Today we want to talk about the most important thing in a survey. Perhaps you are expecting us to discuss the design of questions. Sure, the questions we ask are important but they are not the most important thing. The most important thing is the strike rate or cooperation that is achieved. We would argue that it doesn’t matter how good your questions are, if you haven’t any respondents, you haven’t got a survey.
And herein lies the problem. Cooperation rates have fallen like a stone over the years. It is not that people are busy, they have always been busy; it is their cynicism about the value of market research. It is not unusual in some surveys to have cooperation rates that are way below 10%. We researchers get nervous about the representation of a sample when cooperation rates fall below 50%!
There are a number of things that we can do to make that cooperation rate as high as possible. This is our list of suggestions:
It goes without saying that the starting point for good cooperation is a good list of sample respondents. By this, we mean a list of respondents who are absolutely appropriate for the survey. Of course cooperation will be poor if we are trying to speak to inappropriate people.
If the survey is carried out by telephone, the interviewers must have warm and engaging voices. People like speaking to someone who sounds interesting and charming.
Again, if the interview is administered on the telephone, the interviewer must be intelligent and trained. They must be skilled in overcoming objections and answering respondents’ quite reasonable queries. Respondents can be anticipated to ask who is commissioning the survey. The interviewer must be ready with an appropriate answer. Respondents will want to know how long the interview will take and they expect a truthful answer. They may well ask what the survey is about in order to be assured that they are the right person to speak to. An interviewer who is flummoxed by such questions will be pushed back.
If the survey is carried out online, a clear and attractive presentation of the survey is a must.
Short surveys such as those with just five questions will, for sure, have a much higher cooperation rate than those with 50 questions. Although we may champion short surveys, the reality is that many surveys need lots of questions. What is more, we researchers have lots of questions to ask. We must be careful that questionnaires don’t become unwieldly.
Whether the survey is online or by telephone, it should have a short introduction that sells the benefits of taking part. Consider these two versions of an introduction:
We are conducting a random survey of companies that have applied for factoring finance. Depending on your answers, the survey is likely to take about 20 minutes.
We are speaking to people that have applied for factoring finance to see how suppliers can improve their service to organisations such as yours. The discussion is likely to take 15 to 20 minutes.
Cooperation rate: 30%
Cooperation rate: 50%
So remember when next designing a survey – the quality of the survey is just as much influenced by the response rate as it is the questions themselves. Make sure you spend time and thought to get your response rate as high as possible.
Further reading on questionnaire design: