The first week in June brought together a diverse group of people from Facebook, Instagram, Twitter, Slack, Reddit, Coca-Cola, StockTwits, Morgan Stanley, the UN, and the CIA to discuss the role of social data in industry and society at Big Boulder. The conference covered the many applications of social media data – the world’s largest set of anthropological data ­– across marketing, finance, and government.

Several themes emerged from Big Boulder as to how to extract value from social data and why it’s an important signal for any organization that values insights from people’s behavior.

1. Early Detection

Social media is often the first place people turn when reacting to an event happening in the world. Those reactions are captured in the data which, with the right algorithms, manifest as signals that can alert an organization to a trend before anything else.

By detecting complaints about food prices, Daniel Pedraza explained how UN Global Pulse can predict problems with food security. In a similar vein, the CIA uses social data to forecast the propagation of violence up to 10 days in advance and develop an action plan ahead of time.

For marketers, early detection sheds light into how consumers think and feel about a particular product. Joel Lunenfeld discussed how conversations around the red carpet during awards shows led to Dove’s Real Beauty campaign. Brad Ruffkess reiterated the point when describing how Coca-Cola developed its #ThatsGold campaign for the 2016 Olympics 2 years in advance utilizing trends in social data.

2. Networked Audiences

Following the Fyre Festival, social influencers are getting a bad rap. But, when people lend their name in support of a product they neglect to understand, they should be held accountable. It is also forcing organizations to reconsider how to think about influencers.

Social media is revealing powerful insights into the structure of social groups and the nodes that have the power to influence others. One of the problems with social influencers is organizations simply chase a person that has the most followers rather than an audience or node that has the power to move other people.

Deb Roy of MIT Media Lab mined Twitter data to understand the tribes making up the American political landscape during the US election. Comparing how the social conversation diverged from media coverage, MIT Media Lab identified the topics people cared about in relation to what was being reported on, such as campaign finance and the vice presidential debate.

Focusing on identifying influential audience segments delivers results for brands. Devon Wijesinghe of InsightPool recounted when Advance Auto Parts discovered a female audience interested in Formula One in lieu of the expected audience of men who like NASCAR. Advance Auto Parts ran a campaign targeting the new audience for the first Formula One driver, which resulted in a 10x higher engagement rate.

3. Integrating Multiple Signals

The increasing value of data was a theme throughout several of the sessions. Marketers Gloria DeCoste from Nestle, Brad Ruffkess from Coca-Cola, and Beverly Jackson from MGM International Resorts described data as a capital asset that should be owned by organizations. And, more important than any single data set, is having multiple differentiated data sets that provide new information.

For these marketers, social data delivers understanding of what consumers care about and how they’re experiencing a product in market enabling brands to create more personalized consumer experiences. Insights gained from social are also helping optimize content, such as images that drive response, across other channels.

Organizations increasingly want to connect social data to other data sets. Brands expressed interest in leveraging first and second party data more as well as connect data to offline transactions to understand attribution.

4. Unfiltered Interactions

One of the key advantages of social media, which is more important today than ever, is the ability to communicate directly on a person-to-person basis. These raw interactions are what make social data incredibly valuable.

Reddit discussed its ability for people to express themselves and find communities. In doing so, Reddit’s AMA has become a necessary stop on the campaign trail for politicians to connect with their constituents. Similarly, social media has become a powerful force, being adopted by candidates so they can deliver their messages unfiltered. They also enable anybody in the world to become famous, like #NuggsForCarter. This feature, makes the medium more powerful than any other when it comes to understanding people.

5. Visual Insights

Content that was mostly text-based when social platforms were initially introduced has exploded into visual content. New features and increased bandwidth is enabling images and video in feeds as the camera replaces the keyboard. Vishal Shah from Instagram explained how social content is moving from text to images and video because of the latter’s storytelling power.

With text analysis approaching maturity and visual content dominating feeds, organizations are focusing their attention on extracting insights from images and video. Understanding the context of images will help organizations understand the people behind them. For marketers this means identifying the when and where a brand shows up in an image or video, or being able to identify dog owners and create a target audience for a campaign.

While this technology is emerging in the social space, it’s mature in other spaces such as TV. With our Teletrax technology, we’ve been helping brands and broadcasters understand the context of their TV content for more than a decade.

Big Boulder highlighted the versatility and transformative capability of social data for organizations when it’s processed by sound data science. And, the view from the conference wasn’t bad either.