While it's tempting to rush to market as your transform your great idea into a reality, a lack of reliable market research can doom your dream. Know your market before you go to market.
The term "information age" defines our world today. All businesses require accurate and timely information to be successful. Whether your company is large or small, financing, equipment, materials, talent, and experience alone are not enough to succeed without a constant flow of the right business information.
Many large companies make market research into a very sophisticated and lengthy process so they can find out everything possible about their customers. Philip Kotler, author and economist, writes in Marketing Management,
Coke knows that we put 3.2 ice cubes in a glass, see 69 of its commercials every year, and prefer cans to pop out of vending machines at a temperature of 35 degrees . . . We each spend $20 per year on flowers; Arkansas has the lowest consumption of peanut butter in the United States; 51 percent of all males put their left pants leg on first, whereas 65 percent of women start with the right leg; and . . .Proctor and Gamble once conducted a study to find out whether most of us fold or crumple our toilet paper. . .
While you probably can't afford a separate marketing research department to gather and monitor all the information that could possibly help you, all successful business owners must know their markets, competitors, customer wants and needs, and "what it takes to be competitive." It is not enough to know the answers to what, where, when, and how questions about our businesses. We also need to know why people buy our products and services. You should expect to budget at least a minimal amount of time and money for research, especially if you are starting a new business or branching out into a new direction.
Determine your market research needs and goals
The first step in conducting market research is to decide what you really need to find out. The kind of information you are seeking should determine the type of research you will do (although budgetary constraints will play a part in your decision).
Do you need to obtain a general feel for how key target buyers think about your product category and its various types of items, brands, and buying occasions? If so, interviewing groups of target buyers in focus groups may be the way to go, even though this type of research indicates only directional trends and may not be statistically reliable. Or is the confirmation of general trends in your industry sufficient? In that case, reading information from outside information services, industry trade associations, and industry experts may be all that you need to do.
You may wish to conduct blind tests of different formulas before finalizing recipes for a new product. In that case, you can do "laboratory" tests, where brands, packages, and names of products are not revealed to the test subjects, and achieve statistically reliable results at the 90 percent to 95 percent confidence level of predictability. Or perhaps you have completed extensive product development and testing and are now ready for a field test of your prototype products.
Market research takes a variety of forms
Generally, market research breaks down into the following categories:
- Primary research. Research involving customized data-gathering about the specific usage patterns, product feature likes and dislikes, etc., of target buyers or current users of your products. This is research done specifically for you or by you.
- Secondary research. With secondary research, someone else has gathered data and complied it into non-customized reports. Most of us are familiar with secondary research from doing library research with books and periodicals. While extremely valuable for general market trends, it does not provide you with information about how your specific customers view your specific goods or services.
These two main types of research can be further broken down into sub-categories, as follows:
As you've likely already surmised, some market research options may not be affordable or appropriate for your company. You may have to be more creative and proactive in:
- deciding to do a limited amount of necessary research and setting affordable budgets
- working with market research specialists or outside experts to define research problems and the design of the research
- accepting the possibility of a greater number of errors or a "lower confidence level" in the mathematical predictability of research results due to:
- small budgets
- small sample sizes
- samples chosen in a manner that's not completely random
- conducting and analyzing the necessary market research yourself or with company personnel
Secondary market research can provide insights inexpensively
Secondary research is something every student has completed at one time or another, usually by doing library research with books and periodicals for a school report. This is usually the cheapest and easiest type of research for small businesses to conduct. However, it may be less useful than primary research because the information you obtain was not developed with your particular problem or situation in mind.
Nevertheless, for some types of information, secondary market research is the only kind available. Examples of information best gleaned from secondary research are questions about your competitor's market share or the absolute numbers of potential customers for your product or services.
Secondary research can be divided into two categories:
- External research. External research involves looking at data gathered by industry experts, trade associations, or companies that specialize in gathering and compiling data about various industries. This is exactly the same type of information that you would use to determine your customer demographics.
- Internal research. Internal research is data gathered by your company for purposes other than market research, but which you can use to gauge what the market will do in the future. For example, a sales reports broken down by product line can point to a growth area, even though it was originally generated for sales rep commission payments.
External secondary market research provides broad information
All businesses, large or small, need to know key information about their marketing environment, competitors, and target buyer/users. Smaller businesses may not be able to afford to purchase Nielsen data for their industries at a cost of thousands of dollars per month. However, total market size, major competitors by category, and target buyer/user profile information is often available free from industry publications and trade associations.
The most commonly utilized external information, with some sample websites, includes:
- trade and industry associations and their publications and databases;
- government databases
- companies and consulting firms that track the effectiveness of marketing and advertising, as well as national and regional trends.
Internal company data from competitors may be available by interviewing competitor company executives, attending industry trade shows, and asking the right questions from industry "experts." They may be unaffordable as consultants but willing to direct you to free databases that you would not ordinarily know of or have access to. And don't overlook your competitor's suppliers. They can be excellent sources of information to aid your research.
In addition, consider these other external secondary research sources:
- Online searches, using Google, Bing or other major search engines
- Local civic organizations and clubs like the Chamber of Commerce, Lions Club, and Rotary Club.
- Colleges and universities that might have departments/experts related to your field.
- State and local government agencies.
- Industry-specific advertising, promotion, and public relations agencies.
Internal information can be leveraged for market research
Information generated by your own company can often be a source of market research. Here we're talking about information that was gathered for purposes other than marketing. For example, it may have been gathered for financial or management purposes. Some of the most common sources of information are:
- daily, weekly, monthly, and annual sales reports, broken down by geographical area, by product line, or even by product
- accounting information (e.g., spending and profitability)
- competitive information gathered by the sales force
If you're in retailing or wholesaling, you have a wealth of information at your disposal if you keep detailed information about your sales, by product. You may be able to determine not only the types of products that sell best at various times of the year, but even the colors and sizes that your customers prefer. There are a number of inventory tracking software products on the market that can help you keep track of all this information, not only for financial and tax purposes, but for marketing purposes as well.
Primary market research answers your specific questions
Primary research involves the design and implementation of original research during which data is collected from the source to answer your specific questions. This is the major advantage of primary research—you get information on the specific question or problem you need answered, not information that merely applies to your industry or type of business in general.
Primary research is generally divided into two categories: "experimental" research and "non-experimental" research. As a practical matter, most small companies bypass expensive experimental research and utilize the real market environment to conduct field studies. However, for the sake of completeness, we will provide an overview of experimental research.
Experimental research involves manipulating variables
During experimental research, the researcher controls and manipulates elements of the research environment to measure the impact of each variable. For example, test subjects (who are consumers meeting certain criteria, such as frequent users of the particular product or service in question) are shown several television commercials, and after each one they are asked questions designed to measure the likelihood that they'll purchase the product advertised.
Experimental research is often used by large consumer goods companies to test:
- the effectiveness of new advertising, or competitors' advertising;
- the effect of various prices on sales of a product;
- consumer acceptance of new products in trial and repeat-purchase levels; and
- the effect of different package designs on sales.
There are two major types of experimental research: laboratory studies and field studies. In laboratory studies virtually all variables are controlled except the one being tested. These are generally done on the premises of the research company. In a field study, the testing is done in the "real world," often by test marketing the product in a few locations to see whether consumers will buy it.
Laboratory studies are accurate, but cost prohibitive
Laboratory studies are a type of primary marketing research that are often used by larger consumer products companies. Although these tests have a high degree of reliability and correlation with actual market performance when the test product, pricing, and advertising are similar to those actually used in the real market, they are cost prohibitive for nearly all smaller companies. A well-conducted laboratory study can easily run well over $250,000.
In a typical laboratory study, potential test respondents are approached and prescreened for type of products used and brand preference. Then they are exposed to some advertising that's being tested and are given a chance to purchase the brand among a competitive array of other products. Finally, the consumers are given some of the product to take home and use. The researcher follows up with the test respondents to see whether they used the product, how they liked it, and whether they would or did purchase it again. Laboratory tests are generally based upon a minimum of 100 to 300 test-respondents per location, with the tests completed in a few days at a mall or other location where target respondents tend to cluster.
Small companies can conduct small-base simulated testing with local research services or do their own research by showing target buyers advertising, placing prototype products in homes, and following up with carefully constructed questionnaires. However, even 100 to 300 test respondents may still not be a statistically reliable indication of real world conditions and responses. Smaller companies and local research services are not likely to have a broad database of test response histories or access to sophisticated mathematical computer modeling software. Thus, field studies are often the method of choice for small businesses who want to test a new business idea.
Field studies can be reliable and cost effective
Field studies are a type of experimental, primary market research that is more accessible to small businesses. They are generally real-world tests in a controlled group of stores or in a single city. For large and small companies, the best test of a product may often be actual market conditions. And for many small companies, a real world test is the only experimental research available — anything else is just too expensive.
Research experts often decry the non-representativeness of the test sites as compared to nationwide market research, and they dislike the lack of control over environmental influences. Another complaint that professional market researchers make is the difficulty of accurately monitoring test results and then estimating larger market acceptance upon national roll out, where environment, consumers, sales forces, competitors, and trade makeup may vary significantly in each regional area. However, your small business may have no intention of ever rolling out nationally, so a local field test may be all that you'll ever need.
While there are considerable opportunities for bias in local or regional market sites, consumer behavior and habits, and local competitive products, a single retail store or small city test market can provide significant real world feedback. It is not unusual for thrifty entrepreneurs to design, test, package, produce initial inventories, and test-market a new consumer retail product at a cost of $10,000 or less.
On the other hand, larger national and multi-national companies often spend hundreds of thousands of dollars in market research prior to launching a test market for a single product. A large company may spend over $1 million in a single test market for the first six months. While you may not have the desire or the resources to do such extensive testing, the fact that large companies are willing to invest so much money in this type of research should convince you that a small field test of your business idea is worth the effort, before you commit all your funds to the project.
There are two major field testing options to consider:
- controlled store testing—you make your product available in a single store or limited group of stores, and
- city or regional test markets—you make your product available in several stores within a city or a region.
Place products in one store for controlled store testing
Many small businesses cannot afford to hire outside consultants and researchers to provide an extensive product development program. However, even a one-person firm can conduct its own "in-house" controlled testing, which involves placing its products in selected stores. Controlled store test-marketing can reliably simulate real-market conditions and buyers. The information uncovered can reduce risks and save time.
Many entrepreneurs have successfully designed and conducted their own controlled test-marketing by finding a receptive store owner and placing their products in a single store. You can then refine your products before you expand sales beyond this home-grown test vehicle. Both products and services can be tested in this simple, low-cost format.
A single-person car-detailing company conducted a test of the business with a local car repair shop. This entrepreneur spent weekends at the repair shop offering customers on-the-premise cleaning and home pickup of their cars for cleaning, washing, waxing, and other car-detailing services. He refined his "bundle" of services and prices prior to conducting his business full-time and eventually offered his services as a sub-contractor to many other auto repair shops in the city.
For business owners contemplating opening a single new store, providing demonstrations of the new product prototype to potential target consumers/buyers can provide a wealth of qualitative data at low financial risk. This informal research can point you in the right direction for refinement of product features/benefits prior to committing scarce dollars for expensive molds, production equipment, or plant and office.
City, regional test markets provide broader feedback
City and regional test markets differ from controlled store test markets only in the number of stores and size of the geographical area involved. If you have the finances and the time to devote a disproportionate amount of sales and marketing attention to the product test sites, this may be an excellent way to determine whether your product has market appeal.
Non-experimental research is research done in the normal course of business, where the environment cannot be as closely controlled as in experimental research. Also the many variables of the "business" can't be as easily isolated. This research centers on measuring the entirety of a project rather than its separate parts.
Non-experimental research is divided into two categories:
- qualitative research, which seeks to obtain many subjective reactions from a limited number of test subjects and
- quantitative research, which seeks to obtain the reactions of many test subjects to a limited number of questions.
Non-experimental research is often used by companies to test:
- buyer responses to new products and product improvements (qualitative)
- buyer evaluation of advertising, packaging, and brand positioning (qualitative)
- effect of a 10 percent price increase on buyer purchase intent (quantitative)
- testing of a new formula against a similar competitive formula (quantitative)
Qualitative research focuses on individuals, not numbers
Qualitative research is original company research ("primary") on a subject in the normal course of company business ("non-experimental"). It is primarily concerned with getting a subjective "feel" for the research topic, not a numerical, statistically predictable measure. More concretely, you can think of qualitative research as in-depth subjective interviewing or conversing with target buyers or potential users of your product or service.
Qualitative research results might not accurately represent the entire market.
Asking friends and neighbors about how they "feel" about political candidates and their election platforms is technically qualitative research. However, a good researcher would make sure they all qualify as registered voters who do cast their ballots regularly in order to provide a comfort level for the results of these conversations about political candidates in a given election. However, if you conclude that 100 percent of your friends and neighbors will vote for a particular party, this information may not be reliable in predicting election trends. The information is biased by your friendship with the respondents, their demographic and geographic location, and lifestyle influences.
The value of qualitative research depends upon these factors:
- the subject matter,
- the background of research respondents, and
- the skill of the researcher.
The problem with qualitative research is that your confidence level, that the results are indicative of the general population or key target group trends, will be somewhat low. In general, qualitative data analysis is subject to large statistical errors in accurately predicting the behavior of the total market of users. This can be a particular problem in a focus group—especially if you are not using a professional market research company who will have a moderator trained to handle difficult people and to probe for accurate information. For example, if everyone in a focus group likes your product, that may be a ringing endorsement to proceed. However, maybe only one "loudmouth" in the focus group liked it, and everyone else was afraid to voice a contrary opinion. Because of this statistically low level of predictability, you should be looking for broad general indicators or trends to guide the next steps.
Focus groups and in-depth one-on-one interviews are common qualitative methods.
Focus groups, individual interviews are key to qualitative research
Focus groups can be thought of as "group interviews," where a manageable number of target buyers are brought together, presented with an idea or a prototype product, and asked to discuss their opinions with a moderator and with each other.
Focus group participants generally expect to be paid for their time. Rates might range from $30 to $100 or more per participant for a two-hour session, though you might be able to get away with providing a free meal instead. Generally groups that are discussing business products or services are paid more than groups discussing consumer products.
You can hire a market research company to locate the focus group members according to criteria you specify, and to conduct the session using a professional moderator, while you watch from behind a one-way mirror. Or, you can do it the economical way by conducting the sessions yourself, using target buyers you've located via the phone book. Regardless of the method you choose, you must give careful thought to exactly what you want to learn.
Even better, have a prototype for people to examine and try. Assuming you have already researched secondary databases and found that the size of the industry available to you is large enough to sustain your company with a modest market share, conduct interviews with a short questionnaire.
One-on-one interviews can provide detailed insights
Rather than conducting group interviews via a focus group, you can opt to conduct your in-depth individual interviews with potential target buyers or with people who already purchase a competitive product.
To get an accurate handle on the what the market's reaction will be, at least 25 qualified people should be interviewed for each significant product difference or formula. This will provide the basis for directional evaluation and changes in the prototype, or product positioning.
For smaller companies, selecting interviewees from local surroundings (e.g., local neighborhood for a single retail store) may be the most practical alternative. For companies doing business in a larger region or nationally, attempts should be made to obtain as representative a sample of interviewees as possible throughout the company's marketing area.
The questionnaire for qualitative research should include:
- demographic information (what is their age, sex, occupation, home locale, income range, etc.)
- confirmation that they use the product or service you're testing
- which brands are used or purchased
- how often brands are purchased
- why they like different brands
- what is disliked about different brands
- discussion of different product attributes
- discussion of the importance of various product attributes
- evaluation of product prototype
Ideally, your test results should be confirmed by quantitative research or a real-market field test.
Quantitative research seeks information from many respondents
Quantitative research is a type of non-experimental market research that provides numerical measurement and reliable statistical predictability of the results to the total target population. Like qualitative research, this is original company research ("primary") on a subject in the normal course of company business ("non-experimental").
Quantitative research is distinguished from qualitative research primarily by the large numbers of people who are questioned ("sampled respondents") and the type of questions asked. Generally, sample sizes of 100 are adequate for simple "yes/no" questions to get results that are 95 percent reliable as being accurate for the entire market of buyers. To increase the accuracy to 97 percent to 99 percent, the sample sizes would have to increase to 400 to 2,000 or more, depending upon the subject matter and complexity of questioning.
For example, a company might design a prototype product that it evaluates using qualitative research through focus groups made up from a target consumer group. Once the features and benefits of the prototype have been refined to consumers' satisfaction, and communication of the product's brand positioning in the marketplace has been discussed and modified, quantitative testing may be done.
At this point, larger companies continue to refine the prototypes and may conduct a series of blind tests, in-home usage studies, and even market forecast simulations costing up to $100,000. For smaller companies, it may be less expensive, faster, and just as accurate to do a small field-study test in the real market, despite the risks that the results of the test may not translate to other markets.
To do good quantitative research, you need the following three elements:
- a well-designed questionnaire
- a randomly selected sample
- a sufficiently large sample
Even for small companies, the best recommendation for choosing the optimum sample size is to consult a professional market researcher or a nearby school with a statistics department for help in designing, constructing questionnaires, conducting the research, and analyzing results.
Well-designed questionnaires are the linchpin of quantitative research
The design of a good quantitative questionnaire depends upon careful consideration of:
- Which decisions are going to be based upon the test results?
- What key information do you need to make these decisions?
- What information was gathered in qualitative research that would be useful?
- How should test respondents be screened for demographic and lifestyle backgrounds?
- How many respondents are necessary for statistical reliability for different questions?
- How will you tabulate and analyze questionnaire results?
Questions may be posed in writing, by fax, or over the phone, but generally phone interviews have a better response rate. If you use the phone, you will want the telemarketer to use a script, to be sure that each respondent is answering the same questions.
Quantitative questionnaires are similar to qualitative questionnaires but usually gather more information that can be numerically tabulated with significant statistical predictability. Questions should be based upon "common sense" and good communication practices. All questions should be directly related to providing useful information for decision-making. For example, a questionnaire could include:
- demographic information (age, sex, occupation, home locale, income range, etc.)
- confirmation that the respondent uses the product or service you're testing
- which brands are used or purchased
- how often brands are purchased
- why the respondent likes different brands
- what is disliked about brands
- importance of different brand images
- ranking of brands by preference
- whether price makes a difference to the frequency of purchasing different brands
- evaluation of different product attributes
- ranking of product attribute importance for buyers
- evaluation of brand positionings and advertising
- purchase intent on a five-point scale (definitely, maybe, indifferent, maybe not, definitely not)
- brands that would be replaced by the new product prototype
Construct questions that allow test respondents to easily understand and answer them. Questions should be ones that your targeted test respondents will most likely know the answers to and would be willing to provide information on. Avoid:
- vague questions;
- non-useful background questions;
- trick questions;
- questions that ask for two pieces of information in a single question; and
- questions outside the expected knowledge and experience base of respondents.
Small companies often provide simple questionnaires to customers when they come into the store or purchase products and services. They may use the questionnaires to obtain a qualitative "pulse," or check, mainly to verify that nothing is going terribly wrong in their day-to-day operations. Or, they may use the questionnaires to measure the effectiveness of local advertising media in generating store traffic. Customer database-building is another possible objective.
Over time, you may obtain results that are almost as good as quantitative test results, particularly if you ask simple "yes/no" questions (on customer satisfaction, for example).
Validity of quantitative research hinges on sample selection
When conducting quantitative market research, there are two ways to select test respondents:
- probability samples (randomly selected samples)
- non-probability samples
With probability sampling, each individual in the target population has an equal chance of being selected for testing. This means that test results have a better chance of being representative of the entire target population. In a non-probability sample, the selection of the respondents is not random. For example, if you went door to door on a Tuesday morning until you got 50 people to answer your questionnaire you would be using non-probability sampling. Rather than including information from the entire target population for your product, your responses would be skewed in favor of those who would be home on a Tuesday morning (most likely stay-at-home parents or a retirees.)
Many small companies utilize only non-probability sampling methods in their research. This may be due to budget constraints or historical practice.
But the difference between probability and non-probability methods can be significant. While only probability sampling provides a true representation of the total target population, accurate predictability, and distribution levels, non-probability sampling is not necessarily bad when doing research and does not automatically render your results invalid. This is especially true if you have taken the time to consider what market segments are likely to be your best customers. In the example given above, if the most likely users of your product were young parents, then your results might be very relevant.
Aim for the largest feasible number of respondents
If you're doing quantitative market research, in most cases, the sample size for the number of respondents you'll test is determined by your available budget and by the confidence levels that you desire or can accept. The larger the sample size, the greater degree of accuracy, not only for predictions of total population behavior, but also for the degree of variation in that behavior.
This is the basis for determining confidence levels in predictability of the test base compared to the entire target population. The larger the sample size, the smaller the standard error—the possibility that the test results will not mirror the behavior of the target population. On the other hand, if your sample grows beyond a certain size, you will not greatly increase your accuracy level, but you will definitely incur more research costs.
Some Statistical Background
At least 100 test respondents should be selected from a probability sample for all quantitative tests with the objective of 68 percent to 95 percent confidence levels in predictability of test results.
When sample sizes are at least 100, if the results are quantified and displayed on a graph, the results will tend to approximate what is called the "normal curve" of distribution. That is, the majority of people will give you an "average" response, a smaller number will give you a "below average" or an "above average" response, and a very small number will give you an "exceptionally below average" or an "exceptionally above average" response. This distribution is also known as a bell curve. The mathematical probability that a given test observation will fall within a range of values from the middle of this normal distribution curve is called a "standard deviation."
There is a direct relationship between your sample size and the degree of reliability, based on the statistically predictable behavior of respondents' test results clustering in the pattern of a normal curve. This is the basis for quantifying the confidence level of test results, e.g., stating that you can have a 95 percent confidence level that your test results mirror the general population.
Mathematically, under a normal curve, 68.3 percent of all observations fall within plus or minus one standard deviation of the middle of the curve; 95.5 percent of test observations fall within two standard deviations of the middle of the normal curve and 99.7 percent of test observations fall within three standard deviations. The key point is that the larger the sample size, the greater the probability that the test results will fall within one to two standard deviations of the middle of the normal curve of population behavior.