Survey Completion Rate vs. Number of Questions [New Data]

Our friends at SurveyMonkey had a great post recently about the effect of adding more survey questions on the completion rate of your survey.  We all know that adding more questions to a survey (or a landing page) can decrease the response rate because it is more work to fill it out.   They found that adding just 5 questions to a 10 question survey will drop the completion rate by another 2%.

But, what surprised me most about this was that the dropoff rate was not that big.  Now, this only included people who started the survey (submitted 1 page of results) so there is a bias in the data. There is a huge drop off in people who open your survey but never answer any questions, and that is not shown in this graph. However, this seems to say that if people start your survey, they are pretty likely to finish it.

What does this mean for marketers?

1. Keep surveys and landing pages short , or at least make sure you are only gathering the information you really need. This will help you get more responses.
2. Measure the conversion rates your company , because the conversion or completion rates can vary a lot.  I will say that even though our forms (when we use them) at HubSpot are somewhat long, I am pretty happy with our conversion rates so we keep the longer form because we do use and value the information we receive.
3. Overall, the drop off rates were not huge.  So perhaps for surveys you should just ask all the questions you need to ask, and make sure you get the right data.  Cut the really unnecessary stuff, but if you get value from it, keep it.

What do you think of this data?  Have you seen a similar effect on your surveys and landing pages?