uses five steps and is shown in Figure 8–1. Although the five-step approach described here focuses on marketing decisions, it provides a systematic checklist for making both business and personal decisions.

LO 8-2Describe the five-step marketing research approach that leads to marketing actions.

Every marketing problem faces its own research challenges. For example, the marketing strategy used by LEGO’s toy researchers and designers in Denmark illustrates the wide variations possible in collecting marketing research data to build better toys.

LEGO Mindstorms EV3: From TRACK3R, whose interchangeable bazooka and hammer can operate after only 20 minutes of assembly …

LEGO’s definition of “toy” has changed dramatically in the last half century—far beyond the 40 billion LEGO bricks sitting in rooms around the world today. In 1998, LEGO introduced the Mindstorms kit, a “toy” that integrates electronics, computers, and robots with LEGO bricks. Developed with the help of the Media Lab at the Massachusetts Institute of Technology, the Mindstorms kit appeals to a diverse market—from elementary school kids to world-class robotics experts. The kits can be found in homes, schools, universities, and industrial laboratories.
A simplified look at the marketing research for the LEGO Mindstorms EV3—introduced in late 2013 at a price of $350—shows the two key elements in defining a problem: setting the research objectives and identifying possible marketing actions.
Set the Research Objectives
Research objectives are specific, measurable goals the decision maker, in this case a LEGO manager, seeks to achieve in conducting the marketing research. For LEGO, let’s assume the immediate research objective is to decide which of two new Mindstorms designs should be selected for marketing.

Page 196


… to EV3RSTORM, whose infrared sensors let it walk or skate.

In setting research objectives, marketers have to be clear on the purpose of the research that leads to marketing actions. The three main types of marketing research, explained in more detail later in the chapter, are as follows:
1.Exploratory research provides ideas about a vague problem. LEGO was concerned that middle school kids would be overwhelmed by the 500-plus pieces in Mindstorms kits and quickly lose interest. LEGO’s brainstorming—an example of exploratory research—revealed kids need to have a basic device up, running, and doing tricks in 20 minutes.
2.Descriptive research generally involves trying to find the frequency with which something occurs or the extent of a relationship between two factors. So if LEGO wants to know which of the two Mindstorms kits is of greatest interest to middle school versus high school students, it might ask them. LEGO can then assess the relationship by doing a cross tabulation (discussed later in the chapter) of school level versus kit preference.
3.Causal research tries to determine the extent to which the change in one factor changes another one. Changing key pieces in a Mindstorms kit affects how quickly the newly built device can do tricks—affecting acceptance by kit users. Test markets, discussed later, use causal research.
Identify Possible Marketing Actions
Effective decision makers develop specific measures of success, which are criteria or standards used in evaluating proposed solutions to the problem. Different research outcomes, based on the measure of success, lead to different marketing actions. For LEGO, assume the measure of success is the total time middle schoolers spend playing with each of the two Mindstorms kits until they produce a device that can do simple tricks. This measure of success leads to a clear-cut marketing action: Market the design that produces an acceptable device in the least amount of playing time.
Marketing researchers know that defining a problem is an incredibly difficult task. If the objectives are too broad, the problem may not be researchable. If they are too narrow, the value of the research results may be seriously lessened. This is why marketing researchers spend so much time defining a marketing problem precisely and writing a formal proposal that describes the research to be done.
The second step in the marketing research process requires that the researcher (1) specify the constraints on the marketing research activity, (2) identify the data needed for marketing actions, and (3) determine how to collect the data.
Specify Constraints
The constraints in a decision are the restrictions placed on potential solutions to a problem. Examples include the limitations on the time and money available to solve the problem.
What constraints might LEGO set in mid-2012 in developing the LEGO Mindstorms EV3, whose market launch is scheduled for late 2013? LEGO might establish the following constraints on its decision to select one of the two improved designs: The decision (1) must be made in five weeks (2) using 10 teams of middle schoolers playing with the two improved Mindstorms kits.

Page 197

Identify Data Needed for Marketing Actions
Effective marketing research studies focus on collecting data that will lead to effective marketing actions. In the Mindstorms case, LEGO’s marketers might want to know students’ math skills, time spent playing video games, and so on. But that information, while nice to know, is largely irrelevant because the study should focus on collecting only those data that will help them make a clear choice between the two Mindstorms designs.
Determine How to Collect Data
Determining how to collect useful marketing research data is often as important as actually collecting the data—Step 3 in the process, which is discussed later. Two key elements in deciding how to collect the data are (1) concepts and (2) methods.
Concepts    In the world of marketing, concepts are ideas about products or services. To find out about consumer reactions to a potential new product, marketing researchers frequently develop a new-product concept, which is a picture or verbal description of a product or service the firm might offer for sale. For example, the LEGO designers might develop a new-product concept for an innovative Mindstorms EV3 robot, like that shown, capable of walking or skating, perhaps by adding an infrared sensor.
Methods    Methods are the approaches that can be used to collect data to solve all or part of a problem. To collect data, the LEGO designers might use a combination of (1) observing the behavior of Mindstorms users and (2) asking them questions to meet its “20-minute” measure of success for young users. Observing people and asking them questions—the two main data collection methods—are discussed in the section that follows.
How successful is LEGO’s marketing research and design strategy for its Mindstorms robots? Among younger users alone, in 2012 more than 20,000 elementary and middle school teams faced off in competitions around the world.
How can you find and use the methods that other marketing researchers have found successful? Information on useful methods is available in tradebooks, textbooks, and handbooks that relate to marketing and marketing research. Some periodicals and technical journals, such as the Journal of Marketing and the Journal of Marketing Research, both published by the American Marketing Association, summarize methods and techniques valuable in addressing marketing problems.
Special methods vital to marketing are (1) sampling and (2) statistical inference. For example, marketing researchers often use sampling by selecting a group of distributors, customers, or prospects, asking them questions, and treating their answers as typical of all those in whom they are interested. They may then use statistical inference to generalize the results from the sample to much larger groups of distributors, customers, or prospects to help decide on marketing actions.

learning review
8-1.What is marketing research?
8-2.What is the five-step marketing research approach?
8-3.What are constraints, as they apply to developing a research plan?
Page 198


LO 8-3Explain how marketing uses secondary and primary data.

Collecting enough relevant information to make a rational, informed marketing decision sometimes simply means using your knowledge to decide immediately. At other times it entails collecting an enormous amount of information at great expense.
Figure 8–2 shows how the different kinds of marketing information fit together. Data, the facts and figures related to the project, are divided into two main parts: secondary data and primary data. Secondary data are facts and figures that have already been recorded prior to the project at hand. As shown in Figure 8–2, secondary data are divided into two parts—internal and external secondary data—depending on whether the data come from inside or outside the organization needing the research. Primary data are facts and figures that are newly collected for the project. Figure 8–2 shows that primary data can be divided into observational data, questionnaire data, and other sources of data.
Secondary Data: Internal
The internal records of a company generally offer the most easily accessible marketing information. These internal sources of secondary data may be divided into two related parts: (1) marketing inputs and (2) marketing outcomes.
Marketing input data relate to the effort expended to make sales. These range from sales and advertising budgets and expenditures to salespeople’s call reports, which describe the number of sales calls per day, who was visited, and what was discussed.
Marketing outcome data relate to the results of the marketing efforts. These involve accounting records on shipments and include sales and repeat sales, often broken down by sales representative, industry, and geographic region. In addition, e-mails, phone calls, and letters from customers can reveal both complaints and what is working well.

Types of marketing information. Researchers must choose carefully among these to get the best results, considering time and cost constraints.


Page 199

Secondary Data: External
Published data from outside the organization are external secondary data. The U.S. Census Bureau publishes a variety of useful reports. Best known is the Census 2010, which is the most recent count of the U.S. population that occurs every 10 years. Recently, the Census Bureau began collecting data annually from a smaller number of people through the American Community Survey. Both surveys contain detailed information on American households, such as the number of people per household and the age, sex, race/ethnic background, income, occupation, and education of individuals within the household. Marketers use these data to identify characteristics and trends of ultimate consumers.

Scanner data at supermarket checkout counters provide valuable information for marketing decisions.

The Census Bureau also publishes the Economic Census, which is conducted every five years. These reports are vital to business firms selling products and services to organizations. The 2012 Economic Census contains data on the number and sales of establishments in the United States that produce a product or service based on each firm’s geography (state, county, zip code, etc.), industry sector (manufacturing, retail trade, etc.), and North American Industry Classification System (NAICS) code. Data for the 2012 Economic Census will be released beginning in 2014.

QR 8-2
Census 2010 Video

Several market research companies pay households and businesses to record all their purchases using a paper or electronic diary. Such syndicated panel data economically answer questions that require consistent data collection over time, such as, “How many times did our customers buy our products this year compared to last year?” Examples of syndicated panels that provide a standard set of data on a regular basis are the Nielsen TV ratings and J.D. Power’s automotive quality and customer satisfaction surveys.
Some data services provide comprehensive information on household demographics and lifestyle, product purchases, TV viewing behavior, and responses to coupon and free-sample promotions. Their advantage is that a single firm can collect, analyze, interrelate, and present all this information. For consumer product firms such as Procter & Gamble, sales data from various channels are critical to allocate scarce marketing resources. As a result, they use tracking services such as IRI’s InfoScan to collect product sales and coupon/free-sample redemptions that have been scanned at the checkout counters of supermarket, drug, convenience, and mass merchandise retailers.
Finally, trade associations, universities, and business periodicals provide detailed data of value to market researchers and planners. These data are now available online via the Internet and can be identified and located using a search engine such as Google or Bing. The Marketing inSite box on the next page provides examples.
Advantages and Disadvantages of Secondary Data
A general rule among marketing people is to obtain secondary data first and then collect primary data. Two important advantages of secondary data are (1) the tremendous time savings because the data have already been collected and published or exist internally and (2) the low cost, such as free or inexpensive Census reports. Furthermore, a greater level of detail is often available through secondary data, especially U.S. Census Bureau data.
However, these advantages must be weighed against some significant disadvantages. First, the secondary data may be out of date, especially if they are U.S. Census data collected only every 5 or 10 years. Second, the definitions or categories might not be quite right for a researcher’s project. For example, the age groupings or product categories might be wrong for the project. Also, because the data have been collected for another purpose, they may not be specific enough for the project. In such cases, it may be necessary to collect primary data.

Page 200


Marketing inSite

Online Databases and Internet Resources Useful to Marketers
Marketers today in search of secondary data can consult online databases available via the Internet. Indexes to articles in periodicals and statistical or financial data on markets, products, and organizations can be accessed either directly or via Internet search engines or portals through keyword searches.
Statistical and financial data on markets, products, and organizations include:
The Wall Street Journal (, CNBC (, and Fox Business (, which provide up-to-the-minute business news and video clips about companies, industries, and trends that affect the marketing environment.
•STAT-USA ( and the Census Bureau ( of the U.S. Department of Commerce, which provide information on U.S. business, economic, and trade activity collected by the federal government.
Portals and search engines include:
• (, the portal to all U.S. government websites. Users can click on links to browse by topic or enter keywords for specific searches.
•Google (, the most popular portal to the entire Internet. Users enter keywords for specific searches and then click on results of interest.
Some of these websites are accessible only if you or your educational institution have paid a subscription fee. Check with your institution’s website.


learning review
8-4.What is the difference between secondary and primary data?
8-5.What are some advantages and disadvantages of secondary data?
LO 8-4Discuss the uses of observations, questionnaires, panels, experiments, and newer data collection methods.

Primary Data: Watching People
Observing people and asking them questions are the two principal ways to collect new or primary data for a marketing study. Facts and figures obtained by watching, either mechanically or in person, how people actually behave is the way marketing researchers collect observational data. Observational data can be collected by mechanical (including electronic), personal, or neuromarketing methods.
Mechanical Methods    National TV ratings, such as those of Nielsen shown in Figure 8–3, are an example of mechanical observational data collected by a “people meter.” The device measures what channel and program is tuned in and who is watching. The people meter (1) is a box that is attached to TV sets, VCRs, DVRs (digital video recorders), cable boxes, and satellite dishes in about 20,000 households across the country; (2) has a remote that operates the meter when a viewer begins and finishes watching a TV program; and (3) stores and then transmits the viewing information each night to Nielsen. Data are also collected on TV viewing using less sophisticated meters or TV diaries (a paper-pencil measurement system).

What determines if American Idol stays on the air? For the importance of the TV “ratings game,” see the text.

In early 2012, Nielsen introduced “cross-platform television ratings,” which combine Nielsen’s existing TV ratings with its new online ratings. These ratings include consumer viewing of TV programs seen not only on regular TV but also streamed online or wirelessly on PCs, smartphones, tablet devices, and video game consoles.
On the basis of all these observational data, Nielsen then calculates the rating and share of each TV program. With 115.6 million TV households in the United States, a single rating point equals 1 percent, or 1,156,000 TV households. For TV viewing, a share point is the percentage of TV sets in use tuned to a particular program. Because TV and cable networks sell over $64 billion annually in advertising and set advertising rates to advertisers on the basis of those data, precision in the Nielsen data is critical.

Page 201


Nielsen Television Index Ranking Report for network TV prime-time households for the week ending May 19, 2013. The difference of a few share points in Nielsen TV ratings affects the cost of a TV ad on a show and even whether the show remains on the air.

Source: Nielsen © 2013. The Nielsen Company Broadcast Ranking Report for the week ending May 19, 2013. Viewing estimates include live viewing and DVR playback on the same day, defined as 3 A.M. to 3 A.M. Rank is based on the number of U.S. viewers (in millions) from Nielsen’s National People Meter Sample.

A change of 1 percentage point in a rating can mean gaining or losing millions of dollars in advertising revenues because advertisers pay rates on the basis of the size of the audience for a TV program. So as shown by the green rows in Figure 8–3, we might expect to pay more for a 30-second TV ad on NCIS than one on Castle. Broadcast and cable networks may change the time slot or even cancel a TV program if its ratings are consistently poor and advertisers are unwilling to pay a rate based on a higher guaranteed rating.
But TV advertisers today have a special problem: With about three out of four TV viewers skipping ads with TiVo and DVRs, or channel surfing during commercials, how many people are actually seeing their TV ads? Now services such as Nielsen offer advertisers minute-by-minute measurement of how many viewers stay tuned during commercials. Recently, TiVo also expanded its service to allow TV advertisers to see how many and what kind of users are watching their commercials. The viewership data in Figure 8–3 include not only live TV but also programs recorded on DVRs. With these more precise measures of who is likely to see a TV ad, buying TV ads is becoming more scientific.

Is this really marketing research? A mystery shopper at work.

Personal Methods    Observational data can take some strange twists. Jennifer Voitle, a laid-off investment bank employee with four advanced degrees, responded to an Internet ad and found a new career: mystery shopper. Companies pay mystery shoppers to check on the quality and pricing of their products and the integrity of and customer service provided by their employees. Jennifer gets paid to travel to exotic hotels, eat at restaurants, play golf, test-drive new cars, shop for clothes, and play arcade games. But her role posing as a customer gives her client unique marketing research information that can be obtained in no other way. Says Jennifer, “Can you believe they call this work?”

Page 202

Watching consumers in person or recording them are two other observational approaches. For example, Procter & Gamble watches women do their laundry, clean the floor, put on makeup, and so on because they comprise 80 percent of its customers! And Gillette records consumers brushing their teeth in their own bathrooms to see how they really brush—not just how they say they brush. The new-product result: Gillette’s Oral-B CrossAction toothbrush.
Ethnographic research is a specialized observational approach in which trained observers seek to discover subtle behavioral and emotional reactions as consumers encounter products in their “natural use environment,” such as in their home or car. Recently, Kraft launched Deli Creations, which are sandwiches made with its Oscar Mayer meats, Kraft cheeses, and Grey Poupon mustard, after spending several months with consumers in their kitchens. Kraft discovered that consumers wanted complete, ready-to-serve meals that are easy to prepare—and it had the products to create them.
Personal observation is both useful and flexible, but it can be costly and unreliable when different observers report different conclusions when watching the same event. And while observation can reveal what people do, it cannot easily determine why they do it. This is the principal reason for using neuromarketing and questionnaires, our next topics.
Neuromarketing Methods    Global brand expert Martin Lindstrom believes that most traditional marketing research—like focus groups and surveys—is wasted because consumers’ feelings toward products and brands reside deep within the subconscious part of their brains. Lindstrom used brain scanning to analyze the buying processes of more than 2,000 participants. Lindstrom merged neuroscience—the study of the brain—with marketing! His controversial findings using “neuromarketing” are summarized in his breakthrough book Buyology.

“Neuromarketing” often uses a cap with dozens of sensors to measure brain waves to try to understand consumers better. For some changes made by Campbell Soup Company based on neuromarketing, see the text.

Based on the results of neuromarketing studies, Campbell Soup Company recently changed the labels of most of its soup cans. Some of the changes: Steam now rises from more vibrant images of soup; the “unemotional spoons” have disappeared; and the script logo is smaller and has been moved to the bottom of the can.
Primary Data: Asking People
How many dozens of times have you filled out some kind of a questionnaire? Maybe a short survey at school or a telephone or e-mail survey to see if you are pleased with the service you received. Asking consumers questions and recording their answers is the second principal way of gathering information.
We can divide this primary data collection task into (1) idea generation methods and (2) idea evaluation methods, although they sometimes overlap and each has a number of special techniques. Each survey method results in valuable questionnaire data, which are facts and figures obtained by asking people about their attitudes, awareness, intentions, and behaviors.
Idea Generation Methods—Coming Up with Ideas    In the past, the most common way of collecting questionnaire data to generate ideas was through an individual interview, which involves a single researcher asking questions of one respondent. This approach has many advantages, such as being able to probe for additional ideas using follow-up questions to a respondent’s initial answers. However, this method is very expensive. Later in the chapter we’ll discuss some alternatives.

Page 203


Focus groups of students and instructors were used in developing this textbook. To see the specific suggestion that may help you study, read the text.

General Mills sought ideas about why Hamburger Helper didn’t fare well when it was introduced. Initial instructions called for cooking a half-pound of hamburger separately from the noodles or potatoes, which were later mixed with the hamburger. So General Mills researchers used a special kind of individual interview, called a depth interview, in which researchers ask lengthy, free-flowing kinds of questions to probe for underlying ideas and feelings. These depth interviews showed consumers (1) didn’t think it contained enough meat and (2) didn’t want the hassle of cooking in two different pots. The Hamburger Helper product manager changed the recipe to call for a full pound of meat and to allow users to prepare it in one dish, leading to product success.
Focus groups are informal sessions of 6 to 10 past, present, or prospective customers in which a discussion leader, or moderator, asks for opinions about the firm’s products and those of its competitors, including how they use these products and special needs they have that these products don’t address. Often recorded and conducted in special interviewing rooms with a one-way mirror, these groups enable marketing researchers and managers to hear and watch consumer reactions.

QR 8-3
Trend Hunter Video

The informality and peer support in an effective focus group help uncover ideas that are often difficult to obtain with individual interviews. For example, to improve understanding and learning by students using this textbook, focus groups were conducted among both marketing instructors and students. Both groups recommended providing answers to each chapter’s set of Learning Review questions. This suggestion was followed, so you can see the answers by turning to a section at the back of the textbook or by tapping your finger on the question in the textbook’s iPad version.

Wendy’s spent over two years remaking its 42-year-old burger. The result: Dave’s Hot ’N Juicy, named after Wendy’s founder, Dave Thomas. See Figure 8–4 for some questions that Wendy’s asked consumers in a survey to discover their fast-food preferences, behaviors, and demographics.

Finding “the next big thing” for consumers has caused marketing researchers to turn to some less traditional techniques. For example, “fuzzy front end” methods attempt early identification of elusive consumer tastes or trends. Trend Hunter is a firm that seeks to anticipate and track “the evolution of cool.” Trend hunting (or watching) is the practice of identifying “emerging shifts in social behavior,” which are driven by changes in pop culture that can lead to new products. Trend Hunter has identified about 200,000 “micro-trends” through its global network of 118,000 members and features several of these trends on its daily Trend Hunter TV broadcast via its YouTube channel (trendhuntertv).
Idea Evaluation Methods—Testing an Idea    In idea evaluation, the marketing researcher tries to test ideas discovered earlier to help the marketing manager recommend marketing actions. Idea evaluation methods often involve conventional questionnaires using personal, mail, telephone, fax, and online (e-mail or Internet) surveys of a large sample of past, present, or prospective consumers. In choosing among them, the marketing researcher balances the cost of the particular method against the expected quality of the information and the speed with which it can be obtained.
Personal interview surveys enable the interviewer to be flexible in asking probing questions or getting reactions to visual materials but are very costly. Mail surveys are usually biased because those most likely to respond have had especially positive or negative experiences with the product or brand. While telephone interviews allow flexibility, unhappy respondents may hang up on the interviewer, even with the efficiency of computer-assisted telephone interviewing (CATI).
Increasingly, marketing researchers have begun to use online surveys (e-mail and Internet) to collect primary data. The reason: Most consumers have an Internet connection and an e-mail account. Marketers can embed a survey in an e-mail sent to targeted respondents. When they open the e-mail, consumers can either see the survey or click on a link to access it from a website. Marketers can also ask consumers to complete a “pop up” survey in a separate browser window when they access an organization’s website. Many organizations use this method to have consumers assess their products and services or evaluate the design and usability of their websites.

Page 204


To obtain the most valuable information from consumers, this Wendy’s survey utilizes four different kinds of questions discussed in the text.


The advantages of online surveys are that the cost is relatively minimal and the turnaround time from data collection to report presentation is much quicker than the traditional methods discussed earlier. However, online surveys have serious drawbacks: Some consumers may view e-mail surveys as “junk” or “spam” and may either choose to not receive them (if they have a “spam blocker”) or purposely or inadvertently delete them, unopened. For Internet surveys, some consumers have a “pop-up blocker” that prohibits a browser from opening a separate window that contains the survey; thus, they may not be able to participate in the research. For both e-mail and Internet surveys, consumers can complete the survey multiple times, creating a significant bias in the results. This is especially true for online panels. In response, research firms such as SurveyMonkey have developed sampling technology to prohibit this practice.
The foundation of all research using questionnaires is developing precise questions that get clear, unambiguous answers from respondents. Figure 8–4 shows a number of formats for questions taken from a Wendy’s survey that assessed fast-food restaurant preferences among present and prospective consumers.

Page 205

Question 1 is an example of an open-ended question, which allows respondents to express opinions, ideas, or behaviors in their own words without being forced to choose among alternatives that have been predetermined by a marketing researcher. This information is invaluable to marketers because it captures the “voice” of respondents, which is useful in understanding consumer behavior, identifying product benefits, or developing advertising messages.
In contrast, closed-end or fixed alternative questions require respondents to select one or more response options from a set of predetermined choices. Question 2 is an example of a dichotomous question, the simplest form of a fixed alternative question that allows only a “yes” or “no” response.
A fixed alternative question with three or more choices uses a scale. Question 5 is an example of a question that uses a semantic differential scale, a five-point scale in which the opposite ends have one- or two-word adjectives that have opposite meanings. For example, depending on the respondent’s opinion regarding the cleanliness of Wendy’s restaurants, he or she would check the left-hand space on the scale, the right-hand space, or one of the three other intervening points. Question 6 uses a Likert scale, in which the respondent indicates the extent to which he or she agrees or disagrees with a statement.

Page 206


Visitors to Frito-Lay’s Facebook Page voted on new potato chip flavors by clicking on “I’d Eat That” to show their preferences.

The questionnaire in Figure 8–4 provides valuable information to the marketing researcher at Wendy’s. Questions 1 to 8 inform him or her about the respondent’s likes and dislikes in eating out, frequency of eating out at fast-food restaurants generally and at Wendy’s specifically, and sources of information used in making decisions about fast-food restaurants. Question 9 gives details about the respondent’s personal or household characteristics, which can be used in trying to segment the fast-food market, a topic discussed in Chapter 9.
Marketing research questions must be worded precisely so that all respondents interpret the same question similarly. For example, in a question asking whether you eat at fast-food restaurants regularly, the word regularly is ambiguous. Two people might answer “yes” to the question, but one might mean “once a day” while the other means “once or twice a month.” However, each of these interpretations suggests that dramatically different marketing actions be directed to these two prospective consumers.
The high cost of using personal interviews in homes has increased the use of mall intercept interviews, which are personal interviews of consumers visiting shopping centers. These face-to-face interviews reduce the cost of personal visits to consumers in their homes while providing the flexibility to show respondents visual cues such as ads or actual product samples. A disadvantage of mall intercept interviews is that the people interviewed may not be representative of the consumers targeted, giving a biased result.
Electronic technology has revolutionized traditional concepts of interviews or surveys. Today, respondents can walk up to a kiosk in a shopping center, read questions off a screen, and key their answers into a computer on a touch screen. Fully automated telephone interviews exist in which respondents key their replies on a touch-tone telephone.
Primary Data: Other Sources
Four other methods of collecting primary data exist that overlap somewhat with the methods just discussed. These involve using (1) social media, (2) panels and experiments, (3) information technology, and (4) data mining.
Social Media    Facebook, Twitter, and other social media are revolutionizing the way today’s marketing research is done. In developing a new potato chip flavor, Frito-Lay substituted Facebook research for its usual focus groups. Visitors to its Facebook Page were polled, allowing them to suggest new flavors, three of which appeared in 2013. All they had to do was click an “I’d Eat That” button to show their preferences. And Estée Lauder asked social media users to vote on which discontinued shades to bring back.
Carma Laboratories, Inc., the maker of Carmex lip balm, is a third generation, family-owned business with a history of accessibility to customers. In fact, founder Alfred Woelbing personally responded to every letter he received from customers. Today, Carma Labs relies on social media programs to help promote its products.

Carmex’s home page topics range from its Twitter Giveaways to win Carmex products to its family history. For how it uses social media marketing research, see the Using Marketing Dashboards box.
Page 207


Are the Carmex Social Media Programs Working Well?
As a marketing consultant to Carmex, you’ve just been asked to assess its social media activities for its lip balm product line.
Carmex has recently launched new social media programs and promotions to tell U.S. consumers more about its line of lip balm products. These include Facebook and Twitter contests that allow Carmex fans and followers to win free samples by connecting with Carmex. A creative “Carmex Kiss” widget allows users to upload their photo and to send an animated kiss to a friend.
Your Challenge    To assess how the Carmex social media programs are doing, you choose these five metrics: (1) Carmex conversation velocity—total Carmex mentions on the Internet; (2) Facebook fans—the number of Facebook users in a time period who have liked Carmex’s Facebook brand page; (3) Twitter followers—the number of Twitter users in a time period who follow Carmex’s Twitter feed; (4) Carmex share of voice—Carmex mentions on the Internet as a percentage of mentions of all major lip balm brands; and (5) Carmex sentiment—the percentage of Internet Carmex share-of-voice mentions that are (a) positive, (b) neutral, or (c) negative.
Your Findings    Analyzing the marketing dashboard here, you reach these conclusions. First, the number of both Facebook fans and Twitter followers for Carmex is up significantly for 2013 compared to 2012, which is good news. Second, the Carmex share of voice of 35 percent is good, certainly relative to the 48 percent for the #1 brand ChapStick. But especially favorable is Carmex’s 12 percent increase in share of voice compared to a year ago. Third, the Carmex sentiment dashboard shows 80 percent of the mentions are positive, and only 5 percent are negative. Even more significant is that positive mentions are up 23 percent over last year.
Your Actions    You conclude that Carmex’s social media initiatives are doing well. Your next step is to probe deeper into the data to see which ones—such as free samples or the Carmex Kiss—have been especially effective in triggering the positive results and build on these successes in the future.

Carmex lip balm is meant to reduce cold sore symptoms and soothe dry and chapped lips. It is packaged in jars, sticks, and squeezable tubes. The U.S. Carmex product line includes original, strawberry, lime twist, vanilla, pomegranate, and cherry flavors. Although Carmex lip balm sales trend behind ChapStick and Blistex, Carmex consumers tend to be loyalists—true zealots.
One opportunity for Carmex ( is to conduct marketing research using social media listening tools to understand the nature of online lip balm conversations. Lip balm is a seasonal product, with both sales and online activity peaking during the cough–cold season of November through March.
The Using Marketing Dashboards box shows how Carmex uses marketing metrics to assess its social media programs for its line of products. Data have been modified to protect proprietary information.

Page 208


To discover how Walmart used test markets to help develop its internationally successful supercenters, such as this one in China, see the text.

Carmex uses several social media metrics, such as conversation velocity, share of voice, and sentiment. These metrics are tracked by electronic search engines that comb the Internet for consumers’ behaviors and “brand mentions” to calculate share of voice and determine whether these brand mentions appear to be “positive,” “neutral,” or “negative” in order to calculate “sentiment.” A widely used Facebook metric measures the number of likes, which refers to the number of Facebook users opting in to a brand’s messages and liking the brand.
Marketing researchers increasingly want to glean information from sites to “mine” their raw consumer-generated content in real time. However, when relying on this consumer-generated content, the sample of individuals from whom this content is gleaned may not be statistically representative of the marketplace.
Panels and Experiments    Two special ways that observations and questionnaires are sometimes used are panels and experiments.
Marketing researchers often want to know if consumers change their behavior over time, so they take successive measurements of the same people. A panel is a sample of consumers or stores from which researchers take a series of measurements. For example, the NPD Group collects data about consumer purchases such as apparel, food, and electronics from its Online Panel, which consists of nearly 2 million individuals worldwide. So a firm like General Mills can count the frequency of consumer purchases to measure switching behavior from one brand of its breakfast cereal (Wheaties) to another (Cheerios) or to a competitor’s brand (Kellogg’s Special K). A disadvantage of panels is that the marketing research firm needs to recruit new members continually to replace those who drop out. These new recruits must match the characteristics of those they replace to keep the panel representative of the marketplace.
An experiment involves obtaining data by manipulating factors under tightly controlled conditions to test cause and effect. The interest is in whether changing one of the independent variables (a cause) will change the behavior of the dependent variable that is studied (the result). In marketing experiments, the independent variables of interest—sometimes called the marketing drivers—are often one or more of the marketing mix elements, such as a product’s features, price, or promotion (like advertising messages or coupons). The ideal dependent variable usually is a change in the purchases (incremental unit or dollar sales) of individuals, households, or organizations. For example, food companies often use test markets, which offer a product for sale in a small geographic area to help evaluate potential marketing actions. In 1988, Walmart opened three experimental stand-alone supercenters to gauge consumer acceptance before deciding to open others. Today, Walmart operates over 4,000 supercenters around the world.
A potential difficulty with experiments is that outside factors (such as actions of competitors) can distort the results of an experiment and affect the dependent variable (such as sales). A researcher’s task is to identify the effect of the marketing variable of interest on the dependent variable when the effects of outside factors in an experiment might hide it.
Information Technology    Information technology involves operating computer networks that can store and process data. Today, information technology can extract hidden information from large databases, such as those containing retail sales collected through barcode scanners at checkout counters and households’ product purchases and TV viewing behavior.
Figure 8–5 shows how marketers use information technology, data, models, and queries to obtain results that lead to marketing actions. Today’s marketing managers can be drowned in an ocean of data; they need to adopt strategies for dealing with complex, changing views of the competition, the market, and the consumer. The Internet and PC help make sense out of this vast amount of information. The marketer’s task is to convert it into useful information that will lead to marketing actions.

Page 209


How marketing researchers and managers use information technology to turn information into action.


As shown at the top of Figure 8–5, marketers use information technology that consists of computers linked together through sophisticated communication networks to access and retrieve data from internal and external sources. These data sources are stored, organized, and managed in databases. Collectively, these databases form a data warehouse.
As shown at the bottom of Figure 8–5, marketers use computers to query the databases in the data warehouse with marketing queries or questions. These questions go through statistical models that organize and manipulate the data to analyze and identify the relationships that exist. The results are then presented using tables and graphics for easier interpretation. When querying a database, marketers can use sensitivity analysis to ask “what if” questions to determine how hypothetical changes in product or brand drivers—the factors that influence the buying decisions of a household or organization—can affect sales.
Traditional marketing research typically involves identifying possible drivers and then collecting data. For example, we might collect data to test the hypothesis that increasing couponing (the driver) during spring will increase trials by first-time buyers (the result).

At 10 P.M., what is this man likely to buy besides these diapers? For the curious answer that data mining gives, see the text.

Data Mining    In contrast, data mining is the extraction of hidden predictive information from large databases to find statistical links between consumer purchasing patterns and marketing actions. Some of these are common sense: Since many consumers buy peanut butter and grape jelly together, why not run a joint promotion between Skippy peanut butter and Welch’s grape jelly? But would you have expected that men buying diapers in the evening sometimes buy a six-pack of beer as well? Supermarkets discovered this when they mined checkout data from scanners. So they placed diapers and beer near each other, then placed potato chips between them—and increased sales on all three items! On the near horizon is RFID (radio frequency identification) technology using “smart tags” on the diapers and beer to tell whether they wind up in the same shopping bag.

Page 210


Making Responsible Decisions ethics

No More Personal Secrets: The Downside of Data Mining
eXelate, Intellidyn, Rapleaf, Google Ad Preferences, Yahoo!, BlueKai, Alliance Data, … yes … and Facebook and Twitter, too!
The common denominator for all these is their sophisticated data mining of the Internet and social media that reveals an incredible amount of personal information about any American. Time journalist Joel Stein, using both online and offline sources, discovered how easily outsiders could find his social security number and then found a number of other things about himself—some correct, some not.
For example, he likes hockey, rap, rock, parenting, recipes, clothes and beauty products, and movies. He makes most of his purchases online, averaging only $25 per purchase. He uses Facebook, Friendster, LinkedIn, MySpace, Pandora, and StumbleUpon. He bought his house in November, which is when his home insurance is up for renewal. His dad’s wife has a traffic ticket.
And he uses an Apple iMac and is an 18- to 19-year-old woman???!!!
OK, OK, sometimes data mining errors occur!
These data are collected many ways from the Internet—from tracking devices (like cookies, discussed in Chapter 21) on websites to apps downloaded on a cellphone, PC, or tablet device that reveal a user’s contact list and location.
These personal details have huge benefits for marketers. Data mining enables one-to-one personalization and now enables advertisers to target individual consumers. This involves using not only demographics such as age and sex but also “likes,” past buying habits, social media used, brands bought, TV programs watched, and so on.
Want to do some sleuthing yourself? Download Ghostery at It tells you all the companies grabbing your data when you visit a website.

Data mining related to the Internet and social media is exploding. In 2012, companies were expected to spend $840 million for online data, double the 2009 amount. This increased investment in data mining also includes a greater focus on the more than $2 billion spent annually on social media advertising. And the cost to an advertiser for that one bit of special information about you? Two-fifths of a cent. For how much online data mining can reveal about you personally and the ethical issues involved, see the Making Responsible Decisions box.
Advantages and Disadvantages of Primary Data
Compared with secondary data, primary data have the advantages of being more flexible and more specific to the problem being studied. The main disadvantages are that primary data are usually far more costly and time-consuming to collect than secondary data.

Wendy’s Customer Satisfaction Survey allows customers to provide feedback on their meal experience. To take the survey, go to You’ll need a receipt from a recent visit, which has an 8-digit code to start the survey. Your reward? A printable coupon that can be used on your next visit!

Analyzing Primary Data Using Cross Tabulations
Suppose top management at Wendy’s wants to use the questionnaire in Figure 8–4 to survey a sample of U.S. households to assess how often customers of different ages eat at fast-food restaurants. Management suspects that as the age of the head of the household increases, visits to fast-food restaurants decline. The data provided by the questionnaire confirm this, but the information is not in a format that suggests ideas for viable marketing actions. Using cross tabulations will provide answers leading to actions.

Page 211


Two forms of a cross tabulation relating age of head of household to frequency of fast-food restaurant patronage.


Developing Cross Tabulations    A cross tabulation, or cross tab, is a method of presenting and analyzing data involving two or more variables to discover relationships in the data.
The Wendy’s questionnaire in Figure 8–4 includes many questions that might be paired to understand the fast-food business better. For example, to try to answer the question in which Wendy’s top management is interested, we can pair the question regarding the age of the head of the household in Figure 8–4 (question 9d) with the question that asks how often the respondent eats at a fast-food restaurant (question 3).
Using the answers to question 3 as the column headings and the answers to question 9d as the row headings gives the cross tabulation shown in Figure 8–6, based on answers from 586 respondents. The figure shows two forms of cross tabulations:
•The raw data or answers to the specific questions are shown in Figure 8–6A. For example, this cross tab shows that 144 households in the sample whose head was under 25 (shaded red) ate at fast-food restaurants once a week or more. It also shows the loyalty of many customers of fast-food restaurants; the number of customers who visit them once a week or more is more than double the number who visit them once a month or less, as indicated by the totals shaded brown in Figure 8–6A.
•Answers on a percentage basis, with the percentages running horizontally, are shown in Figure 8–6B. Of the 215 households headed by someone under 25, 67.0 percent ate at a fast-food restaurant at least once a week and only 8.8 percent ate there once a month or less. Also, across all age groups, 46.4 percent—almost half—ate in a fast-food restaurant once a week or more.
Two other forms of cross tabulation using the raw data shown in Figure 8–6A are described in problem 7 in Applying Marketing Knowledge at the end of the chapter.
Interpreting Cross Tabulations    A careful analysis of Figures 8–6A and 8–6B shows that patronage of fast-food restaurants is related to the age of the head of the household. The percentages on the diagonal (in orange) in Figure 8–6B reveal that younger households are far more likely than older households to visit fast-food restaurants once a week or more.

Page 212

So if we want to reach frequent users of fast-food restaurants, we should target those whose head of household is under 25 years of age and who tend to visit these restaurants once a week or more, as shown in Figure 8–6B. Marketers often use special efforts to reach these loyal, frequent users. So Wendy’s might advertise to the segment of households headed by a man or woman under 25 years old. But Figures 8–6A and 8–6B do not tell us what media to use to reach them—such as by television ads or social networks. For those answers, we need to relate the age of the head of household again to the answers given to question 7 in Figure 8–4—the source of information that households use.
Probably the most widely used technique for organizing and presenting marketing data, cross tabulations have some important advantages. The simple format permits direct interpretation and an easy means of communicating data to management. Cross tabs offer great flexibility and can be used to summarize questionnaire, observational, and experimental data.
Cross tabulations also have some disadvantages. For example, they can be misleading if the percentages are based on too few observations. Also, cross tabulations can hide some relationships because each cross tab typically shows only two or three variables. Balancing both advantages and disadvantages, more marketing decisions are probably made using cross tabulations than any other method of analyzing data.

learning review
8-6.What is the difference between observational and questionnaire data?
8-7.Which type of survey provides the greatest flexibility for asking probing questions: mail, telephone, or personal interview?
8-8.What is cross tabulation?

Mark Twain once observed, “Collecting data is like collecting garbage. You’ve got to know what you’re going to do with the stuff before you collect it.” So, marketing data and information have little more value than garbage unless they are analyzed carefully and translated into findings, Step 4 in the marketing research approach.

LO 8-5 Explain how information technology and data mining lead to marketing actions.

Analyze the Data
Schwan Food Company produces 3 million frozen pizzas a day under brand names that include Tony’s and Red Baron. Let’s see how Teré Carral, the marketing manager for the Tony’s brand, might address a market segment question in early 2013. We will use hypothetical data to protect Tony’s proprietary information.

How are sales doing? To see how marketers at Tony’s Pizza assessed this question and the results, read the text.

Teré is concerned about the limited growth in the Tony’s brand over the past four years. She hires a consultant to collect and analyze data to explain what’s going on with her brand and to recommend ways to improve its growth. Teré asks the consultant to put together a proposal that includes the answers to two key questions:
1.How are Tony’s sales doing on a household basis? For example, are fewer households buying Tony’s pizzas, or is each household buying fewer Tony’s pizzas? Or both?
2.What factors might be contributing to Tony’s very flat sales over the past four years?
Facts uncovered by the consultant are vital. For example, is the average household consuming more or less Tony’s pizza than in previous years? Is Tony’s flat sales performance related to a specific factor? With answers to these questions Teré can take actions to address the issues in the coming year.

Page 213


These marketing dashboards present findings to Tony’s marketing manager that will lead to recommendations and actions.
Source: Teré Carral, Tony’s Pizza.


Present the Findings
Findings should be clear and understandable from the way the data are presented. Managers are responsible for actions. Often it means delivering the results in clear pictures and, if possible, in a single page.
The consultant gives Teré the answers to her questions using the marketing dashboards in Figure 8–7, a creative way to present findings graphically. Let’s look over Teré’s shoulder as she interprets these findings:
•Figure 8–7 A, Annual Sales—This shows the annual growth of Tony’s Pizza is stable but virtually flat from 2010 through 2013.
•Figure 8–7 B, Average Annual Sales per Household—Look closely at this graph. At first glance, it seems like sales in 2013 are half what they were in 2010, right? But be careful to read the numbers on the vertical axis. They show that household purchases of Tony’s pizzas have been steadily declining over the past four years, from an average of 3.4 pizzas per household in 2010 to 3.1 pizzas per household in 2013. (Significant, but hardly a 50 percent drop.) Now the question is, if Tony’s annual sales are stable, yet the average individual household is buying fewer Tony’s pizzas, what’s going on? The answer is, more households are buying pizzas—it’s just that each household is buying fewer Tony’s pizzas. That households aren’t choosing Tony’s is a genuine source of concern. But again, here’s a classic example of a marketing problem representing a marketing opportunity. The number of households buying pizza is growing, and that’s good news for Tony’s.

Page 214

•Figure 8–7 C, Average Annual Sales per Household, by Household Size—This chart starts to show a source of the problem: Even though average sales of pizza to households with only one or two people are stable, households with three or four people and those with five or more are declining in average annual pizza consumption. Which households tend to have more than two people? Answer: Households with children. Therefore, we should look more closely at the pizza-buying behavior of households with children.
•Figure 8–7 D, Average Annual Sales per Household, by Age of Children in the Household—The real problem that emerges is the serious decline in average consumption in the households with younger children, especially in households with children in the 6-to-12-year-old age group.
Identifying a sales problem in households with children 6 to 12 years old is an important discovery, as Tony’s sales are declining in a market segment that is known to be one of the heaviest in buying pizzas.
Effective marketing research doesn’t stop with findings and recommendations—someone has to identify the marketing actions, put them into effect, and monitor how the decisions turn out, which is the essence of Step 5.
Make Action Recommendations
Teré Carral, the marketing manager for Tony’s Pizza, meets with her team to convert the market research findings into specific marketing recommendations with a clear objective: Target households with children ages 6 to 12 to reverse the trend among this segment and gain strength in one of the most important segments in the frozen pizza category. Her recommendation is to develop:
•An advertising campaign that will target children 6 to 12 years old.
•A monthly promotion calendar with this age group target in mind.
•A special event program reaching children 6 to 12 years old.
Implement the Action Recommendations
As her first marketing action, Teré undertakes advertising research to develop ads that appeal to children in the 6-to-12-year-old age group and their families. The research shows that children like colorful ads with funny, friendly characters. She gives these research results to her advertising agency, which develops several sample ads for her review. Teré selects three that are tested on children to identify the most appealing one, which is then used in her next advertising campaign for Tony’s Pizza. This is the ad shown to the left.

Marketing research at Tony’s Pizza helped develop this colorful, friendly ad targeted at families with children in the 6-to-12-year-old age group.

Evaluate the Results
Evaluating results is a continuing way of life for effective marketing managers. There are really two aspects of this evaluation process:
Evaluating the decision itself. This involves monitoring the marketplace to determine if action is necessary in the future. For Teré, is her new ad successful in appealing to 6-to-12-year-old children and their families? Are sales increasing to this target segment? The success of this strategy suggests Teré should add more follow-up ads with colorful, funny, friendly characters.
Evaluating the decision process used. Was the marketing research and analysis used to develop the recommendations effective? Was it flawed? Could it be improved for similar situations in the future? Teré and her marketing team must be vigilant in looking for ways to improve the analysis and results—to learn lessons that might apply to future marketing research efforts at Tony’s.
Again, systematic analysis does not guarantee success. But, as in the case of Tony’s Pizza, it can improve a firm’s success rate for its marketing decisions.

Page 215


learning review
8-9.In the marketing research for Tony’s Pizza, what is an example of (a) a finding and (b) a marketing action?
8-10.In evaluating marketing actions, what are the two dimensions on which they should be evaluated?
LO 8-6Describe three approaches to developing a company’s sales forecast.

Forecasting or estimating potential sales is often a key goal in a marketing research study. Good sales forecasts are important for a firm as it schedules production. The term sales forecast refers to the total sales of a product that a firm expects to sell during a specified time period under specified environmental conditions and its own marketing efforts. For example, Betty Crocker might develop a sales forecast of 4 million cases of cake mix for U.S. consumers in 2014, assuming consumers’ dessert preferences remain constant and competitors don’t change prices.
Three main sales forecasting techniques are often used: (1) judgments of the decision maker, (2) surveys of knowledgeable groups, and (3) statistical methods.
Judgments of the Decision Maker
Probably 99 percent of all sales forecasts are simply the judgment of the person who must act on the results of the forecast—the individual decision maker. A direct forecast involves estimating the value to be forecast without any intervening steps. Examples appear daily: How many quarts of milk should I buy? How much money should I withdraw at the ATM?
lost-horse forecast involves starting with the last known value of the item being forecast, listing the factors that could affect the forecast, assessing whether they have a positive or negative impact, and making the final forecast. The technique gets its name from how you’d find a lost horse: go to where it was last seen, put yourself in its shoes, consider those factors that could affect where you might go (to the pond if you’re thirsty, the hayfield if you’re hungry, and so on), and go there.
For example, New Balance recently introduced its Minimus, a shoe that is 50 percent lighter than other lightweight shoes. Its unique features are designed to make running easier and limit foot injuries. Suppose a New Balance marketing manager in early 2014 needs to make a sales forecast through 2016. She would take the known value of 2013 sales and list positive factors (good acceptance of its high-tech designs, great publicity) and the negative factors (the economic recession, competition from established name brands) to arrive at the final series of sales forecasts.35

How might a marketing manager for the New Balance Minimus running shoe create a sales forecast through 2016? Read the text to find out.
Page 216

Surveys of Knowledgeable Groups
If you wonder what your firm’s sales will be next year, ask people who are likely to know something about future sales. Two common groups that are surveyed to develop sales forecasts are prospective buyers and the firm’s salesforce.
survey of buyers’ intentions forecast involves asking prospective customers if they are likely to buy the product during some future time period. For industrial products with few prospective buyers, this can be effective. There are only a few hundred customers in the entire world for Boeing’s large airplanes, so Boeing surveys them to develop its sales forecasts and production schedules.
salesforce survey forecast involves asking the firm’s salespeople to estimate sales during a forthcoming period. Because these people are in contact with customers and are likely to know what customers like and dislike, there is logic to this approach. However, salespeople can be unreliable forecasters—painting too rosy a picture if they are enthusiastic about a new product or too grim a forecast if their sales quota and future compensation are based on it.
Statistical Methods
The best-known statistical method of forecasting is trend extrapolation, which involves extending a pattern observed in past data into the future. When the pattern is described with a straight line, it is linear trend extrapolation. Suppose that in early 2000 you were a sales forecaster for the Xerox Corporation and had actual sales data running from 1988 to 1999 (see Figure 8–8). Using linear trend extrapolation, you draw a line to fit the past sales data and project it into the future to give the forecast values shown for 2000 through 2012.
If in 2013 you want to compare your forecasts with actual results, you are in for a surprise—illustrating the strength and weakness of trend extrapolation. Trend extrapolation assumes that the underlying relationships in the past will continue into the future, which is the basis of the method’s key strength: simplicity. If this assumption proves correct, you have an accurate forecast. However, if this proves wrong, the forecast is likely to be wrong. In this case, your forecasts from 2000 through 2012 were too high, as shown in Figure 8–8, largely because of fierce competition in the photocopying industry. The spike in 2010 sales revenues is mainly due to new acquisitions.

Linear trend extrapolation of sales revenues at Xerox, made at the start of 2000.


Page 217


learning review
8-11.What are the three kinds of sales forecasting techniques?
8-12.How do you make a lost-horse forecast?

LO 8-1 Identify the reason for conducting marketing research.
To be successful, products must meet the wants and needs of potential customers. So marketing research reduces risk by providing the vital information to help marketing managers understand those wants and needs and translate them into marketing actions.
LO 8-2 Describe the five-step marketing research approach that leads to marketing actions.
Marketing researchers engage in a five-step decision-making process to collect information that will improve marketing decisions. The first step is to define the problem, which requires setting the research objectives and identifying possible marketing actions. The second step is to develop the research plan, which involves specifying the constraints, identifying data needed for marketing decisions, and determining how to collect the data. The third step is to collect the relevant information, which includes considering pertinent secondary data (both internal and external) and primary data (by observing and questioning consumers) as well as using information technology and data mining to trigger marketing actions. The fourth step is to develop findings from the marketing research data collected. This involves analyzing the data and presenting the findings of the research. The fifth and last step is to take marketing actions, which involves making and implementing the action recommendations and then evaluating the results.
LO 8-3 Explain how marketing uses secondary and primary data.
Secondary data have already been recorded prior to the start of the project and consist of two parts: (a) internal secondary data, which originate from within the organization, such as sales reports and customer comments, and (b) external secondary data, which are created by other organizations, such as the U.S. Census Bureau (which provides data on the country’s population, manufacturers, retailers, and so on) or business and trade publications (which provide data on industry trends, market size, etc.). Primary data are collected specifically for the project and are obtained by either observing or questioning people.
LO 8-4 Discuss the uses of observations, questionnaires, panels, experiments, and newer data collection methods.
Marketing researchers observe people in various ways, such as electronically using Nielsen people meters to measure TV viewing behavior or personally using mystery shoppers or ethnographic techniques. A recent electronic innovation is neuromarketing—using high-tech brain scanning to record the responses of a consumer’s brain to marketing stimuli like packages or TV ads. Questionnaires involve asking people questions (a) in person using interviews or focus groups or (b) via a questionnaire using a telephone, fax, print, e-mail, Internet, or social media survey. A cross tabulation is often used with questionnaire data to analyze the relationships among two or more variables to lead to marketing actions. Panels involve a sample of consumers or stores that are repeatedly measured through time to see if their behaviors change. Experiments, such as test markets, involve measuring the effect of marketing variables such as price or advertising on sales. Collecting data from social networks like Facebook or Twitter is increasingly important because users can share their opinions about products and services with countless “friends” around the globe.
LO 8-5 Explain how information technology and data mining lead to marketing actions.
Today’s marketing managers are often overloaded with data—from internal sales and customer data to external data on TV viewing habits or grocery purchases from the scanner data at checkout counters. Information technology enables this massive amount of marketing data to be stored, accessed, and processed. The resulting databases can be queried using data mining to find statistical relationships useful for marketing decisions and actions.
LO 8-6 Describe three approaches to developing a company’s sales forecast.
One approach uses the subjective judgments of the decision maker, such as direct or lost-horse forecasts. A direct forecast involves estimating the value to be forecast without any intervening steps. A lost-horse forecast starts with the last known value of the item being forecast and then lists the factors that could affect the forecast, assesses whether they have a positive or negative impact, and makes the final forecast. Surveys of knowledgeable groups, a second method, involve obtaining information such as the intentions of potential buyers or estimates provided by the salesforce. Statistical methods involving extending a pattern observed in past data into the future are a third approach. The best-known statistical method is linear trend extrapolation.

Page 218

cross tabulation
information technology
marketing research
measures of success
observational data
primary data
questionnaire data
sales forecast
secondary data
1Suppose your dean of admissions is considering surveying high school seniors about their perceptions of your school to design better informational brochures for them. What are the advantages and disadvantages of doing (a) telephone interviews and (b) an Internet survey of seniors requesting information about the school?
2Wisk detergent decides to run a test market to see the effect of coupons and in-store advertising on sales. The index of sales is as follows:
What are your conclusions and recommendations?
3Nielsen obtains ratings of local TV stations in small markets by having households fill out diary questionnaires. These give information on (a) who is watching TV and (b) the program being watched. What are the limitations of this questionnaire method?
4The format in which information is presented is often vital. (a) If you were a harried marketing manager and queried your information system, would you rather see the results in tables or charts and graphs? (b) What are one or two strengths and weaknesses of each format?
5(a) Why might a marketing researcher prefer to use secondary data rather than primary data in a study? (b) Why might the reverse be true?
6Look back at Figure 8–4. Which questions would you pair to form a cross tabulation to uncover the following relationships? (a) Frequency of fast-food restaurant patronage and restaurant characteristics important to the customer, (b) Age of the head of household and source of information used about fast-food restaurants, (c) Frequency of patronage of Wendy’s and source of information used about fast-food restaurants, and (d) How much children have to say about where the family eats and number of children in the household.
7Look back at Figure 8–6A. (a) Run the percentages vertically and explain what they mean. (b) Express all numbers in the table as a percentage of the total number of people sampled (586) and explain what the percentages mean.
8Which of the following variables would linear trend extrapolation be more accurate for? (a) Annual population of the United States or (b) annual sales of cars produced in the United States by Ford. Why?
To help you collect the most useful data for your marketing plan, develop a three-column table:
1In column 1, list the information you would ideally like to have to fill holes in your marketing plan.
2In column 2, identify the source for each bit of information in column 1, such as an Internet search, talking to prospective customers, looking at internal data, and so forth.
3In column 3, set a priority on information you will have time to spend collecting by ranking each item: 1 = most important; 2 = next most important, and so forth.

Page 219


Carmex® (A): Leveraging Facebook for Marketing Research

QR 8-4
Carmex (A) Video Case

“What makes social media ‘social’ is its give and take,” says Jeff Gerst of Bolin Marketing, who manages the Carmex® social media properties. By “give” Gerst is referring to the feedback consumers send on social media; “take” is what they receive—such as news and coupons. “For Carmex, Facebook isn’t just a way to share coupons or the latest product news, but it is also a marketing research resource. We have instantaneous access to the opinions of our consumers.”
“While some people think of social media as ‘free,’ that is not true. However, almost everything in social media can be faster and cheaper than in the offline world,” adds Dane Hartzell, general manager of Bolin Digital. “Many platforms have been prebuilt and we marketers only need to modify them slightly.”
Although Carmex has been making lip balm since 1937, only in the last five years has it made serious efforts to stress growth and become more competitive. For example, Carmex has:
•Extended its lip balm products into new flavors and varieties.
•Expanded into nearly 30 international markets.
•Developed the Carmex Moisture Plus line of premium lip balms for women.
•Launched a line of skin care products, its first venture outside of lip care.
Carmex has used social media tools in developing all of these initiatives, but the focus of this case is how Carmex might use Facebook marketing research to grow its lip balm varieties in the United States.
Brands can leverage Facebook and all social media platforms to test what topics and themes its audience engages with the most as well as validate concepts and ideas. In 2012 Carmex identified the growing trend of consumers seeking product customization. Carmex combined research with Facebook engagement data, which helped to validate consumer interest and led it to develop two new lines of limited-edition lip balm products that launched in 2013.
The first line was a set of three different Carmex “City Sticks” featuring New York, Chicago, and Las Vegas versions of the Carmex lip balm stick with recognizable landmarks from each city on them. The brand partnered with Walgreens to exclusively sell the “City Sticks” in each of the three cities. During this time Carmex leveraged its social media channels on Facebook and Instagram to solicit photos of fans holding up their favorite style of Carmex in front of a landmark in their own city. Carmex then used these photos to help it decide on new locations for future limited-edition “City Sticks.”

Page 220

Carmex’s second line of new products was four fashion-forward “glamorous” designs of Carmex Moisture Plus. Carmex researched current design trends in the women’s fashion industry to come up with the four different styles, and it had seen good engagement from its Facebook community on “fashion themed” posts, which helped validate the concept. The four styles were: “Chic,” a black and white houndstooth; “Fab,” with bright purple circles; “Adventurous,” a leopard print; and “Whimsical,” with blue, orange, green, and pink intertwined ribbons. Carmex first announced the line to its Facebook fans to generate interest and they were brought to market in the summer of 2013.
“We have three potential new flavors and we can only put two into quantitative testing,” explains Jeff Gerst to his team. “So we have two goals in doing marketing research on this. One is to use Facebook to help us determine which two flavors we should move forward with. The second goal is to drive our Facebook metrics.”
The two key Facebook metrics the Carmex marketing team has chosen to help narrow the flavor choices from three to two are “likes” and “engagement.” “Likes” are the number of new “likers” to the brand’s Facebook Page. This metric measures the size of the brand’s Facebook audience. In contrast, “engagement” measures how active its Facebook audience is with Carmex. Any time a liker posts a comment on the Carmex Wall, likes its status, or replies to one of its posts, the engagement level increases.
The easiest way for Carmex to grow the number of “likes” on its Facebook Page is through contests and promotions. If it gives away prizes, people will be drawn to its site and its likes will increase. However, these people may not actually be fans of the Carmex product so at the end of the promotion, they may “unlike” Carmex or they may remain fans but not engage with the Carmex Page at all.
“One of the biggest challenges facing Facebook Community Managers for brands is how to grow your likes without hurting the level of engagement,” says Holly Matson, director of experience planning at Bolin Marketing.
“Depending on how we go about conducting the research,” Gerst adds, “we can drive engagement with our existing Facebook community, we can use this as an opportunity to grow our Facebook community or, potentially, we could do both.” The benefits of this Carmex Facebook strategy are twofold: (1) narrowing the number of flavors to be researched from three to two and (2) enhancing the connections with the Carmex Facebook community.
Carmex’s Facebook activity can benefit (1) by using a poll to increase engagement, (2) by launching a contest to increase the number of likers, and (3) by trying to increase both engagement and likers through combining a poll with a contest.
The “Engagement” Strategy: Use a Poll
Let’s look at two ways to use the engagement strategy showing actual Facebook screens. First, Carmex can post a somewhat open-ended question on its Facebook Wall, such as, “Which Carmex lip balm flavor would you most like to see next: Watermelon, Green Apple, or Peach Mango?” (Figure 1). However, consumers are less likely to respond to a question if they have to type in a response and have their name attached to it.

Facebook Open-Ended Poll Question


Alternatively, Carmex can post the same question on its Wall as a fixed-alternative poll question (Figure 2). Then consumers need only click on a flavor to vote; this is quick, anonymous, and will drive more people to vote, where more votes means more engagement. Within five minutes Carmex will have several dozen votes and, by the end of a business day, Carmex can very easily have over 500 responses.

Facebook Fixed-Alternative Poll Question


Page 221

In this scenario, the consumers are content because they are able to engage with a brand they like and have their opinions heard. Carmex is content because it has engaged hundreds of its fans on its Facebook Page, and it gains results that are very helpful in deciding which flavors to put into testing. This scenario gets an answer quickly and drives fan engagement with existing fans but does not drive new likers to the Carmex Facebook Page.

Potential Results from Three Possible Facebook Strategies


The “Likes” Strategy: Use a Contest
If Carmex wants to grow the size of its Facebook community, which means the number of its brand page “likes,” it can adopt a different strategy. Carmex can announce a contest where, if consumers “like” Carmex on Facebook and share a comment, they will be entered to win three limited-edition flavors. The chance to win limited-edition flavors is exciting to Carmex enthusiasts, and a contest like this will draw new consumers to the page. Carmex can ask the winners to review the limited-edition flavors and see if there is a consensus on which flavors should move on to quantitative testing. Setting up a contest, developing official rules, promoting the contest through Facebook ads, and fulfilling a contest can be costly and time-consuming.
The Combined Strategy: Use Poll and Contest
Carmex can also choose to layer these two strategies into a combined strategy where it runs the limited-edition flavor contest to promote new likes and meanwhile posts the poll question on its Facebook Wall to drive engagement.
Figure 3 shows the potential results from the three Facebook strategies being considered—the poll only, the contest only, or both strategies together. Assume the Carmex marketing team has sought your help in selecting a strategy and needs your answers to the questions below.
1What are the advantages and disadvantages for the Carmex marketing team in collecting data to narrow the flavor choices from three to two using (a) an online survey of a cross-section of Internet households or (b) an online survey of Carmex Facebook likers?
2(a) On a Facebook brand page, what are “engagement” and “likes” really measuring? (b) For Carmex, which is more important and why?
3(a) What evokes consumers’ “engagement” on a brand page on Facebook? (b) What attracts consumers to “like” a brand page on Facebook?
4(a) What are the advantages of using a fixed-alternative poll question on Facebook? (b) When do you think it would be better to use an open-ended question?
5(a) If you had a limited budget and two weeks to decide which two flavors to put into quantitative testing, would you choose a “poll only” or a “contest only” strategy? Why? (b) If you had a sizable budget and two months to make the same decision, which scenario would you choose? Why?


"Is this question part of your assignment? We Can Help!"