New Products and Consumer Preferences

Before a new product can be made, or an existing one varied, someone must have the idea of doing so. Since market research as a business activity is conducted and paid for by firms, the investigation of consumer wants and preferences is fairly well defined by the spheres of interest that exist within the firm. There is little point in asking the public the open-end questions, "What do you need, or want? What shall we make?" The consensus appears to be that consumers themselves scarcely ever originate new product ideas, that development of a specific product represents the functioning of creativity in business, and that such creativity may be found among technical research, production engineering, market research, or sales staff, all of whom have been responsible for one idea or another.

Considering the magnitude of choice that confronts the American public, it is easy to understand why this is true. Any household consumer is likely to deal in a far greater variety of products and purchases than any firm, and any consumer's suggestions for new items cover a broader field than that of any business interest. The firm is oriented to production and marketing; the consumer, to his own needs and to an unlimited range of alternatives. The consumer may say, for example, in answer to the open-end question, "What do you want?" "We'd like something to make shoe polishing less of a messy job." For the manufacturer of shoe polish, however, this does not offer clear instructions on how to adapt his paste, liquid, or spray toward even greater convenience. Nor does it produce for the manufacturer of shoes a formula for a plastic shoe that can be so cheaply made, it will be discarded when soiled. One acceptable alternative, for consumers, might conceivably be for drug companies to develop a toughening agent for feet so that shoes are no longer needed for protection, and then for the world of fashion to insist that bare feet are the epitome of style. Thus, consumers' suggestions rarely furnish new product ideas, either because consumers are inarticulate about their own needs, or because they lack the technical competence to specify feasible alternatives.

But consumer "wants and needs," more narrowly defined, have been intensively explored by firms that have been seeking new product ideas. Surveys of household spending patterns, can be analyzed for clues to potential markets, and then augmented by investigations of different design. For example, although food consumption data have been collected since Engel's time, nearly all reports have been in terms of dollars spent, or quantities bought, or nutrients consumed, or different types of foods. Yet the consumer generally eats food in the form of meals: the same amount of bread per week may mean noonday sandwiches for one family; for others, it may mean bread pudding for dinner, French toast for breakfast, scalloped tomatoes for luncheon or cinnamon toast for a midnight snack. A detailed study of eating habits (known as "the menu study") was conducted in 1958 by the Market Research Corporation of America, to record the forms of food used and the types of meals served. Such a special study of consumer behavior -- of what consumers do -- can be broadly or narrowly defined: it may take in a household activity, such as cleaning chores, or a part of the shopping process, such as having appliances serviced; it may require the consumer to cooperate by recording purchases in a diary, or to be oblivious to the observation of her buying behavior in supermarkets; it may analyze broad social trends, such as the increasing number of married women in the labor force, or the opinions that consumers give about such developments.

A fairly common framework for direct investigation of consumer wants and needs is to set up a discussion topic with a few participants, sometimes under the direction of a psychologist who is versed in small-group dynamics. For example, a food company's market-research staff organizes a panel of teenagers to talk about the ways they like to entertain and their association of "party" with "food"; or a psychologist who is working for an oil company talks with mothers about vacation travels and taking care of young children on the road. The use of depth interviews and motivational research into what people think and feel about such topics, as opposed to what they say about them, has become less controversial now than when it first appeared. But consumer comments in such discussions are apt to be impressionistic or imprecise. For example, after the U.S. Housing and Home Finance Agency convened a "Women's Congress on Housing" in 1956, McCall magazine sponsored an "Annual Congress on Better Living," attended by 100 women from all over the country, who aired their views on housekeeping and homemaking. This conference reported that 8 out of 10 delegates were favorably inclined toward colored light bulbs, yet at the second conference a year later, while more than one-fourth of the delegates had used colored light bulbs, very few recommended them. Consequently, the final report warned that "the true substance of the Congress is reflected in the thousands of quotations found in the report....Interpretations given should not be projected to the entire population."

All such techniques look for provocative comments with which to stimulate the staff. For a business which traces the origin of a new product to its "surveys of consumer wants and needs," the consumer has provided the catalyst, not the end result.

Once the idea for a new product is at hand, consumers can evaluate it through concept research. Testing consumer reaction before the product itself exists requires that the product be described by words or pictures or by a dummy package. The description is of the concept involved in the product rather than of the item itself; it is kept vague precisely so that consumers can fill in specifications. For example, a manufacturer of paper products envisages throwaway doormats of absorbent stock, or a vending-machine company thinks of installing frozen-food dispensers next to a coin-operated electronic oven. The research staff then submits a brief statement to consumers for discussion, asking what they like about the idea; how they would use such a product; what their family's reaction would be; where it should be sold; what items a store should drop from stock in order to carry it; what advantages and disadvantages they can find after thinking about the notion for a time; how they would change it; what else it makes them think about; do they want it? need it? is it good for anything?

The simplest result of such an investigation is a "stop" or a "go" signal. Once consumers have registered wholehearted approval, the company can then ask its technical staff whether the product can, in fact, be made; if the audience interviewed shows little interest in the idea, or violent disapproval of it, the company can shelve the project. Some firms state flatly that concept research to determine consumer approval must precede any directions to their technical or engineering staffs, whose resources are too valuable to waste in developing products that cannot be sold. Some firms even use concept research to set priorities for its developmental work: they list projects under consideration and ask consumers to rank them in order of their importance. More often, such research turns up suggestions for changing the product or using it differently. According to one of the leading experts in this field, "The consumer allowed to define the product herself may well find more value in it than the manufacturer alone could ever see." 8 Most companies agree that this first screening of a new product does not require a representative sample of consumers, for the firm is attempting chiefly to discover all types of reaction. Because the concept is vague, comments from the audience may be irrelevant: they may express dislike for a feature that the firm knows can be changed, or else suggest a modification that the company knows will be technically difficult to achieve. There is no reason to elaborate such reactions by securing them from a "properly designed" sample. If the concept survives this very preliminary test, the firm can then move on to more systematic research into consumer preferences that are relevant to the product itself.

On the other hand, concept research is not suitable for dealing with many new products, nor for some variations of existing products. Some investigators are convinced that a consumer must handle the product, use it, or "get involved with it," before any useful reaction can be obtained. In certain cases, the consumer's ability to visualize the new product may be distrusted. For example, once aerosol cans had become available, concept research established rather readily that consumers would favor their use for condiments, polishes, cosmetics, and paints. But if insecticides and window-cleaners had not already familiarized consumers with the use of pressurized sprays, concept research might have been less successful in exploring reactions to the aerosol can itself. In cases in which style will be of major importance to the finished product, companies agree that early consumer approval means little, for, in matters of fashion, the general public still follows rather than leads. Finally, where an existing product is to be changed, the maker can ask consumers directly for their reactions or submit very specific suggestions for their approval.

There are some fairly early examples of consumer research designed to work out the details of a new product or its variation. In the early '20s, the N. W. Ayer Company, advertising agency for Cannon towels, asked housewives for their opinions about buying and using colored towels, which till then had always been white. The manufacturer gained considerable competitive advantage from introducing differentiated products -- particularly, when these were in the colors and specific designs that met consumer preference. In the same decade Cluett, Peabody & Company ordered an extensive consumer survey to gauge the market for collar-attached shirts and to find out what exactly wearers wanted in a shirt. The results -- which specified well-anchored buttons, comfort and "fit" in the collar -- were used by the firm to design shirts that it introduced in 1928, when it shifted the bulk of its output away from collarless shirts and separate starched collars. What consumers wanted most, the survey showed, was to have both shirts and collars that did not shrink; this knowledge was considerable encouragement for the company to adopt Sanford Cluett's preshrinking process, despite the extensive educational campaign that it required for both retailers and consumers.

Market research techniques today seek to discover consumer specifications for a product, by means of questionnaires, interviews, and product sampling. For example, suppose that disposable doormats have passed preliminary screening: using concept research, investigators have found housewives receptive to the idea and have heard suggestions for making use of the product in offices, schools, and other public buildings. A cat fancier saw possibilities of using the mats to line cages and baskets, and one woman pointed out that a material that was strong enough for a doormat might do for umbrellas, which one could use once and then throw away, since they were frequently lost in any case. The firm's market analysts began to collect figures on the sales of fiber, rubber, and plastic matting -- the "close substitutes" or competitive alternatives for consumer choice; the sales research staff drew up tentative plans for distribution through variety stores and supermarkets, with different price and discount structures; and technical personnel developed several prototypes, differing in weight and size and in a special process that made the material absorbent on one side and waterproof on the other. At this point, consumer research may enlist company personnel to take home samples for use and to place the doormats at strategic points in offices and factory. Samples may be donated for use in a neighboring school or theater. Visitors to the firm's display center may be asked to view three different mats and then fill out a simple questionnaire about preference. Field workers may seek out housewives who agree to use the product for two weeks and then report the reactions of family members and guests. Sometimes, the firm will invite small groups of consumers to visit the laboratory and handle or taste the new appliance or food item; for other products, it is the reactions of the retail trade that are of paramount importance:

The one thing to be sure of, however, is not that you think the new gadget or the new gimmick is clever or good or cheap or whatever -- the important thing to find out ahead of time is whether the fellow you are going to sell it to thinks it is good or smart or new or something, and also whether he thinks he can make money doing with it because he is not interested, fellows, in your making money.

It is significant that consumer research is a fundamental step for many firms that make no consumer goods or sales to retailers. The experience of DuPont in promoting nylon and other synthetic textiles is well known: although the company made only the fiber, its market research program called for developing not only fabrics but garments, and for commissioning designers to create models specifically for the new materials. Such a program is typical of industries that have created new materials -- such as aluminum, synthetic fibers, and plastics. All such firms have devoted major energies to developing final consumer goods that use their output as a material, and to securing consumer acceptance of these goods. Typical of their reasoning is the following explanation by an executive of Celanese, Inc.:

We must know the habits, attitudes, preferences and needs at each level of the flow of our goods starting with the ultimate consumer and working back through the retailer and the various wholesale and process levels until we arrive at our direct customer, in our case the weaver or the knitter, etc.

In our case we spend a very large part of our effort in obtaining a thorough understanding of the ultimate consumer in particular. That is the man or woman who buys our product and makes the final decision as to whether or not it is worth what he paid for it. We do this through our own direct mail panels, through field surveys of attitudes and preferences and by working closely with retailers and following up the goods that they have sold to customers through their normal channels.

The use of consumer research is therefore not restricted to socalled "consumer goods" firms, and a positive buyer reaction to new products and opinions on their specific characteristics may be essential in order to achieve proper distribution of the innovation.

Depending on the product and the company, this stage of consumer research may or may not involve careful sampling procedures. The process attempts to enlist consumers as judges; their indications of potential losers, however, may be more important than their selections of a winner. If the market for the product is fairly well defined, samplers and testers are chosen from that market: the reactions of bank presidents or their secretaries to a synthetic fabric designed for heavy-duty overalls would be somewhat irrelevant. Some consumer panels are exclusive: a nationally known ice-cream company and an equally famous proprietary drug manufacturer have both asked children for their comments on flavor and taste; a panel of canines supplies useful information to a firm that makes dog food; new mothers are enlisted by their friends to participate in evaluating babies' wear for a clothing manufacturer. In other cases, however, the potential market is still undefined, so that almost any consumer reaction may be relevant. Comments or criticisms direct the technicians and engineers to change the prototypes along given lines, and the firm revises and alters its product in a series of successive approximations to what consumers most prefer. Again, the most useful result may be a simple "stop" or "go" signal; with a prototype in existence, the firm can begin the cost and profit analyses that may also provide such a signal. If consumers agree that a specific modification, whose cost can be predicted, is essential, and if sales research shows that the market for an item priced to cover such costs is limited (or worse, occupied by an entrenched competitor, with a high degree of consumer loyalty), then the company had best shelve the whole project. Consumer preference and cost estimates thus together provide an evaluation of new products before they are placed on the market.

Assuming, however, that product specifications to fit consumer preferences have been evolved and that cost analyses predict no obvious losses, still the size of the market has yet to be determined. The company needs to learn the potential repeat purchases, or -- in the case of a major appliance or some other item infrequently purchased -- the potential consumer satisfaction that will build the market. In short, although the firm has so far proceeded with a crude "go" signal, it now needs more precise information on just how closely the new product matches consumer preferences.

If the product is a variation, or faces competition from close substitutes, consumer research attempts to learn whether buyers will find it significantly differentiated and whether, once they have tried the product, they will find a level of utility satisfactory enough to warrent repeat purchases. Business executives are keenly aware that it is consumer preference that determines the success of a competitive move to vary the product:

Don't ever test a product that is nearly as good -- or just as good ...if you haven't got a definite advantage then spend your time, effort and money to get a better product or a different product or one with some built-in advantages....Don't bother at all unless you know there is a need for the product. This is one you can't think about. It is so darn obvious and still the mistake is made so many times.

Firms that are introducing a new product or a product variation that will confront powerful competition use blind product testing, to let consumers decide whether or not a "definite advantage" exists. Samples of the new product and of competing items, labeled as A and B or some other anonymous identification, are supplied to households for use under ordinary conditions. Consumers then report their preferences and the details of their experience, including suggestions for product improvement, or for packaging, directions for use, and so on. There are those who find this type of consumer research to be essential:

Don't waste marketing dollars on products which fail to develop a clear and decided product superiority in the customary blind product tests. A new product in a worthwhile field will undoubtedly buck up against one or more strong consumer franchises. Companies owning these established brands will be stimulated to extra effort by the introduction of your new brand. Their combined effort can effectively stunt the new brand's growth unless it is recognized as clearly superior by consumers.

Once again, business credits consumer choice with having sufficient power to defeat competitive effort.

The anonymity of blind product testing is not essential for all new items, especially if no close substitute exists. But household use of the product under normal conditions, whether or not the maker's identity is known, nearly always proves an effective form of consumer research. Visitors to a laboratory who sample several alternative formulas for a food item may agree unanimously on one choice. But most families do not have meals accompanied by a sip of coffee, and then one of tea, followed by milk and a taste of cocoa; it is only rarely that two products are used simultaneously for the same purpose. If the housewife uses product A for one week and then tries product B for another, the differences discerned in side-by-side tests may wash out, or else other reasons for preferring one to the other may show up. Home use can provide useful suggestions for further development. Pretesting of early models of the Princess telephone "considerably reduced the risks involved in introducing this new telephone," for consumers complained that the base of the instrument was too lightweight and hard to grasp, objected to the flat dial, and indicated their preferences for a night light and various colors. Reporting on this experience, a company executive explained that "the modifications we plan to make will be in line with the findings of the studies."

Various homes offer various conditions of use: not only differences in climate, utilities, and the surroundings of household operations, but different degrees of use (and abuse), depending on the consumer's skill and that of the family members, as well as their total way of living. The prospective appliance, which gives outstanding performance when it is carefully adjusted by a technician, may prove thoroughly unsatisfactory to a nearsighted, short-tempered, or hasty housewife. A toilet preparation, food package, or liquid dye, which gives excellent results when measured and prepared in the laboratory, may fail in the face of the lack of calibrated instruments and of ideal storage conditions in the home. On the other hand, use tests may benefit from the inventiveness of consumers: a survey of families who bought cotton-tipped sticks that had been designed for baby care disclosed that these items were also used for children's paints, sportsmen's guns, coffeepots, cosmetics, and glue.

This last example illustrates a major problem in drawing a sample of consumers, either for blind product testing or for reports on use under normal conditions. To gain information from prospective users, the firm wishes to sample potential buyers rather than a cross-section of the total population; yet in many cases the market for the product may not be readily apparent. If one is to rely on the results from any sample, it must be drawn in terms of the universe. For example, to determine the average number of words to a page in this book, one need count the words on only a few pages, selected at random within the book: these pages represent a sample of the universe that includes all the pages within this book. But the universe of buyers of a new product cannot be defined in this way. It would seem obvious to select mothers of young babies in order to test baby oil; yet bachelors have been known to buy gifts for babies, and some enterprising consumer may discover that the oil offers an ideal lubricant for stuck zippers. It follows that the quantitative results from such consumer research can be variously interpreted. For example, home testing of the disposable paper doormat may show overall approval by 70 per cent of all users, but this may mean approval by 85 per cent of families living in the suburbs and only 40 per cent of city families. This may suggest hypotheses for further testing: does the difference represent only location? or does it reflect the presence of children and dogs in suburban families? or the physical difficulties of disposing of trash in the city? Is 70 per cent approval sufficient for the firm to take its product to market? What level of consumer acceptance is "enough" for the firm to proceed with its marketing plans? If the nonusers are violently opposed to the product, while the majority who do approve lack great enthusiasm, how much weight should be given to intensity of feeling? Most firms use quantitative measurement as guidelines; precisely because a new product is new and different, each situation is judged independently.

Beyond the details of consumer preference and acceptance gained by the research methods so far outlined, the firm that develops a new product also seeks information on the likelihood of its success in the market. Whether or not the tasters and testers and users like the samples, how many consumers will buy how much at what price? What will be the consumer choice, once the product is introduced into market competition?

Full-scale test marketing can provide some clues to the answer, but such investigations are feasible only with certain types of products, usually with those that have a small unit price and frequent consumer purchase. Food and household supplies, the items stocked by supermarkets and variety stores, predominate in test-marketing activities. New models of automobiles or appliances or furniture, or items such as clothing, for which fashion and style set the pace, cannot be as readily tested in one local market prior to national distribution. Where it can be used, however, test marketing covers a wide range of alternatives for consumer choice.

The simplest question for test marketing is the level of sales: over a three-month period how many items did consumers purchase? if there are close substitutes, how did sales compare to those of competitors? if not, how does the sales level, projected, compare to cost and profitability measures? Most test marketing, however, seeks more information than that; it wants to find out about the volume of repeat sales, the comments from users who are buyers (and therefore differ from previous consumers, whose reported reactions were to free samples), the impact of advertising or promotional campaigns, and so on. By testing in more than one area, the firm can experiment with various product designs, sizes, or prices. To ensure reliable results from such tests, the different consumers who are offered different alternatives must be in distinct groups, preferably not only separate from each other but, even more, unaware of the other possibilities being tested. Such complicated aspects of test marketing have led to the development of highly sophisticated techniques; like all of consumer research, these warrant more lengthy discussion than this brief introduction permits.

Source: Random House