Foley on Microsoft

Microsoft and LinkedIn: What’s Next?

• By Mary Jo Foley
• 03/09/2017

Since Microsoft's deal for LinkedIn closed in early December, Microsoft CEO Satya Nadella has provided a short list of some integrations the company plans to execute first. Tying the LinkedIn platform with Outlook, Office/Office 365, Windows, Active Directory and Dynamics 365 figures prominently. Based on Microsoft’s recent job postings, it appears the pair also are working on a suite of human-capital management (HCM) software.

Nadella gave a nod to LinkedIn during Microsoft’s Q2 FY17 earnings call in late January, noting that Microsoft was not just cloud-first/mobile-first, but also "member first." During that call, Chief Financial Officer Amy Hood said Microsoft had no immediate plans to try to move LinkedIn to Microsoft’s Azure cloud. Prior to the Microsoft acquisition, LinkedIn was building out its own datacenter.

Instead of trying to corral LinkedIn into the Microsoft Productivity and Business Processes pen, Microsoft’s current management seems more interested in figuring out how Redmond might integrate some of LinkedIn’s ways into its other products and services.

The decision to elevate Kevin Scott, LinkedIn’s head of engineering, Microsoft’s own chief technology officer might indicate that Microsoft may be getting even more serious than it already has become about continuous development. Given his charter at LinkedIn, it appears Scott also will provide some LinkedIn influence regarding the way Microsoft bakes machine learning smarts into its software and services.

The reason Microsoft was willing to spend so much to acquire LinkedIn was for its data and algorithms. By combining Microsoft’s evolving knowledge graph with LinkedIn’s professional graph, Microsoft would be able to "make professionals more productive." Microsoft’s graph is a collection of information pertaining to entities like contacts, messages, calendar entries and documents. LinkedIn’s graph centers around entity information regarding jobs, co-workers, learning, prospects and recruiting/hiring.

LinkedIn has been building its knowledge base by pulling user-generated content from members, recruiters, advertisers and company administrators, and combining that with data extracted from the Internet. Machine learning techniques are key to how LinkedIn has built and refined its graph, according to an engineering blog about the LinkedIn graph from the company.

LinkedIn’s graph is a sizeable one. As of October, it held data on 450 million members, 190 million historical job listings, 9 million companies, 200-plus countries, 35,000 skills in 19 languages, 28,000 schools, 1,500 fields of study, 600-plus degrees, 24,000 titles in 19 languages, and 500-plus certificates, among other entities, LinkedIn officials said. Inferring relationships among these entities is the core to this graph, and LinkedIn has been training binary classifiers for each kind of relationship and collecting the training data.

While Microsoft and LinkedIn brass have been trumpeting the culture synergies between the two companies, the graph and machine learning integrations between the two are where things will become quite interesting.

Sure, in the near term Microsoft will instill its software and services with more people-focused smarts thanks to LinkedIn. But Microsoft’s longer-term future is dependent on cloud services, AI and machine learning, and cross-platform productivity software that’s updated in near-real time. And LinkedIn has amassed some valuable assets in these areas. Recruiting bots, Cortana-enabled scheduling that’s enhanced with LinkedIn profile information, and better "professional" search results in Bing are all real possibilities as the two companies begin integrating their respective graphs and AI/machine learning capabilities.

As other company watchers have noted, Microsoft doesn’t have a stellar track record with many of its larger acquisitions. But if the ’Softies can manage to walk the fine line between giving LinkedIn some independence while still benefitting from its engineering advances, the deal may be a big win for the company. And professionals might actually become more productive as a result.