Jeffrey Henning of Vovici had a recent post about “Causes of Survey Incompletes: Why Panelists Say They Abandon Surveys” (http://bit.ly/g2Xyue).
I would agree with Jeffrey and James Sallows of Lightspeed Research — the quality of the surveys is at least as responsible as the respondents themselves.
The primary reasons for incomplete surveys according to Lightspeed Research:
1. Subject matter = 35%
2. Media downloads = 20%
3. Survey length = 20%
4. Grids = 15%
5. Too many open-ends = 5%
The findings are not surprising.
Just like with e-mail and advertising, relevance counts. If the subject matter is not relevant and meaningful to the respondent, they’re less likely to respond.
Media downloads likely correlates with age. Older people likely do not have the high-speed connections of younger people and are less comfortable waiting for downloads than 18 to 34 year-olds who are used to downloading videos and music.
Survey length is critical. According to Lightspeed, incompletes steadily increase after 15 minutes. I think it’s rude to ask for more than seven or eight minutes of someone’s time unless you’re paying a very nice stipend.
Making grids mandatory causes people to drop out. They are maddening, imposing and inconsiderate. I would guess these are more prolific with less sophisticated market researchers.
I was surprised to see that too many open-ends leads to greater incompletion. Don’t make open-ends mandatory, but definitely include them so respondents have the opportunity to tell you what’s on their mind — especially those things you were not smart enough to ask about.
Answers to open-ended questions is where insights lie. That’s why I always like to end my surveys with “what did we fail to ask you that you believe we need to know to . . . ?”
What do you do to your surveys to increase response rates and decrease incompletes?