Can question selection improve your survey results? Check out these stats now!


Surveys are a great way to get information from customers to help you grow your business. They are like a net that catches all this fantastic data if – and only if – we ask the right questions in the right way. What does “right” look like?

We’d like to share the results of our latest study on surveys; in particular, how to create effective questions for your surveys. We think it will round out our recent discussion on surveys (see: Who knew surveys could be so easy!) and give you some solid data to help you get the most out of your customer research this year. So let’s begin!

Types of Questions

We analyzed the responses of 3000 study participants and, of the three basic types of questions, the radio button was the most respondent-friendly, with 92.8% answering those questions. Participants liked the checkbox questions almost as much with 83.3% responding, but only 67.3% answered the input text type questions. So we can easily draw the conclusion that close-ended questions are more user-friendly than open-ended ones. Our advice? Use textual questions sparingly!

Number of Questions

Of course, even a respondent-friendly question type can’t perform miracles when there are too many questions in a survey. The results clearly showed that the more questions, the less probability of completing a survey: 99,09% of surveys composed of 1-2 questions were completed successfully, as compared to only 63.6% completing a survey of 21 questions. Quite a gulf, isn’t it?

Now, the average number of survey questions in our study was 3.77 (let’s call it 4). The vast majority, 96.86%, of surveys had from 3-5 questions, a good sign that most of you have already discovered the optimal number of questions needed to get a good response. The final results showed surveys with 1-2 and 3-5 questions were by far the most popular among participants (81% of total number). Therefore we’d suggest to include max 10 questions per survey (more than 96% surveys completed).

Survey Links

First of all, if you send a newsletter to your customers with a survey link, make sure it’s clearly marked and accessible. Now to the results! We analyzed the open and click through rates (CTR) rates of messages containing survey links. For HTML messages with links, the open rate came in at 51.6% and the CTR at 17.2%. Plain text messages didn’t fare as well, with a CTR ratio of only 11.2% and open rates too low to track!

Conclusions and Actions

Let’s draw some conclusions and  formulate some helpful tips based on the results:

  • If a number of your surveys are not completed, try to find out the reason why. Maybe the questions were too long, or there were too many, or they were the wrong type for your audience, etc. Use the data from our study to analyze your surveys and create a list of best practices going forward.
  • We should all agree now that open-ended questions invite too wide a range of answers, making them very difficult to measure and quantify. Plus, they take too long to finish, so people drop out of your survey!
  • Don’t overwhelm respondents with difficult, thought-provoking questions early on. Start with some easy, radio button-type questions, then sprinkle in 1-2 open-ended questions, then back to the easy ones.
  • Add your survey link to an HTML email message for better results every time!

We hope that our report data and tips prove helpful when planning, creating, and sending your surveys. Please feel free to share with us your comments, ideas, or your own suggestions on how to take surveys up a notch!