Surveys have a great new platform: smartphones! We have access (we can get to people pretty much anytime) and identification (we can be pretty sure who is taking the survey). It’s all for us to screw up. Is the survey industry going to do what it always does? As Betty Adamou pointed out in her paper at the Newmr (www.newmr.org) conference last year, survey research takes a communication medium and beats it to death. Telephone? Web? All victims of over use and poorly designed, boring, and long surveys.
There has been a “140 characters per question text limit” meme floating around recently started by Annie Pettit from Conversition Strategies. This made me think, “How can we apply this idea to smartphone surveys?” The 140 characters for the question text makes sense, but what about the rest? So here are my proposals for NOT destroying the smartphone platform:
10 questions per survey. This seems enough to me, about 5 minutes per survey. Given it that the device is mobile, and the places from which respondents will respond now varies quite a bit, 5 minutes of attention seems pretty reasonable.
7 alternatives per question. The psychologist George Miller pointed out decades ago in the “Psychology of Communication” that 7 alternatives is about all we can cope with. He called it the “span of immediate apprehension” and this construct has stood the test of time (more or less). Seven alternatives should be enough. So we have to give up the 25 brands we usually ask about, did we really think people remembered that much about what they bought? No more endless grids, in any case they are impossible to navigate on a smartphone. No more 11 point scales! Yes!!!
140 characters per question. 140 characters is actually quite a lot. You should be able to convey enough in 140 characters. If you can’t then there is the risk the respondent gets lost along the way. This is a smartphone, remember, so the screen size is limited. Twitter has grown massive on 140 characters, so at least there is precedence.
We can’t let smartphones fall prey to the destructive tendencies of current survey practices. The 10-7-140 principle gives a framework for engaging survey respondents, not mistreating them as we usually do. It is a chance to reverse the habits of a lifetime – who agrees?