A couple of years ago I earned 15 minutes of fame by suggesting, at the MRS Annual Conference, that surveys would be dead in 20 years. Two years on I thought I ought to look at the issue again and see whether I now thought surveys had just 18 years to go.
The answer is yes and the reasons are pretty much the same as they were two years ago.
- Many research users are not showing any signs of improving their core surveys, i.e. long and boring. Indeed many brand trackers and customer satisfaction surveys seem to be getting worse - which is going to reduce co-operation rates still further.
- Traditional surveys provide data which is too slow, too expensive, and too shallow for many clients' needs.
- The focus is moving to discourse and 'real' conversations.
However, I do not believe that surveys will be replaced by social media listening/monitoring research. I doubt that passive listening will generate more than 10% of the research pie at any time in the next 50 years (contact me in 2062 to see if I am right).
So what do I think will replace the survey? I think that two elements will replace surveys
- Single questions
I see a future populated by Community Panels and enhanced Access Panels. A community panel is linked to single brand or topic whereas an access panel is available to anybody willing to pay the hire charge. In both cases the databases will increasingly store ever larger amounts of information, observational data, survey data, transactional data, and social media data. When we need to know something new we will typically just ask a single question, or a few questions, probably via a mobile device to build on what we already know about people. This is a very different prospect to being asked a complete a whole survey. I also see many of those single questions as being open-ended.
Discussions are a much more natural way to find out what people think, believe, know, and intend. I see bots being created to search social media and transactional databases looking for clues, asking questions and engaging in discourse.
Will there be any surveys in in 18 years? Yes, I think there will be some. Social research will sometimes need to reach people who are not in communities or access panels, so surveys will be needed. Some psychology will need surveys, and some choice based exercises will utilise experiences that could loosely be described as surveys. Mini-surveys, will also exist, five or six questions, capable of being answered on a mobile. But the standard survey of 2011, the 20, 30, or even 40 minutes of grids, check boxes, and piped lists is doomed!
In the short-term there are things we can do to lengthen the lifespan of surveys, namely make them more engaging, improve the targeting, and utilise stored data to reduce questions.