Recently, I’ve been advising on survey design for a consultation exercise and emphasising 2 key points:
- Start with the end in mind
- Keep it simple
I wrote a while ago about lessons learnt for designing customer and staff surveys. I described 5 mistakes people make:
- Failing to link survey questions to survey objectives
- Asking too many questions
- Designer bias
- Analysing answers to individual questions
- Poor characterisation
The second of those lessons was reinforced when I read this article on the Harvard Business Review Blog. Learn more by asking fewer questions describes some research where response rates for surveys with fewer questions were over 11 times better than for a “full” questionnaire. When one factored in the declining quality of “tail end” answers — people clearly just “box-ticking” the final five or six answers to be done with it — less proved to be more.
They concluded: Getting a robust response for the five most important questions was, on every dimension, far more valuable than puny responses for the top twenty questions.
This also links strongly to one of the other mistakes I wrote about – failing to link the questions to the survey objectives. In my experience, “woolly objectives” leads to surveys with excessive numbers of questions, simply because the designer hasn’t thought about what the survey needs to achieve.
The other issue with “big” surveys, is that they annoy people! Not the greatest way to engage with staff or customers.
You can find my “Quick Guide” to survey design here. (PDF)
Download my full article here. View the presentation below: