How much longer do I have to go to complete this thing? As someone who is actively involved in online survey creation, not to mention someone who enjoys taking surveys, I am greatly concerned about survey length and how we convey that length to our participants.

One of the more common ways of setting expectations is via the survey invitation or the welcome page. It is best practice to set expectations up front before asking the participant to become a respondent. Typical language might include:

"This survey will take less than three minutes of your time to complete"

If the survey is longer than 3- 5 minutes, we have to be concerned about respondent drop off. This is particularly the case in B2B marketing research where respondents have very limited time. There is no magic time limit, and it depends upon the survey’s ability to engage the respondent, and the respondent’s level of engagement with the brand. Suffice it to say though, the longer the survey the lower your completion rate will be.

Another method for reinforcing the concept of “making progress” toward survey completion is the use of a progress bar or similar option. The graphic below shows the user they are 73% complete. Other options include a page count (e.g. Page X of Y) or a progress bar without a percentage. The final option is to not include a measure at all. The latter approach seems to be common amongst the large panel companies who frequently send out very lengthy surveys.

Does the inclusion of such a measure increase completion rate or make the process more enjoyable for the participant? The first question can be addressed via an A/B test, preferably a series of tests over multiple surveys. The second question can be answered with a question such as “Did the inclusion of a progress bar make the survey experience more enjoyable?”

Progress bars can be useful tools in encouraging respondents to complete the survey. Test them, before making a wholesale conversion.