22 Jun Effective Mobile Survey Design for Market Research
Market research has recently reached a tipping point. The ubiquity of smartphones and tablets has given market researchers the ability to gather insights from anyone at any time anywhere in the world (with an Internet connection). It’s never been easier or more convenient to reach such a large portion of the population with mobile methodologies, and companies have taken note. The latest GRIT Report claims that 67 percent of research buyers and suppliers are currently using mobile surveys, and another 23 percent are considering them. While software companies are making some strides in mobile compatibility, researchers will have to change the way surveys are designed and implemented to account for the growing number of mobile respondents. Here are the top four things to keep in mind when designing a mobile-compatible survey:
- Keep it short
This applies to both question length and survey length. Question text should be as short and as simple as possible so as not to force the respondent to scroll down the page. The length of the survey should also be as short as possible and only one question should appear per page. The correlation of survey length with abandonment rates was tested by Carey Stapleton in 2013 on desktops and smartphones. She found that abandonment rates on a 44-question unoptimized mobile survey reached 15.9 percent, while a 26-question survey that was optimized for mobile had an abandonment rate of only 6.3 percent. Clearly, making a survey concise and mobile-compatible can lower recruitment costs and save respondents time.
- Use small or no pictures
Pictures and graphics, even when shrunk to fit a smartphone screen, can be unassumingly large files. Most smartphones still do not have the processing power to load large pictures as quickly as desktops and will force the respondent to wait while pictures and graphics load.
- Use multiple choice and drop-downs whenever possible
Question types become extremely important on mobile screens as they impact user experience and response validity. The two most compatible question types on a smartphone are drop-downs and vertical radio buttons. Radio buttons should be displayed vertically as opposed to horizontally to prevent the respondent from having to scroll to the side and potentially missing one or more answer options. One study by Peytchev and Hill in 2010 tested horizontal radio buttons on a smartphone and revealed that 23 percent of respondents either did not know there was another answer choice to which they had to scroll or that it was too much effort to scroll and see it.
This notion was further illustrated by Carey Stapleton in 2013 in a survey that tested horizontal radio buttons, some of which were hidden off-screen. Respondents were asked to rate their satisfaction with the survey and were split into two groups: Some saw the unsatisfied side of the scale first while others saw the satisfied side of the scale first. Even on a desktop, users will naturally pick the first option displayed slightly more often – 5.7 percent in this case – but this effect was exaggerated with horizontal radio buttons on a mobile device and elicited an 11.6 percent difference in responses from the two groups. Smartphones allow much more room for vertical radio buttons, which should be used whenever possible because of the decreased likelihood of having to scroll to reveal hidden answer options.
One of the hardest question types to display on a mobile screen is the matrix. Matrix questions should not be displayed in a grid with the traditional horizontal Likert scale on mobile screens as the options and labels will end up too close together and might require horizontal scrolling. Matrix questions are best displayed as drop-downs that reveal the answer options, or as vertical radio buttons with one question/statement per page. Use embedded data and display logic to display mobile-friendly questions when necessary.
Traditional horizontal Likert scale matrix:
Matrix converted to drop-down with Likert scale:
Matrix converted to vertical radio buttons:
- Test on multiple devices
It is imperative that researchers test their surveys on multiple devices, operating systems and browsers to get a sense of the compatibility of their questions on different screen resolutions and in browsers with varying capabilities. One method of measuring the compatibility on different devices is to include a question in the survey that asks for the respondent’s device type. If there is a noticeable difference among drop-off rates, this likely means the survey is not displaying correctly on some devices.
Try out these four steps on your next survey and make sure you are able to test the results by using metadata or embedded data to track the type of device each respondent is using. Let me know how it goes and if you have any tips or tricks to add to the conversation!