Crafting Surveys with the Mobile User in Mind – Part 1

Many news outlets have picked up the results from a recent report from the Pew Internet & American Life Project on “Just-in-time Information through Mobile Connections.” Here’s the topline: “Some 70% of all cell phone owners and 86% of smartphone owners have used their phones in the previous 30 days to perform at least one of the following activities: Coordinate a meeting or get-together … Solve an unexpected problem that they or someone else had encountered … Decide whether to visit a business, such as a restaurant … Find information to help settle an argument they were having … Look up a score of a sporting event … Get up-to-the-minute traffic or public transit information to find the fastest way to get somewhere … Get help in an emergency situation.”

I’m thankful that I haven’t encountered all of these in the past month. But I’ve certainly encountered all of them at one point or another, and I’ll bet you have, too. They’re so common, so “normal,” that they almost go unnoticed. Together they represent one of the main categories of things that mobile phones, especially smartphones, are “for” today. (Though I’m kind of surprised that “Get help using, setting up, or fixing something” are missing from the list.)

Pew calls these occasions “just-in-time,” which seems right, but there’s another dimension involved: they take place while people are in the midst of doing something else. Around C+R we’ve been calling this kind of mobile usage “interstitial,” and it is a particularly relevant construct, as marketing research needs to rely more on reaching people via mobile devices. Our current methods are very much “extrastitial.”

Everyone in MR knows that online surveys are going to rely increasingly on respondents accessing them on mobile phones, principally smartphones. But that’s just not happening yet. At least here at C+R, smartphone-based attempts to access online surveys targeted to the general population (i.e., not specifically targeted to smartphone users) is generally in the single digits and generally flat over time. Why the disconnect between high and increasing mobile use and low and flat mobile access?

Online surveys (and most online qualitative methods) are designed to be viewed on a desktop/laptop computer, “leaning forward” to view a large screen, with input provided via a keyboard and mouse, and – most important – with the user fully focused on the task at hand, often for a significant period of time. Standard online surveys are designed to be “web first, mobile second,” while smartphone experiences (and the most successful apps) are “mobile first, web second” – “with you all the time and … used in moments of downtime.”

If this argument is right, it has huge implications for survey research.

The industry has never really tried to attract mobile users as a general-purpose sample source. Would the limited sample pool and dismal response rates being generated by the web first, mobile second approach be surpassed by a true mobile first, web second effort? Would a true mobile-oriented survey approach yield a much more representative, responsive sample?

Even more speculatively, as more and more online usage migrates to mobile devices, should we start thinking about mobile surveys as potentially becoming the eventual dominant form and traditional web surveys – conducted on “Immobile” or perhaps “Heritage” devices – the aging, declining technology? If so, then we’ll have to re-invent surveys to become interstitial. How will we do that and still be able to collect the extensive case data that some analyses require? (It feels like designing interstitial online qualitative experiences would be much more straightforward, but that’s a discussion for another day and post.)

Obviously we won’t be able to field 45-minute questionnaires anymore – a good thing from many angles – but how do we collect the information that we need for something like major category segmentations in a form that allows for interstitial usage?

Can we just encourage respondents to spread a long-form survey experience over several days? Something like 5 questions a day over the course of a week, giving the respondent the option at each juncture to continue now or resume tomorrow? This might look like the “remind me later” feature in Outlook: “What do you want to do now? (1) Answer 5 more questions, (2) Continue later” with “Continue later” bringing up a set list of delay periods – “remind me in XX hours” or something like that.

That’s logical, but it’s not appropriate for interstitial usage because it assumes scheduling and pre-planning. Interstitial usage “just pops up” between doing other things: talking to a friend, wondering about a sports team, figuring out where to go for lunch, waiting for a bus or train. (Go back and look at the Pew list of JIT occasions at the top of this post for confirmation on this.)

In our next post, we’ll explore Google Surveys and what they mean for MR business.