Previous Page Table of Contents Next Page

3 Researching consumer attitudes to your products

Main points in Chapter 3
Researching consumer attitudes to your products

Before beginning commercial production you must be sure that consumers will want to buy your products.
The chapter reviews ...

It is important to test people’s reactions to your products before going into full-scale production.


As we have seen in Chapter 2 and will discuss again in the following chapters, much market information can be obtained from observing the existing situation in shops and asking questions of retailers and wholesalers. However, this type of research will be wasted if, when you come to sell your products, people do not buy them. Before you invest in a factory you must be sure that people will want to buy what you are going to produce, and that they will keep on buying it. To achieve this confidence it is desirable to both carry out surveys of consumers and to arrange for potential consumers to taste your products and give their opinions.


Consumer surveys are used to find out:

Consumer questionnaires are mainly used when the type of product to be sold is already well-established on the market. If people have never tasted nor used the type of product you plan to produce then it will be difficult for them to answer questions about it. For a product with which most people are familiar, such as juice or jam, it is a relatively easy matter to ask questions and get useful information.

Carrying out a survey

Before you carry out a survey of potential customers you need to decide who to survey and how you are going to carry out the survey. If you decide to use questionnaires, then you also have to design the questionnaire so that it will provide you with the information you need.

In developed countries interviews by telephone are widely used. However, this is rarely appropriate for developing countries. Normally you have to interview potential customers in person, either individually or as members of a group. For small and medium-sized processors individual interviews are probably easier, as organizing to get groups of people together can be rather difficult. Here we shall concentrate on individual interviews, although you could consider inviting groups of people to meetings, at which you could discuss their attitudes to the products you plan to compete with. You could also use such groups to discuss possible “brand” names for your products (see Chapter 6).

It is clearly not possible to interview all potential customers. For that reason you will have to take a sample. The question of sampling is fairly complex and the subject of much debate among market research professionals. For fairly small organizations it is necessary only to observe a few basic rules:[4]

Carry out interviews in a relevant location. If you have no idea about where you will be able to sell your product then you may need to carry out a random survey of the entire population in a particular area. However, if you are planning to sell your product through supermarkets it may be a good idea to carry out interviews of people as they walk out of a supermarket.

Make sure you carry out interviews in a systematic way. Standing in the street or at a marketplace and talking to people you see there may produce biased results. For example, if the interviewer is a man he may choose to interview the most attractive women; conversely, a woman may choose to interview the most handsome men. Neither approach is likely to give you an accurate view of people’s ideas about the product you plan to produce. You need to decide in advance how you will choose the people you interview, using simple sampling techniques.

Use simple sampling techniques. The simplest method is to interview a percentage of people in a structured way. For example, if you decide to interview ten percent of all people leaving a supermarket on a particular day then you could interview the first to leave, the eleventh, the twenty-first, the thirty-first, etc. This removes the possible bias caused by the interviewer choosing to interview someone just because they look nice or because they look friendly and willing to cooperate. However, a problem with this method is that it is difficult to interview someone and count people leaving a store at the same time, so you may have to employ two interviewers.

Limit interviews to potential customers. You should concentrate your interviews on the people you feel are most likely to buy your products. For example, it is probable that most purchases of iced lollies are by children (you should confirm your ideas about potential customers by talking to shopkeepers first) or by parents for their children. It is more useful to interview children, either alone or with their parents, as they leave the shop. Even more obviously, children are unlikely to buy beer, so there is little point in interviewing them about beer. In the above examples, therefore, you could survey every tenth child about iced lollies, or every tenth adult about beer.

Systematic interviewing, such as asking questions of every tenth person ...

... avoids the risk of the interviewers picking the most attractive people to interview!.

Stratify. Let’s suppose that you interview every tenth child leaving a supermarket in order to ask questions about iced lollies. At the end of the day you find you have interviewed fifteen girls and five boys. There may be reasons for this, unrelated to iced lollies. For example, the shop may be close to a girls’ school or parents may be more likely to ask their daughters to do shopping for them (girls may go to the shop to buy other things). However, a sample of fifteen girls and five boys may not be representative of all children shopping in all shops where you could sell your lollies. You can overcome this problem by taking a stratified sample in which you plan to interview ten girls and ten boys. You start out by interviewing every tenth child until you have spoken to ten girls. After that you interview just boys. Other ways of stratifying include age and income. For example, you could further stratify the sample of children by planning to interview five boys and five girls under eleven years old and five of each over eleven years old.

Minimize the cost. The amount of consumer research you can do is, of course, related to the expected size of your business. The number of interviews you need to do will depend on how large an area you intend to supply and on the amount of money and time you think you can afford to spend on interviews. You will have to do more interviews if your product is something that is not purchased by everyone than if it is a product that everyone buys. For example, if you interview 50 people you may find that almost all eat bread but that only ten eat yoghurt. After a time, you will discover that you are not getting much, if any, useful information from additional interviews and it is time to stop. Interviewing people in a public place is more cost-effective than visiting them in their home, although if you need to ask a lot of questions home interviews may be more appropriate (people are unlikely to want to stand in the street answering your questions for more than a few minutes). Random sampling of people according to their addresses is expensive because of the costs of going from one home to another. You could concentrate interviews in a few randomly selected areas (i.e. cluster sampling)[5] or just carry out interviews in those parts of town where you think you will be able to sell your product.

Designing questionnaires[6]

Questionnaires should be designed in such a way that the information obtained from each interview can be compared and much of it can be put in the form of tables of data. If you are planning a small business and will do the interviews yourself you still need a structure for the interview. If you are going to employ people to ask questions you should give them a very clear questionnaire with instructions to ask every question exactly as written. The questionnaire should be clear and easy to understand and as brief as possible. Apart from a few polite opening comments it should not contain questions that will not be of use to you in planning your business. In planning the questionnaire you need to:

Types of questions that can be used include:

Opening questions. These are polite introductory questions to put the person being interviewed at ease. For example...

“I am doing a survey about the fruit juices that people like to buy. Do you mind helping me by answering a few questions?”

Closed questions. These questions have “yes” or “no” as answers and can be used to establish basic facts. For example...

“Do you drink orange juice?”

Multiple-choice questions. These require the person being interviewed to choose one answer from a list of possible answers you have pre-selected. For example, a number of answers are possible for the question...

“How often do you buy orange juice?”

among these are...


more than once a week


about once a week


once or twice a month


less than once a month

or you may ask...

“Which orange juice do you like best?


ABC juice


XYZ juice


Joe’s juice

Scale questions. These are a type of multiple-choice question where the person being interviewed can give answers from a scale of possibilities, such as...











You could also ask people to rate the importance of a feature of your product. For example...

“Sugar content in fruit juice is to me”:

extremely important

very important


not very important

not important






Other approaches include...

ABC Juice is


very good










ABC Juice is better than Joe’s Juice

strongly agree


neither agree nor disagree


strongly disagree






Note: The advantage of multiple-choice questions is that they simplify the statistical analysis of the questionnaire. The disadvantage is that they force people to give answers they may really not want to give. For example, in response to the question about which juice is preferred some people may want to say, “I like ABC juice and Joe’s juice the same but XYZ juice is awful.” (In this case the problem could be avoided by changing the question to “which of the following juices do you like?”, although, of course, the information obtained will not be exactly the same.)

Open-ended questions. These are questions where the interviewer writes down the answer the respondent provides. It is difficult to do any statistical analysis of the answers to such questions but they can provide useful information, for example by revealing issues that had not been thought about before.

Combination multiple-choice and open-ended questions. These are like multiple-choice questions but have a blank space for a reply not foreseen when the questionnaire was designed. For example, answers other than those listed are possible for the question...

“What is the main reason you buy ABC juice?”


the taste


the price


the packaging


company reputation


others my children like it

The questionnaire needs to follow a logical sequence. For example, you need to establish that people do, indeed, buy fruit juice, before asking them what types of juice they buy. The following sequence is logical...

1. Do you drink fruit juice?
2. Which brand of fruit juice do you normally drink?
3. What do you like about this brand?
4. What would make you consider buying a different brand?

... whereas the next sequence is not logical and is cumbersome...

1. What would make you consider buying a different brand of fruit juice to the one you already buy?
2. What do you like about the brand you drink?
3. What brand do you normally drink?

Once you have designed a questionnaire, you need to first try it out with a few people. Do they understand the questions? Can they answer them without any difficulties? Are they willing to wait standing up for the time it takes to answer the questions? After this test you should amend the questionnaire, if necessary, before doing a further test in a market or outside a shop.

Note: An example of a consumer questionnaire is given in Annex 2.


When to have your products tasted

It is a good idea to carry out consumer surveys before you organize tasting. The surveys may tell you a lot about what people expect from a product, allowing you to change your product’s formulation to meet their preferences.

The need for tasting surveys will depend on the products you plan to market. Foods that taste more or less the same everywhere require little or no advance tasting, but foods that depend on the formulation and on different ingredients do.

Some points to consider are:

Organizing tasting tests

Unless you already have a processing facility and are just planning to add another product to your range, the products you prepare for tasting purposes will probably have to be made in your experimental kitchen, if you already have one, or in someone’s home. For this reason they will not taste exactly the same as they will when you begin commercial processing. You should therefore try to create a situation in the experimental kitchen as close as possible to commercial production. For example, if you think that the fruit you will be using will take two days to reach you from the farmers, don’t use freshly picked fruit for your trials. If your factory will eventually use large caterers’ packs of sugar, make sure that this sugar is the same as that used in your experimental products. Use cooking utensils of the same metal as those you will use in your factory. There are many factors that can influence the taste of processed foods.

The tasting survey you organize will depend on the products you are planning to produce. For example, if you are going to make jam this is a product with which most people are already familiar. You can then ask them to compare your jam with other jams that they already buy (see Figure 2). On the other hand, if you are planning to produce a product that is unknown in your area you will need to find out not only whether people like the product but also whether they would buy it.

When organizing tastings there are several important things to take into account:

Figure 2
Example of a tasting survey form*

For the following characteristics please put a tick under the better jam

Jam A

Jam B



Fruit content


Please give 2 ticks to the jam you liked more and 1 to the jam you liked less

Please put 1 tick under the jam(s) you would consider buying and a cross under the jam(s) you would not buy

*although this form shows a survey for jams, any product can be substituted

Give traders the opportunity to taste your products

When visiting wholesalers and retailers it is useful to take samples of your product. Traders usually have a good idea of what appeals to consumers and can advise on whether your product would be appreciated in their area.

Consumer reaction to packaging and labelling

Tasting surveys should come at an early stage in the development of your business, as there is little point in making further investments if people do not like your products. At this stage you are unlikely to have decided on the packaging and almost certainly will not have designed a label. However, you could ask tasters one or two questions about packaging to help in your planning. Where there are alternative types of packaging you could ask, “in what type of packaging would you prefer to buy this product?” In the case of fruit juices, for example, the possible answers could be plastic bottles, glass bottles, cartons or cans.

After you have decided to go ahead and make the product, you need to finalize plans for the packaging and label (see Chapter 4). When you have chosen the type of packaging you want to use and have a draft design of a label, you could consider doing a further consumer survey to find out whether people like your plans. If possible, you should give people two or more alternatives to choose from, rather than simply asking them whether they like your proposal or not. You could also have samples of competitors’ packaging and ask people to compare these with your prototype.


Interviewing consumers and carrying out tasting surveys should give you a good idea of:

You can then reach conclusions, such as:

When you have decided that people will want to buy your new product, you then need to turn your attention to getting detailed information about the market, as described in the next chapter.

[4] Large companies who wish to understand the topic in more depth may wish to refer to Chapter 7 “Sampling in marketing research” in FAO, 1997 (see “Further reading”).
[5] see FAO, 1997 page 74
[6] see FAO, 1997 Chapter 4 for more detail.
[7] but see the story in Chapter 2 page 19
[8] See Coetzee, H., 2001 for an interesting article on the testing of new food products on illiterate and semi-literate consumers.

Previous Page Top of Page Next Page