I’ve been known to ask a lot of questions. Just ask my team at work, my daughter or my wife, for that matter. (Turns out she doesn’t appreciate being subjected to the Socratic method…go figure.) But even I’ve learned that there are diminishing returns when it comes to asking questions—the 10th one just makes them mad.
And so it is with survey research. Dr. Reg Baker, AKA the “Survey Geek,” has posted on the topic no fewer than nine times in the last few years, quoting many definitive sources including Galesic and Bosnjak’s important study on the impact of survey length on data quality in web studies.
As if all the accumulated evidence is not enough, we now have a new reason to be concerned about survey length: the rise of mobile as a viable option to complete web surveys. More and more, respondents are choosing to do surveys on smartphones, whether we designed them with mobile in mind or not. In this example from a web tracker for a telecom client, mobile survey starts are up 50% in the past year. However, mobile survey completes lag behind. In fact, the average completion rate for mobile survey starters across all the studies we looked at was only about 75%. That’s because, on average, mobile responders experience 25-50% longer survey lengths and break off at nearly twice the rate of non-mobile responders. This may be partially related to issues of mobile usability (e.g., tiny radio buttons or 11-point scale grid questions), but the evidence also implicates survey length and, in particular, the extra length that mobile responders experience when completing a web survey over a slow cellular network.
I’d also argue that mobile responders come to the table with a different set of expectations. This is particularly true when the responder is not an experienced survey taker (i.e., not a member of a panel). For most of us, mobile internet is about convenience and completing tasks quickly and efficiently. This is why the best eCommerce or news sites dramatically simplify their mobile-optimized web offerings. They know their audience wants to get in and out quickly.
In an increasingly mobile world, where mobile is just another option for responding to survey requests—an option, by the way, that is chosen disproportionately by folks who are often challenging to collect data from—we now have another compelling reason to shorten and simplify our surveys. Realistically, we may have no choice.
I invite you to watch my 11-minute presentation, “Getting Friendly with Unintentional Mobile Respondents.” It summarizes our analysis of mobile responders on recent Market Strategies web studies and shares our thoughts on how data collection must evolve to harness the power and the inevitability of mobile.