#1 – Only survey those you want to survey.

• I recently got a survey asking me to rate my satisfaction with one of the website’s products even though I haven’t used it; I’ve only read about the product on their website.
• In a user satisfaction survey, only survey those who’ve used your product, or at least make a response option like “Not applicable” available. Better yet, survey those who’ve used your product several times so that they have more of an informed opinion.
• Sounds obvious, but the website that made this mistake is arguably the most well-known website in the world. Sometimes it’s the obvious that gets overlooked.

#2 – Keep the survey under 3 minutes

• I’ve seen lengthy surveys that measure all sorts of “good-to-know” metrics – brand recall / recognition / awareness, advertisement recall / evaluation, purchase intention, past purchases, product usage, importance of product features, satisfaction with product features, etc.
• How can you trust your results if only 1 out of 10 users invited to participate actually completed your long, boring survey? Odds are, their opinions and behavior do not reflect those of the 90% of users who did not complete your survey.
• The survey I recently got had only seven questions and took less than a minute to do. Short surveys will have a higher response rate and therefore, better quality results. Have the courage to eliminate “good-to-know” questions and stick to the essential ones.
• Note: You can’t compensate for a low response rate with a bigger sample.

#3 – Limit the survey to 1 or 2 pages 

• Many surveys ask just one question per page because doing so is easier from a programming perspective, but these surveys are more cumbersome for your users. User experience should always win.

#4 – Use emails instead of pop-ups

• Users visit your site for various reasons, and answering surveys is not one of them. Don’t deteriorate user experience through pop-up survey invitations. Use email invitations, even if it means you can only sample from those who provided their email.

#5 – Use email reminders and incentives

• Sending email reminders to encourage survey participation is a cheap, effective way of boosting response rate and therefore, the quality of your data. Make sure you keep your unsubscribe email list up-to-date.
• Offering incentives is another great way of boosting response rate, especially for websites that can offer incentives at virtually no cost. Microsoft could offer a one-month free subscription for their Office 365 to those who do their survey. YouTube could show videos ad-free for a month.
• You can also test for the incentive level that would yield the best response rate per dollar by running experiments. It’s worth the investment.