LinkedIn (LNKD.DL) is usually known as a social networking company for professionals, but it might also have a future in the Internet of Things (IoT), which connects everyday devices to each other. Kafka, its open-source platform, was initially created to pass messages across LinkedIn's sites and apps, but it could eventually facilitate communications between IoT objects as well.
LinkedIn recently revealed some staggering numbers regarding Kafka's growth. The platform now handles over a trillion messages per day, compared to a billion daily messages just four years ago. It currently processes 1.34 petabytes (1.34 million gigabytes) of data through its system every week. The platform works by receiving a message from one system then delivering it to other systems that require that data, usually in real time.
For example, if a co-worker updates his or her LinkedIn profile, the notification is delivered to the databases that power the website. The process sounds simple, but it can get complicated quickly when millions of users are all sending data simultaneously.
The market opportunity
Since Kafka excels at delivering real-time data between various services, it would be an ideal platform for delivering notifications to connected devices, like wearables, connected cars, or smart appliances. LinkedIn plans to improve Kafka's encryption and security features over the next two years, which could strengthen its own sites and potentially shield IoT devices from hackers.
The three LinkedIn engineers who originally developed Kafka have since left to establish Confluent, a start-up focused on commercializing Kafka as a paid data-management platform for large companies. Confluent enhances Kafka with advanced features and enterprise support that makes a company's data available as real-time streams. The system works via Kafka plugins, which lets businesses improve their data circulation without modifying their existing databases. The start-up has already raised over $30 million in funding (including an investment from LinkedIn) in less than a year.
Outside of businesses, real-time data streams can be used during live concerts or sporting events. Networking giant Cisco (NASDAQ: CSCO) started integrating real-time analytics functions into its networking hardware last December through its "Connected Analytics for the Internet of Everything" initiative. It showcased this technology with a "Connected Billboard" on Highway 101 in San Francisco, which displayed shorter or longer messages based on the driver's estimated speed and online traffic reports. Cisco estimates that the number of connected devices worldwide could double from 25 billion to 50 billion between 2015 and 2020.
A potential source of growth?
Since LinkedIn has already invested in Confluent, and the founders were its former engineers, it would be logical for the two companies to work together to expand the platform to the IoT market. If LinkedIn expands its version of Kafka and helps other businesses process and receive important data updates in real time, it might eventually become a new stream of revenue for the company.
We've already seen other social networks make similar moves in the big data and IoT markets. Twitter's (TWTR -7.40%) subscription-based data licensing business, which delivers a "firehose" of tweets to businesses, accounted for 10% of its top line last quarter. The company partnered with IBM (IBM 2.06%) last year to refine that data to help businesses predict upcoming trends. Twitter has also developed a data-processing platform called Storm, which processes real-time data for IoT devices, as well as a newer version called Heron.
Facebook (FB 2.13%) owns Parse, a mobile app development platform that provides apps with Facebook connections, user authentication, push notifications, and data analytics. By charging iOS and Android developers for subscriptions, Facebook can actually profit from app sales without running a mobile app store. Facebook recently expanded that system by letting Parse developers integrate their apps with IoT devices. This could eventually help Facebook gather data from smart home and wearable devices to gain even more complete profiles of its users.
The road ahead
Compared to Twitter and Facebook's ambitious big data and IoT efforts, LinkedIn seems to be falling behind. Twitter's firehose of tweets can be a goldmine of data for companies, while its notification system can be used for communications between smart devices. With Parse, Facebook can quietly mine user data from apps and IoT devices, which can improve its targeted ads.
For LinkedIn, its role in the IoT market is less clear. It houses the professional profiles of 380 million members, but there's no obvious way to utilize or monetize that data in the IoT market yet. Since Kafka is open source and already used by about 100 other companies, LinkedIn's version must also include standout features that aren't found elsewhere to be commercialized. But judging from LinkedIn's plans to upgrade Kafka with new features in the future, it's likely investors haven't heard the last of this open-source platform.