Late last year, an American researcher used anonymised, aggregated data from multiple public sources to answer a question that had previously been only guessed at by traditional in-person or phone surveys: How many American men are gay?
For marketers what’s fascinating about this research is not that we finally know how many American men are gay (about 5%, it seems), but that the answer was made possible by marketing’s buzzingest new buzzword: Big Data.
Predictably, most of the media reaction to the research focused on the political implications of the findings (Here’s the headline: US states that are less tolerant of same-sex relationships don’t have fewer gay men, they just keep more of their gay men in the closet). However, the researcher Seth Stephens-Davidowitz revealed in a podcast interview with Dan Savage (warning NSFW) that he personally has little interest in the topic itself.
Multiple sources is the secret sauce.
In his role as a data scientist with Google and a NY Times columnist, Stephens-Davidowitz has been shining the ‘big data’ flashlight on some fairly provocative topics, but his interest is not in being provocative, per se. His work focuses on the potential of data analytics, fueled by the almost unlimited information generated by our use of the internet, to accurately answer questions we could previously only resolve with expensive surveys of limited scope and questionable accuracy.
Rather than relying on a single source of direct-survey information, Stephens-Davidowitz combined aggregate data from Facebook (relationship status), porn keyword searches (obvious enough), Craigslist (casual encounters listings), popular dating sites (profile and seeking data) and Google searches (including prevalence of the query “Is my husband gay?”). He then overlayed these sets with more traditional sources such as the US census and Gallup polls and found very strong correlations: a classic case study of combining multiple-source, multiple format data from unlikely vendors to learn something new and, perhaps, previously unknowable.
Let me tell you what I think you want to hear.
Survey data is notoriously unreliable, for a whole host of reasons. I’ve had B2B publishers confide that readers who claim they are responsible for major purchasing budgets, for example, rarely are. Junior staff talk up their roles to feel important, while more senior execs talk theirs down, knowing that publishers and advertisers are fishing for prospects and they can do without the spam. You can ask any advertising creative who has witnessed a ‘focus group’ in action to supply further examples of what people will say to earn a free sandwich and a turn on the microphone.
While prompted surveys are good for measuring intention, big data is increasingly being used to measure un-provoked behaviour. Which is otherwise known as ‘the truth’. As anyone who’s ever signed up for a gym membership can tell you, intention and behaviour are two entirely different animals.
9 out of 10 Marketers plan to use Big Data in their next campaign.*
It seems that while a lot of the marketing industry talk around “big data” is reveling in the technical wizardry of the tools and offering generic business double-speak about competitive advantages, the real story here is that big data, ultimately, knows the truth.
It will, however, require creative minds to find and combine previously un-related data sets that reveal the truth about our preferences, purchases and behaviours. And this truth, in turn, will set creative marketers free – free to develop counter-intuitive strategies, pursue previously-dismissed niche markets and deliver provocative messages that resonate profoundly with an audience who knows, deep inside, that there is a brand who truly understands them.
* This statistic** is entirely made up, but can you imagine what would prompt a marketer to publicly admit they have no intention of using the latest marketing technology?
** 37% of all statistics are made up.
About the Author: Barrie Seppings blogs about making things better – for clients, brands, agencies and humans. He is currently Regional Creative Director at Ogilvy Singapore and he likes boards surf, skate and snow. Follow him on the Twitter, connect on LinkedIn, or add him on Google+
About the images: all photographs used with the permission of Martin Ollman Photography. Contact Martin directly for rights and commissions.