A Data Science Central Community
With two new members joining per second, LinkedIn continues to earn its title as “the world’s largest professional network on the Internet.” The Mountain View, Calif.-based company was formed in 2003 with the mission to “connect the world’s professionals, to make them more productive and successful.” True to its mission, the company now has more than 187 million members, including executives from all Fortune 500 companies.
The company employs almost 3,200 people worldwide, and earns $252 million in revenue from multiple channels including premium subscriptions, self-serve ads, hiring solutions and marketing solutions. Of the total, 64 percent of LinkedIn’s membership is from outside the United States. LinkedIn has more than 50 million members in Europe the Middle East and Africa, and about 34 million members in the Asia Pacific region. In all, LinkedIn members are from 200 countries worldwide.
LinkedIn continues to earn the respect of the business community, and member profiles are now an indispensable human resource tool for identifying and recruiting viable candidates for entry-level through senior executive positions. Member profiles contain information on previous employment, education history, and the member’s areas of expertise. The profile information is collected, rapidly analyzed, sorted, and made available for recruiters from global companies looking for skills to match their specific needs. More than 5.3 billion professionally-oriented recruitment searches of member profiles took place in 2012.
The voluminous data created by LinkedIn’s growing membership enables the company to customize products and services at the global level, as in the search for the most commonly used words by country. At the individual level, relevant searches include recommendations such as the “jobs you may be interested in”, talent matches, Pandora searches for people, profile browse maps, similar jobs, the referral engine, and “events you may be interested in.”
Analyzing large volumes of membership data to provide useful information to its members is critical to LinkedIn’s success. Data scientists, with backgrounds in computer science, mathematics, data analysis and business management, collect raw data and then explore, analyze, and translate the data into actionable information. The end result is a reduction in the overall time between the discovery of relevant facts, the characterization of the business opportunity, and the steps to act on opportunities like recruiting target candidates or identifying viable job opportunities.
Businesses benefit from LinkedIn products include searching for people of influence and social trends, testing new products and services, providing detailed reports of analysis of advertising revenue, assessing the impact of viral marketing, optimizing the recommendation engine, and creating specialized marketing services to business.
To obtain interesting results from its data, LinkedIn developed its own management application’s data flow, storage, research, network analysis, etc., and dashboards. For this, the company went used market solutions and built its own combination of technologies to meet its specific requirements. Its online data store is based on Oracle and Expresso. Its near line data store is based on Voldemort, Zoie, Bobo, Sensei and D-Graph. Pipeline services are based on Kafka and Databus. The off-line data store is based on Hadoop for machine learning, ranking & relevance, and Teradata for analytics & warehouse. LinkedIn uses MapReduce Analytics for pattern path analysis, clickstream, A/B site testing and data science discovery.
What’s next for LinkedIn? Future plans include developing new products & services like self-service analytics, metadata frameworks, integrating reporting solutions, and limitless mobility possibilities. In order to do this, the company will have to address challenges such as rapidly increasing data volumes, and continually improving data quality, standardization, and integration.
LinkedIn has proved that making data accessible to key stakeholders in a timely manner creates tremendous value. By unlocking its data, the products and services that can be created are countless. All it takes is imagination…and of course, the ability to analyze big data.