The rapid adoption of mobile technology has proven both an opportunity and a threat to market research. It has enabled new approaches and allowed access to the burgeoning world of social media, but also caused compatibility migraines. Painful participant experiences have led to unusable data and contributed to response rate problems. Inparticular, online surveys designed with larger screens in mind have failed to work on smaller ones.
Technology has moved on so fast recently that developments aimed at improving the mobile compatibility of online surveys have often been rendered obsolete only a few months after launch. Thankfully, things now seem to be settling down – the mobile ‘smartphone’ is ubiquitous and its fundamental design and operation has altered little in the last ten years or so, and as far as online surveys are concerned the following differentiation is the only one that matters:
1. Devices with larger screens.
Typically, this means personal computers and mid to large size tablets (including iPads). The tablets in this category are capable of displaying online questionnaires as clearly as they appear on modern desktop monitors, and as such they are not especially problematic.
2. Devices with smaller screens.
Typically, this means smartphones and smaller tablets, and whilst most of these can run very high screen resolutions, a physically small screen size makes content displayed at those resolutions virtually unreadable without fiddly zooming in and out. For that reason, even the latest smartphones will by default load web pages in a lower resolution “reader mode”. For example, see below for how the latest iPhone X, which is capable of a resolution of 1125 x 2436 pixels, will load pages at only a third of that (375 x 812 pixels). Older phones run at these resolutions too (e.g. the Samsung Galaxy S5 runs at 360 x 64 pixels).
Our own data show that, on average, 25% of healthcare professional (HCP) participants take online surveys using their smartphones, varying between about 10% for longer surveys (30+ minutes) and upwards of 50% for shorter surveys (5-10 minutes). Amongst the general public, these proportions are even higher. Given that most quantitative questionnaires contain large and/or complex questions that don’t lend themselves to presentation via smaller screens, designing truly mobile compatible surveys is a pressing challenge.
Our experience at First Line is that solutions to this problem are best home-grown. For instance, whilst the top-end survey authoring software that we use (Confirmit) in theory allows designers to simply toggle between a desktop and a mobile compatible version of a survey, in practice this merely deconstructs questions into their constituent parts and presents them separately, taking no account of impact on usability (e.g. a grid question is converted into a string of single questions). Not only that but much of the essential formatting is lost, so it looks messy too. The result is that the so-called mobile compatible version usually looks worse on the small screen than does the original “desktop” version. Developments are ongoing, but many of the most common types of quantitative questions are yet to be properly catered for.
Fortunately, there are some simple remedies. For example, we found that removing logos, using icons rather than text, and expressing survey progress as a number rather than a graphic increased the available screen space on smaller screens from 60% to 85%. However, many quantitative questions require us to invest in new approaches. Take the ubiquitous “grid” question – items down the side, and a scale (verbatim, numeric, or sliding) across the top. You might also have a “Don’t know” option. We know that such grids can often have 20 or more list items, each to be scored on a scale that may have 10 or more scale points. Such monsters can appear daunting on big screens never mind small ones, but they are a fact of life in quantitative market research.
At First Line we’ve developed the following approach to their presentation on smaller screens. In fact, we like the approach so much it has become our recommended execution for larger screens too.
List items are revealed one by one as the participant works through, with a summary of previous scores shown above. Screen space is maximised, completion is intuitive, and presentation is visually clear and attractive. All the components of the question are configurable, so it could just as easily be a numeric 1 to 7 scale, using different colours, and without the scale-end descriptors or “Don’t know” /“N/A” options. We have taken a similar “think again” approach with other common quantitative questions when it comes to their appearance on smaller screens (e.g. 3D-grids, numeric entry grids, and diary style questions).
All our online surveys auto-detect the type of device and screen size that the participant has chosen to use and, based on that, automatically revert to one of two presentation modes – larger screen or smaller screen – thus ensuring that questions are shown optimally. With response rates being what they are, we simply can’t force willing research participants to use larger screens to get our research done, and it is us that must adapt to the smaller screen world in which we all now live so much of the time.