But Wait, There’s More!
Most of us know people who are very similar to us. We also know people who can be very different from us. And these people are one and the same.
On one hand, big data would show you and your circle to be much the same – similar demographics, incomes, voting habits, occupations, media consumption, social values. Yet on the other hand, we all know that when it really counts, we’re all quite different. The differences within families are a classic example of this – how we can be so similar yet so different.
This is where the age of ‘big data’ both worries and excites market researchers. The worry comes from the fear of ‘traditional’ market research (the stuff that actually involves talking with real people) being replaced by solely data-based research – because it’s a false economy once the loss of insight is factored in. The excitement comes from what can be gained through combining the old with the new.
Think: The interpretation of research results has usually involved the ad hoc addition of insights and information from other sources, be they formal (e.g. comparing sales, TARPs or government data with survey data) or informal (using personal knowledge to help shed light). The latter approach is more subjective and thus prone to error but can also offer great insight; the former is something that should be chased as it can do so much to inform the latter. Got that? Combine data from survey and non-survey data to foster the best insights and greatest value.
These two approaches for adding value to survey data succeed because they combine impartial observations about what people actually do with those about what they profess to think.
This really isn’t new. Jonathan called for a greater integration of survey data with behavioural data in a conference paper back in 2005 and frankly it wasn’t new even then. What is new is the convergence of three factors:
1. The growth of trackable social media;
2. The digitisation of our lives;
3. The ability to process, interpret and present data faster and better.
The first of these factors is getting the most headlines, but it’s also where the most caution is required. Despite the growth of social media, the vast majority of New Zealanders are not regularly tweeting, the majority of Facebook cannot be tracked, and the TradeMe community boards remain off-limits to tracking programmes. So while the apparent potential of social media analysis appears strong, the fact remains that the profile of the people and content available will inevitably be a mismatch to your actual needs.
Secondly: the digitisation of almost every facet of our lives. This is the gold, and it doesn’t stop at overlaying data such as your customers’ digital interactions with your brand onto your survey data and hoping for correlations. Design your research carefully and you can integrate survey data directly into your CRM databases and their ilk. The hurdles (both ethical and technical), are not insurmountable, and should be investigated further by anybody who desires more research value.
Thirdly, improvements in processing, interpreting and presenting data have helped research to actually become trendy – those pesky infographics are everywhere, and the good ones do more than any PowerPoint charts ever could. What makes a good infographic? The story-telling within it. Great researchers are great story-tellers - but they need the content for their stories, and that comes as much from the data you hold within your own business as it does from within the statistics of their surveys. By Jonathan Dodd, Research Director, Ipsos jonathan.dodd@ipsos.com and Ricky Hodge, Custom Analytics Specialist, Ipsos ricky.hodge@ipsos.com