Big Insights on Human Behavior

Do you like the taste of beer? Do you prefer the people in your life to be simple or complex? If you answered “yes” and “simple”, you’re most likely a conservative voter who is not opposed to having sex on your first date, and you probably think gay marriage should be illegal. These are not just some funny fact that someone made up in the social media echo-sphere, it’s statistically relevant, according to OKCupid. OKCupid is an online dating site with a very unique approach. Instead of relying on simple profiles with a list of predetermined questions to match people, it allows for users to create any type of quiz or question to ask potential partners, and these questions then become part of the global set of matching criteria: it uses the answers to find a partner whose answers match up with yours, on the assumption that not opposites but similarities attract. On a blog associated with the site, OKCupid shares insights gleamed from their data. Their blog provides some thought-provocative connections between seemingly unrelated questions and facts. Did you know that iPhone users have more sex?

We know more about you than you yourself
In everyday life, we may not bother too much with examining who we are and what certain behavior says about our inner motivations. We just watch a movie because we think it is fun, we play a game because we’re bored and we share a picture on Facebook just because we like it. But what if someone, or something, is watching these seemingly trivial decisions and drawing far reaching conclusions about what we feel about politics, society, relationships, religion etc.?

Some time ago, a retail company was looking at shopping patterns, and found that certain purchases (statistically) indicated very early on if a woman was likely to be pregnant, sometimes even before the woman herself would be aware of this. Based on this insight, the retailer then targeted specific, pregnancy related ads to the woman in question. This freaked people out: how could they know this? This was getting too close! The solution to this? To mix in some random ads, for example for cars or gardening, so that the women would assume these ads were NOT targeted specifically to them and that the pregnancy ads were mere coincidence. In effect: the retailer was hiding the fact that they knew a lot about you.

Cross-linking information
These examples are each ‘just’ from a single company, and they do not even tap into your online shopping behavior, your credit-card receipts, your social media posts, your location check-ins and your network of friends and colleagues. In fact, these examples are not even truly big data: it is traditional BI, based on data cubes, good questions and good statistical analyses by people with a hunch. It is not real-time and it’s not in huge unstructured datasets from a variety of sources. Part of the intrigue of Big Data is to imagine what will happen if a machine has access to all these data streams and would help us to come up with the connections, the insights and perhaps even starts verifying or acting on them accordingly. Imagine the deep personal insights that Google, Apple, Microsoft, Facebook and Amazon and any other ‘portal’ company could be gathering about you.

Real-time society improves accuracy
Whenever there is a feedback loop, big data will become much more powerful, since the insights and assumptions can be verified and tuned: you can present people with ads and measure the click-through, offer people a coupon and see how many people use it, give people a music recommendation and see if they truly like it, set up a date between two people and see if they actually have a click. The quicker the feedback loop, the better you can tune your interpretations. This is what has been happening in online marketing for years: vary the location, size and content of ads and use real-time response statistics to improve the revenue. And we’re living in a world where everything is becoming real-time…

Ethics beyond privacy
It is already hard to create awareness around ‘simple’ privacy: “do not share your name, date of birth or family composition”. People still easily trade personal information away (forever), for simple games or access to fairly trivial information, and people are still completely oblivious about this next level of privacy: the seemingly irrelevant or funny interactions online that may reveal much more about us than we think. Is this just ‘the end of all privacy’ or the beginning of an era where we finally understand human motivations and behavior, so that we can use it for common good? What is the role of large corporations? Can we expect companies to behave ethically? Or is strong international regulation required? Will the data just be insight, or the basis of literally irresistible offers from all sides? If we combine these insights with advances in neurology and psychology, can we create the ultimate ad that triggers us to buy something even though we may not have a real need for it? Perhaps Big Data can deliver the triggers and insights that Neuromarketing is waiting for to break into mainstream marketing…

About Erik van Ommeren

Director of Innovation at Sogeti USA LLC/VINT Erik van Ommeren is responsible for VINT – the International Research Institute of Sogeti in the USA. He is an analyst, public speaker and co-author of the books: “SOA for profit”, “Collaboration in the Cloud” and “Seize the cloud”.

Speak Your Mind