Market research: Questions and questionnaires
Questions and questionnaires
Whenever market research is undertaken, there are six crucial questions - to which you must have appropriate, detailed and supportable answers prior to commissioning the research:
| The Sample | Who are you going to ask? |
| The Method | How are you going to ask them? |
| The Questions | What are you going to ask them? |
| The Results | What will you do with the information? |
| The Cost | How much do you want to pay for the answer? |
| The Time Scale | By when do you need the information? |
These are explored in detail below and in general apply equally to personal and corporate markets:
The Sample
The number of people asked is an important consideration as if you asked just one person you would get a very accurate picture of her views on anything (probably) but this would not necessarily be representative of the world at large and may lead you down a completely blind alley. How many then should you ask? and what sort of people? How many do you need to ask - is it millions, hundreds of thousands or what? Clearly there is a cost implication as the number increases; both in terms of carrying out the research and also in analysing and interpreting the resulting data. Fortunately a lot of research has been carried out into this area and some surprising conclusions have emerged. When you ask a number of respondents (known as the 'population') the same questions, after a certain number the percentage difference in the answer ceases to vary very much, or at least if it does then the degree of likely error can be calculated with a high degree of accuracy.
This number is known as a 'statistically significant number' and although it needs to be calculated for each type of question, the numbers are surprisingly low. For consumer goods it is in the low hundreds and even for such emotive issues such as politics it is only in low thousands. Key points are that the sample must be homogeneous i.e. sharing the important characteristics - e.g. four-wheel car drivers in the home counties, or British owners of Portuguese villas in the Algarve; in order that the data is comparable and that conclusions are meaningful and also of course targeted. This classification is known as segmentation (see separate paper 'Segmentation - slicing and dicing the cake of customer spend').
The Method There are several methods of obtaining the information as mentioned above and key characteristics of each are explored below.
Telephone research
Using the telephone to collect has one great advantage - it is cheap. One researcher can make many calls in a day without leaving an office. It is also both very focussed as you are initiating the call and it is fast as interviews do not take long and the elapsed time to complete the exercise is also short.
There are some drawbacks however - often people do not like to receive unsolicited calls and it can be very difficult to use in a corporate - business to business - environment. It relies on a structured script and on obtaining answers in the same manner (e.g. interviewers tick boxes on template in front of them).
Written questionnaires
This is probably the most common method of research and everyone will be familiar with it in one form or another. Unless used in the right circumstances it can be a passive method reliant on people to complete and return them and in this case inertia rules and only those with a grievance or an axe to grind return them. By combining it with other activities however, such as checking out of an hotel, installing new software, sending in the guarantee form, or making it part of the application process for loans or life insurance and other services it can be turned into an effective method of research.
To be really effective it is best to use questionnaires that ask for boxes to be ticked or strength of agreement to statements to be indicated (see below). This has the advantage that it is easier and faster for the recipient to complete and also allows direct comparability of answers. It will not perhaps have the same depth as, say, a qualitative survey where respondents write comments; but if it is drawn up well then it is very useful and can cover more subjects than the qualitative type.
Types of questionnaire
There are several types of questionnaire and each is designed to explore different aspects or elicit different responses. Some of the more common include:
- Dichotomous Multiple choice
- Importance
- Bipolar
- Likert
- Rating scale 1 - 5
- Buying propensity
These can be used in any combination as long as the questionnaire is not too long and it is focussed.
In the fictitious examples below the data is merely illustrative.
Dichotomous
This is a fairly typical basic type of question, not too intrusive and merely asks you to answer yes or no. As such it cannot assess the degree of feelings in between the poles:
Multiple choice
This is a question offering three or more answers - and allows a greater breadth of response.
Importance
In this type of question the respondent is asked to rate the importance of an issue to them on a scale of 1 to 5
Bipolar
The question asks for a response to be marked between two opposite ends of the scale:
Likert
This question examines how strongly the respondent agrees with a statement and can help assess the feelings of customers towards issues. 
Rating scale
This question type rates the replies in terms of a scale from e.g. poor to first class. As with all these types of questions it is sometimes necessary to have an even number of boxes (e.g. 4) to avoid the middle of the road response commonly taken by those trying to avoid making a stand.
Buying propensity
This type of question is trying to elicit a customer's future intentions by asking whether they might buy a product and can help assess the needs and likely take up of a new product if developed. Care needs to be taken with these questions as they may reflect wants rather than needs!
All of the above are quantitative type questions. What they ask is for a response within pre-defined parameters that allows input into spreadsheets and hard analysis.
Although this facilitates the input into data analysis sheets and subsequent number crunching - the respondent is not allowed to say what they think. They can only answer the question by marking the pre-designated boxes. This is of course of immense use - especially if the questionnaire has been well thought through and piloted. Sadly this is often not the case and many are rather poor! As a result you do not get the qualifying comment that can often express her real feelings.
Qualitative questions can allow more freedom for answers but are much harder to analyse as each respondent will use her own words. Often the question will be couched along the lines of:
This has the advantage that the respondent can say what she likes, which can yield very interesting information that might not have been thought of at design phase; but on the other side, that she can also respond in an unlimited and often unconstructive manner, making analysis much harder.
Whichever question types are used they must always be designed with the express intention of:
- inconveniencing the customer as little as possible;
- being aimed at an homogeneous segment; and
- having been designed to elicit specific information that supports your marketing initiative.
~ Neil Jones ~
If you have a requirement for training or consultant please call us now on 01565 653330, or use our
simple enquiry form.