Foreseeing events before they happen. This power, largely associated with fortune tellers and Precogs in the movie Minority Report (predict crimes before they happen), is a reality. Companies are now actively collaborating with social media platforms (~Twitter) leveraging user data to predict future purchasing outcomes (predictive analytics) as a result of advances in Big Data/Crowdfunding.

In the National Bestseller “Big Data,” authors Vikto Mayer-Schonberger and Kenneth Cukier show how the retail chain Target “relies on predictions based on big-data correlations.” For example, knowing if customers are pregnant is important to retailers given it shapes shopping behaviors including going into new stores and developing brand loyalties. Target, using purchasing behavior data, determined which customers were most likely pregnant based on prediction scores (derived from a dozen products acting as proxies). These scores allowed Target to “estimate due dates within a narrow range so it could send relevant coupons for each stage of pregnancy.” Though Target’s practice ultimately was deemed controversial and a potential violation of privacy (teenager was solicited though she was pregnant), it illustrates the power of Big Data for predicating future purchasing outcomes.

Big-Data-bookAnother example mentioned in “Big Data” involves “Likes.” Facebook utilizes “Likes” (members approving of user companies, news releases, comments, activities) to empower predictive analysis by associating member activities with future outcomes and customizes user news feeds matched to advertising. And Facebook has a lot of data to analyze given members willingly share information online; members click a “Like” button/leave a comment three billion times per day. Facebook tracks users’ “status updates” and “Likes” and determines the most suitable ads to display on its website to earn revenue.

So what is the intelligence behind Big Data/Crowdfunding enabling companies like Facebook and Target to predict future outcomes? That would be predictive analytic engines including Machine Intelligence, or applying math to large quantities of data in order to infer probabilities, and Data Mining (use of XML to tag words). Predictive analytics has risen in tandem with the social media revolution. When Amazon recommends a book you would like, Google predicts that you should leave now to get to your meeting on time, or Pandora magically creates your ideal playlist, these are examples of machine learning over a Big Data stream. (Source: “Machine Learning and Big Data Analytics: The Perfect Marriage,” Willem Waegeman)

Data is to the information society what fuel was to the industrial economy; the critical resource powering the innovations that people rely on. Let’s take a closer look at three of Silicon Valley’s predictive analytic crown jewels: Geo-Location, Social Graph, and Sentiment Analysis.


A major technology embedded in high-flying tech/app companies like Uber and Wave, geo-location enables users to locate/schedule transportation and detect traffic jams (i.e., assessing the speed of phones traveling on highways). Geo-Location is inherently predictive allowing companies to serve ads based on user locations.

User locations in time and space leads to a wide range of apps delivering personalized content (contextual computing). “As the technologies and data underpinning contextification grow, there will be an increasing ability to actually predict user future context creating market pull rather than market push. In the future, being responsive to consumers will be a ticket to market failure. Rather, being predictive of consumer’s wants and needs will be expected.” (Source: Data Crush, Christopher Surdak)

“A company might know who my friends are in the city I am traveling to, what their availability is to meet with me while I’m in town, and the name of their favorite local restaurant. This information enables a marketing person to create a hyperlinked message.”

Case Study: Smart Glasses/Geo-Location

  • The Jamba Juice app downloaded to iGlass knows you are currently out shopping with your friend John (using facial recognition software on a video you posted on Facebook)
  • John loves Very Berry smoothies from Jamba Juice (known from John’s prior purchasing behavior)
  • You are both near your local Jamba Juice shop (using location data from iGlasses)
  • Jamba Juice app knows to text an offer to the two of you for a two-for-one discount at that Jamba Juice if you both stop by in the next 15 minutes (Source: Data Crush, Christopher Surdak)

Social Graph

In 2013, Facebook had over one billion users interconnected via 100 billion friendships representing 10% of the world’s population. Social Graphs (global mapping of everybody and how they’re related) enable Facebook to assess user preferences when correlated with purchasing behavior including driving the company’s preeminent advertising platform. The more time users spend on Facebook, the more advertisers learn about users and the more valuable eyeballs become to them (average user spends over 400 minutes per month logged into the site).

Sentiment Analysis

Though limited to 140 characters, Tweets are rich in metadata including geo-location, user’s language, and #/names of those they follow. Hedge funds (Derwent Capital, MarketPsych) analyze sentiments in tweets to “signal investment” in the stock market.

Read more, or download the entire article, here.