Your online survey is likely taken on a mobile device

A rapidly increasing share of quantitative market research surveys are taken via smartphones. It is a natural development in a world that is becoming increasingly mobile in every aspect. The planning and designing of your market research studies need to take this into account. 

Replying to survey on a smartphone is a natural development for consumers, as our little digital pocket device has become our favorite tool for almost everything.

Still, much market research is designed as "online" research creating a somewhat false distinction between "online" and "mobile" research: today, they are essentially the same, as even research conducted in panels where respondents are invited to participate via email, is replied to from smartphones. And this number is increasing rapidly.

That means basically all digital surveys should be device-agnostic, and the simplest way to ensure that is to design it as mobile surveys; if it works on mobile, it works on desktop, too.

There are three essentials to remember when designing your research for mobile:

1. Minding the survey length

2. Designing respondent-friendly questions

3. Using visuals right in mobile research.

How to determine the right survey length

Respondents time for your survey is limited and you know about survey fatigue; studies into respondent behavior show very clearly that respondents spend more than double the time on question 1 to 10 than they do on question 25-30. The lack of time spent considering their answers damage the quality of their replies.

Thus, to make a user-friendly questionnaire, you need to think of a minimum viable setup. This means your surveys need to serve one purpose instead of many, and you need to break up - once and for all - with long surveys. The upside? You will get better insights from it. Killing darlings is always difficult but remembering what the key insight you are looking for is, makes it somewhat easier.

Next is the question of how survey length is measured. While the number of questions matters, so does each question's complexity. Naturally, it takes a lot less time and mental energy to state one's age than to fill out a grid with ten questions and a five-point Likert scale. That is why the optimal survey length should be measured in minutes rather than the number of questions, and a good rule of thumb is that your survey should not take more than 10 minutes to complete - and shorter is always better.

How to design good questions for mobile research

You get answers as you ask, the old saying goes. So when designing how to ask and what answer options to use for your mobile research, you need to be mindful of your respondents' time and understanding - and that you do not accidentally bias respondents either. Juggling this along with stakeholder requests and expectations means the questionnaire quality risk being determined by compromise and lack of time.

Common mishaps in questions include asking for more than one thing in the same question, including professional language that respondents are unlikely to know, and long complicated instructions. For answer options, common mishaps include answer options not being mutually exclusive, one-sided (e.g. asking for associations and then only include positive or negative adverbs), or including so many answer options the respondent has to scroll far away from the question to see them.

Unfair, but true, it is also worth noting that your questions and answer options may be good but if the survey is too uniform, respondents are also bored quickly. Likewise, if you have too many open-ended questions as respondents will have to use a long time to reply. Thus, a mix of question types is preferable.