The survey is dead, long live the survey

2015 has been an interesting year for pollsters. Whilst the general election exit poll once again proved to be remarkably accurate, the polls in the run up to election day were lacking in predictive power (even if Ipsos MORI’s were the best of a bad bunch!). Ben Page discusses this in more detail on page 130. Whilst political polls are quite unlike the other 99% of our survey work, the poll post-mortem has raised an interesting question about the value of questioning. As we head into 2016, should the questionnaire still be the default research tool or should we ditch the survey in favour of passive techniques that allow us to monitor actual rather than claimed behaviour? Could we be entering an era of research without questions?

Observing people clearly isn’t new – we’ve been conducting ethnographic research for decades – but advances in technology mean that we can now do this relatively cost-effectively over sustained periods and at scale. We use a combination of bespoke research technology (such as online passive meters, audio tracking and social listening tools) and inexpensive consumer devices repurposed for research (such as smartphones, GoPros and pedometers). This allows us to effectively turn the research participant into the researcher by outsourcing the data collection to them, enabling us to conduct longitudinal and quantitative observation, such as

  • collecting the mobile behaviours via passive meters of c. 1,500 participants – the websites they visit, and the apps they use – to feed into the Institute of Practitioners in Advertising’s TouchPoints media planning tool;
  • using GPS meters to track thousands of people’s journeys every year to help our client, Route, measure the value of out-of-home advertising;
  • monitoring online conversations to evaluate new product launches and analyse, in real time, the response to the general election leaders’ debates on Twitter; and
  • using wearable and fixed cameras to reveal how people interact with products and uncover their habitual behaviours, from how they clean their house, to the unconscious coping mechanisms of those with chronic conditions.

The above are just a few examples, but they illustrate how much we can learn without asking a single question. What’s more, some of the insights we are now able to glean, we wouldn’t have been able to get as quickly, as accurately, or at all, via traditional means. The inputs – the photos, videos, online conversations, location maps etc. – are helping us create much more compelling and immersive outputs for our clients.

But can we truly address every issue without asking a single question? No. We can speculate and hypothesise but, if we want to understand what motivates someone, what is behind observed behaviour, we still need to ask. And ask at the right time, in the right place and using the right channel. We’re producing our most authentic, actionable and inspiring insights by augmenting survey research with other sources of data. So rather than research without questions, we’re focusing on asking fewer, more relevant, more intelligent and timely questions rather than no questions at all.