.excerpt-thumb {display: block !important;}

5 Data Management Lessons from Microsoft's LinkedIn Acquisition

Originally published at Data Informed

There has been no shortage of articles and opinions, both positive and negative, since Microsoft announced a $26.2 billion acquisition of LinkedIn. Headlines and hypotheses range from improving CRM, data monetization, and advertising revenue to more personalized insights that enable Cortana to generate recommended actions. So who’s right? With an acquisition of this size, it’s no surprise that there are many ways in which the combined entity can deliver value.

Being familiar with the technology platforms and offerings of both Microsoft and LinkedIn, we have compiled five important data management lessons from which we believe every enterprise can benefit.

Data is the Lifeblood. Data as a Service is the Transfusion.

Let’s start by stating the obvious: With 433 million members, LinkedIn provides one of the best – if not the best – sources of professional profile data out there, coveted by every marketing and sales person and, of course, recruiters. That value was made evident in May 2015, when LinkedIn started limiting access to their APIs through their partner programs.

For enterprises, having seamless, real-time access to any third-party licensed or free public data for data quality and augmentation of customer profiles is essential. Today, receiving and loading data from third-party providers is still an IT laborious ETL process, with business users unhappy with the speed and accuracy of updates. Enterprises should view Microsoft’s goal to make LinkedIn profile data available on demand to their customers in Office 365, Dynamics CRM, and other applications as a table-pounding case for real-time Data as a Service (DaaS) access from all their third-party data vendors.

Data Quality and Accuracy Will Be the Priority

Microsoft is acquiring LinkedIn even though approximately 100 million of the 400 million (as detailed in LinkedIn’s Q3 2015 results) visit LinkedIn on a monthly basis, and likely not all of them are there to update their information. That doesn’t bode well for the quality and accuracy of the data. LinkedIn relies on the power of self-managed profiles. So information related to each user might only be updated if the user has gotten a new job or is beefing the profile in the hopes of finding a new job.

An approximate 25 percent refresh rate isn’t acceptable for B2B enterprises where corporate customer data is continuously renovated and maintained through a steady stream of internal and third-party sources. Just as Microsoft’s data monetization hopes lie with maintaining and improving data quality, enterprises that have not yet put reliable data measures in place can’t hope to succeed. Fortunately, new cloud-based master data management (MDM) offerings have made data quality, and a unified view of customers and products, affordable for companies of all sizes.

Profiles are Interesting, but the Graph of Relationships Matter

We are in the age of customer engagement, backed by hyper-personalization. While name, contact information, company, and job title are fundamental data constructs of CRM and marketing automation, affiliation, affinity, and peer relationships matter even more.

To achieve this level of personalized, relevant insight, both Microsoft and LinkedIn have to continue to leverage the power and technology of graph to capture, manage, and maintain data across relationships of entities at limitless scale. The power of graph extends beyond people-to-people connections, across all entities (Figure 1). This is often described as a 360-degree view.

Figure 1. Deep connections revealed by graph database technology.

Enterprises that desire similar 360-degree views of their customers, accounts, suppliers, products, and more must look beyond their traditional relational databases by investing in technologies that leverage graph. Graphs and other NoSQL databases, when paired with data quality discipline, form a powerful combination.

Machine Learning and Analytics put Relevant Insight in Context

In almost foreshadowing of the acquisition, LinkedIn and Microsoft announced last October integration from Cortana (the Windows 10 digital personal assistant) to LinkedIn profile data (Figure 2).

But this capability was just a teaser and a precursor to the possibility of using machine learning and analytics to tap into the combined Microsoft Office and LinkedIn graphs.

Once again, the only way this becomes a reality is if machine learning and predictive analytics tools have full access to a comprehensive set of relationships through the combined graphs, backed by a foundation of reliable data. Further, ongoing improvement in recommended actions generated by Cortana can only take place if outcomes from such actions are correlated back to suggestions offered in a closed loop.

Data-driven Applications are the New B2B Gold Standard

Like other social networking and self-curating, content-generating properties such as Facebook and Google – LinkedIn helped blaze the trail in data management, spawning Apache projects such as Hadoop, Voldemort, and more. With the power of the cloud, they have managed to scale to limitless data volumes within a single pool. Eschewing the traditional separation of analytical and operation applications, they instead offer personalized, contextual, role- and goal-focused data-driven applications or services on top of that single pool of data. They also have managed to tap into the power of social collaboration with self-curated updates and endorsements. For analytics and segmentation, they have provided an easy, text-based and filter facet search to allow anyone to narrow in on the combination of demographic and firmographic profile information. Recommended actions come in the form of whom you should consider connecting to, what jobs you may be interested in and, more recently, the addition of new company insights, which tap into the job change dates of LinkedIn users, to form the hiring trends across key departments within an organization. Microsoft will certainly be taking advantage all of these capabilities, as well as learning how to improve their own data-driven apps.

B2B companies need to follow suit. The future of the next generation of enterprise data-driven applications lies in blending together capabilities and technologies such as cloud, DaaSMDMgraphmachine learning, and analytics. IT teams then need to ensure that they can deliver data-driven applications similar to LinkedIn to their demanding business users. The good news is that modern data management PaaS platforms are available today, and they cost a lot less than $26.2 billion.