"Good" Surveys - The Importance of Designing an Effective Survey

Surveys can be extremely useful tools to gather information about our customers and ways wecan improve to more effectively meet their needs. However, irrelevant data is just about as useful to us as researchers as no data at all. So how do we determine what makes a good survey that will provide us with relevant, representative, and beneficial information? It starts at the beginning, with the development of the survey itself.

Design With the End in Mind

There are many factors involved when developing a survey and this begins with forming a clear research objective. Throughout the duration of the study, weas marketing researchers need to frequently reference back to this objective and make sure that the research we are designing is helping us to achieve this goal. When developing a survey, if a question is not goingto provide us with information that is relevant to our objective, it does not need to be included in the survey. Developing these questions may involve several rounds of reviews and draft comments from relevant decision makers before being finalized, and it is important that we have several sets of eyes on these drafts to help us remain objective and focus on the end goal.

What Makes a “Good” Survey Question?

We first need to consider our research objective and start to outline broad topics that we want to cover with our questions. From there, we can begin to narrow down specific questions based on these topics. A mix of categorical, metric, and open-ended questions can be used in a single survey, however for ease of data analysis the bulk of these should be categorical and metric, always with the end goal in mind of meeting our research objective.

Do’s and Don’ts of Good Survey Questions

Good survey questions should be simple, clear, and concise. While generating these questions, we needto make sure that they are easy to understand andnot providing extraneous details that will confuse the user. Each question should focus on one single topic; we need to avoid “double-barreled” questions to ensure we are analyzing the data in the right way. The questions should be worded in such a way that all users are able to interpret them the same way and there is little room for misunderstanding or confusion about what the question is trying to ask. Finally, we need

to avoid “leading” questions that will encourage our users to answer in a particular way; careful wording for all questions can help ensure that the survey remains unbiased and objective.

Sample Sizes and Representative Data

Now that the survey questions are compiled and approved, we still have work to do to maximize our potential for success. One of the most important factors in a successful survey is determining our sample size and selecting our sample. It is not enough to simply email our entire customer database – we need to once again revisit our research objective and determine what parameters we are looking for in our sample. Perhaps we want to specifically target customers who have made a purchase within the past six months, or potential customers who have provided their contact information but have yet to make a purchase. From there, we need to further narrow down our sample size by systematic skip interval sampling to ensure a random, representative group. We should have a pre-determined sample size in mind, as gathering too much data can be counterproductive and costly, and make sure that the sample size is large enough to ensure we can still collect the number of complete surveys we need for accurate results.

By following these steps and designing with the end in mind, we as researchers can make sure the data we are gathering is representative, relevant, and will be of use to our decision makers and stakeholders.