What The White House Learned From LinkedIn And The Use Of Big Data

By on October 11, 2012 at 9:00 AM

One of the best presentations at the recent Big Data Innovation Summit in Boston was by LinkedIn Senior Data Scientist Yael Garten. Garten, who leads LinkedIn’s Mobile Data Analytics team, in a presentation entitled Data Infused Product Design & Insights at LinkedIn provided a glimpse of how big data is used by LinkedIn to explore usage patterns, on mobile devices, for instance.

This is a challenge facing the US Government in the new Digital Government Strategy: namely delivering existing web sites and database information — and eventually the types of big data results that the intelligence and scientific communities have — so that mobile devices can access that information from supercomputers.

For those who haven’t kept up with LinkedIn’s progress of late: It’s mission is to “connect the world’s professionals to make them more productive and successful.” Here are a few facts at glance, from Garten’s presentation:
  • LinkedIn was founded in 2003
  • More than 175 million members
  • Registers two new signups per second
  • Executives from all Fortune 500 companies
  • 80% are “decision makers”
  • Average Household income in US: $86,000
  • Generates about 4 billion people searches annually
  • Reaches more than 200 countries and territories
  • Is available in 17 different languages
LinkedIn can take the data behind each of its members’ professional profiles, and slice and dice by seniority, by job function, etc. so they can ask many interesting questions. The data science team at LinkedIn does the following with that data set:
  • Produce product and business insights
  • Build data products
  • Extract insights that can be shared externally
An amazing example of the latter was done for the White House Council on Economic Advisors.

The team put together one of a variety of data stories on “Growing and Shrinking Industries” and “Where did all the people go from the collapsed financial institutions in 2008?”


The work for the White House is summarized in the LinkedIn blog and the graphic above.

Crunching through LinkedIn’s vast collection of data, between 2007 and 2011, the company’s analysts observed:

  • The fastest-growing industries included renewables (+49.2%), internet (+24.6%), online publishing (+24.3%), and e-learning (+15.9%).
  • The fastest-shrinking industries were newspapers (-28.4%), retail (-15.5%), building materials (-14.2%), and automotive (-12.8%).
  • The volume of job gain / loss by industry (as indicated by the largest bubbles in the figure above) show that even through the recession, the industries with the largest volume of employment growth were internet, hospitals and healthcare, health, wellness and fitness, oil and energy, IT and renewables.
  • Retail, construction, telecommunications, banking, and automotive had the largest volume of job losses between 2007 and 2011.
The answer to the question: Where did all the people go from the collapsed financial institutions in 2008? It turned out, it was to other financial institutions that survived and thrived!

Linkedin CEO, Jeff Weiner, wrote last autumn on how to solve the “other” employment crisis: the ability to connect talent with opportunity is more important now than ever before.

“Our vision is a world where every work opportunity is digitally searchable and linked to the appropriate company and the skills required for that opportunity,” he says. “These industry trends uncovered by our data science team are steps along the way to making LinkedIn’s call for a economic graphs a reality.”


This is also an example of our government using both open (big) data, as well as conventional statistical data, that I wrote about recently. In the words of Margo Anderson, Census historian: “It is definitely different from the past and is both brash and exciting”.

Shares0