Asking open-ended questions in your online surveys is a good idea for many reasons. Ultimately, they allow people to respond in their own words, providing color and richness to the data that helps tell the story behind the numbers. Open-ended responses, commonly known as verbatim in marketing research, also aid data quality and integrity, allowing us to identify respondents who did not take the survey seriously.
To effectively make use of open-ends, you need to find the best point in the survey to place them. One of our favorites is asking a follow-up "Why did you give that rating?" to our client's Net Promoter Score (NPS). Marketers use NPS to benchmark their brand over time and also to measure themselves against competitors, but the real power comes from the follow-up open-end. By asking respondents to explain their rating, we uncover the reasons why Promoters love the brand and why Detractors would not recommend it.
Open-ended data is unstructured in nature - the information can't be interpreted quantitatively until we provide structure. To do this, we read and code every single verbatim response. Many researchers rely on computer software to sort the responses into buckets quickly; however, Evolve prefers the old-fashioned approach of manually viewing each answer which allows us to understand the data intimately and apply context when interpreting the responses.
Manually reading each response also helps us identify cheaters - those who sped through the survey without answering truthfully or carefully. If a respondent skips the opens or provides "throw-away" answers, there's a good chance the rest of the data is low quality as well. It's one more tool in our arsenal to ensure the highest quality data and responses.
As with all questions in your survey, use open-ends strategically. We recommend asking just 2-3 opens, as they take longer to answer than multiple choice or rating questions. Too many opens will frustrate respondents and negatively impact data quality. Asking specific questions is crucial, as they will give you much richer responses than a generic "additional comments" that many surveys employ.