In the following post, D’Amore-McKim School of Business Assistant Professor Keith Smith highlights the importance of strategic survey design and the impact it can have on data-informed decision making.
Quality marketing and management decisions are increasingly driven by the data available to managers. Data is everywhere though, and the volume of data managers have access to has grown to colossal proportions. Furthermore, the quality of that data is often in question. Quality decisions require quality data, and how managers can obtain high quality data from their customers should be forefront in their minds.
Often, customer insights are obtained from customer surveys. Survey research methodology is widely used in marketing, and it is important for both marketing managers and marketing academics to follow stringent guidelines to ensure that meaningful and valid insights are attained.
In a recent article published in the Journal of the Academy of Marketing Science, my co-authors and I outline those guidelines and assess how well marketing academics have stuck to them. While some survey guidelines differ between academics and practitioners, there are a number of suggestions that are relevant for both groups:
- While it may seem obvious, marketing managers would do well to ensure the right customers are being surveyed. Too often surveys are administered to customers who do not have the required knowledge to provide helpful information. Perhaps customers without a car are surveyed about car insurance, or young customers are asked about hospice care. Selecting survey respondents carefully, and then accurately reporting when inappropriate respondents are eliminated, has an incredible impact on the quality of a survey.
- Hand in hand with surveying the right customers is asking the right questions. Frequently, surveys are administered without pretesting the questions, or in the interest of brevity, survey questions sacrifice clarity and completeness. Managers who hope to gain clear insight are best served by fully testing questions with a pilot sample and including more than one question per concept to gain a clearer picture. While some statistical techniques exist to measure survey biases, a more thoughtful design of the questions almost always provides a higher quality survey.
Customer insights are increasingly critical in the modern era of hyper-communication. Yet obtaining those insights remains as elusive as ever. In fact, the explosion of data has made it more difficult to make sense of those insights. Successful analysis and customer insights must start with high quality data, and that data depends on successfully designing customer surveys to more accurately capture the truth of customer opinions.