2013-09-15

Terry Young: Using Big Data to Spot Big Trends


source: Big Think 2013-06-25
All across the Internet, we see little memes popping up, and then petering out. A savvy Internet user might be able to use her intuition to spot the memes that have staying power, but identifying these memes in isolation, while perhaps interesting, is not very useful.

"What makes it useful and what makes it actionable," says Terry Young, founder and CEO of sparks & honey, a data-driven advertising newsroom, "is when you cluster multiple things that look similar together and you begin to analyze the patterns and you begin to quantify it."

In other words, when you are trying to spot a future trend you need to find an organizing principle that surounds it. Macrotrends, as Young explains in the video below, are like living, dynamic organisms. At sparks & honey, Young's team uses data analysis to track the subtle movements of 60 macrotrends. "We use those clusters in order to build content and build relevance for a brand," he says.

Transcript -- When you're trying to predict a future trend you need an organizing principle. And what we have done is created 60 macrotrends. We believe that these are living organisms and that they shift and they shape and they're dynamic in nature, almost like a neuro network exists under each macrotrend. So if I threw out the macrotrend like superhuman or robo apocalypse or living matter, we track on a daily basis all of the little subtle movements that happen, the shifts in consumer conversations, the shifts in influencers, the shifts in new scientific studies that support that macrotrend. And we score each of those components.

The importance of that is that it allows us to understand the dynamic nature and take -- use big data, which feeds in, use real time examples -- we connect it into our network of 60 macrotrends and then we use those clusters in order to build content and build relevance for a brand.

For a macrotrend to be born we have to see X number of manifestations in the marketplace. And it varies by category, but when we have enough activity clustered together we create a new macrotrend. Sometimes these are born from an existing macrotrend, but the things that we watch -- there's a couple of pieces. One is we watch very closely the patterns that we see. The second, we watch for the things that are accelerating the movement.

For every trend there's an accelerator and there's a balancer. And the accelerators -- if you imagine sensor network, which is the idea of taking small little sensors, embedding them into dumb products and making them smart products, one of the things that is accelerating that is everything that's happening around robotics, artificial intelligence, everything that's happening around super human, bionics, singularity movement, so forth and so on. On the flip side, though, we have content networks like digital detox, where people are running away from digital so that they can remove digital from their lives, or incognito, where people are trying to look for ways of masking themselves so that they can have a greater level of privacy, or robo apocalypse, which is the fear that robots are going to take over our world, and you see it manifesting in entertainment, popular culture and books.

Without organizing principle, basically what happens is this: brands see something that burst in culture and they're like, "That's really interesting." And they do nothing with it. Why? Because anything that bursts in isolation is not useful. It's just interesting. What makes it useful and what makes it acitionable is when you cluster multiple things that look similar together and you begin to analyze the patterns and you begin to quantify it.

Directed / Produced by Jonathan Fowler & Elizabeth Rodd