« making users feel smart | Home | Shaking Up Everything, And It’s Good »
Simple tip for your online surveys
By susan | February 20, 2009
When you’re creating an online survey, what is the most important information you wish to receive? The profile information of your respondent, or the answers to your questions?
Profile information, which includes date of birth, sex, and income, may be useful for some purposes. It’s also most likely to be skewed or unanswered on a voluntary survey. So if you ask those questions first, you risk losing people right away. What if those same respondents had answered all of your other questions and then dropped off at the end? Would you rather know their vital stats? Or would you rather be able to factor in their other reponses - which, presumably, is what you are really wanting to know?
Please, please, don’t slow down your own survey at the start! Ask the most interesting questions first, the easy questions, the ones that make people genuinely interested in following the questions to see where they lead. Then, at the very end of it, after you’ve given them something really enjoyable to do (offering their opinions on a topic of interest), you can ask politely if they are willing to share some of their vital statistics.
If you’re getting less than a stellar response rate, this could be one of the reasons.
Recently I’ve received links to a couple of surveys where I opted out right away, for two different reasons. The first one began with open-ended questions on a topic where no open-ended examples came to mind. So I was not clear on what they were asking, or how I could help. It was too broad. Generally it’s great to include some open-ended fields for response, but not so much that it looks like a blue book for an exam.
The second survey began by asking for my birthdate. Besides requiring me to use that little series of fussy drop-down menus for month and day and year, it failed to engage my attention. My birthdate is not a topic of interest for me. It also indicated to me that the survey as a whole might not be well-designed. And so I walked away.
If it starts out as an engaging topic, respondents might be willing to proceed. This is so easy. Just try that next time, and see if it works. Bonus points if the respondent thanks you for how much fun it was to answer your questions!
Topics: consumer behavior, market research, marketing, online marketing |