LinkedIn and DataSift: Data-Driven Marketing and Human Data Intelligence Come Together

18th January 2017 1 Comment

As a partner to some of the world’s largest social networks, we’ve witnessed first hand the continued explosion in the volume, variety, and complexity of social data over the last few years.

However, as volumes grow, we’ve seen that the ability for us mere mortals to analyse, interpret, and identify signals from data is becoming increasingly hard. It’s time for machines to lend a hand. Today, our latest collaboration highlights this major shift towards what we at DataSift call active intelligence – businesses now want answers, not just access to vast pools of data, or beautified graphics showing vanity metrics.

So it’s with great excitement that I announce our strategic partnership with LinkedIn and the launch of PYLON for LinkedIn Engagement Insights, which gives marketers advanced insights into the activities, behaviour, and interests of more than 467 million professionals — all in a way that protects people’s privacy.

Bringing together data-driven marketing and active data intelligence, we can now provide marketers with valuable, real-time answers to questions around audience identification, content engagement and brand awareness/perceptions. Equipped with these insights, they can take action, drive decisions and make changes to their campaigns, content, and strategies.

How does LinkedIn Engagement Insights work?

The overwhelming data volumes involved means algorithms are key to the analysis, doing the heavy lifting required to guide people to insights and answers. Machine Learning, Natural Language Analysis (NLP) and Machine Intelligence have a huge role to play. Think about Amazon’s Echo, which uses an AI assistant application called Alexa to enable you to access music, news, weather, personal organization tools and other information instantly. You ask it a question, your command is sent to Amazon’s servers, processed, analysed and a response sent back within a second. The more questions you ask, the smarter the system becomes. With LinkedIn Engagement Insights, powered by DataSift’s PYLON technology, we’re applying compute-intensive, machine intelligence to answer marketers’ questions and drive business impact. Our goal is to surface answers, recommendations, and actions for marketers, not simply surface metrics for dashboards.

PYLON for LinkedIn Engagement Insights processes LinkedIn’s live newsfeed of published content and the engagement surrounding it (clicks, impressions, shares, likes, comments.) It then applies natural language processing and advanced analytics to understand the text within posts and articles to identify the things (like companies, products, and industry topics). With this multi-dimensional stream of real-time data, we can automatically surface insights into different audience segments, content, and engagement to surface answers and actionable intelligence to marketers.

This means marketers and agencies can:

  • Discover new audiences for brands and products – Research the behavior of high value audiences on LinkedIn and validate assumptions. Confirm audience segments by analyzing the job titles, locations, and skills of audiences that are engaging with relevant content on the network. New audience segments can also be identified based on their engagement with content that intersects with a brand’s existing audience
  • Learn what content works best with audiences – Gain a greater understanding of what topics, types of content and creative resonates most with key audiences. Use content insights to better shape and inform content marketing strategies, create more compelling content for target audiences and more effectively measure success
  • Benchmark brand against the competition  – Analyze brand awareness and perception. Understand how audiences engage with their brand, products or services on LinkedIn relative to industry competitors and peers.

In practice, with LinkedIn Engagement Insights media planners can now identify audience segments engaging with content related to topics such as “Big Data” and surface demographics such as job titles, industries, and geographies. This information can then be used to better target LinkedIn Sponsored Content. Content marketers can now discover the content that is most popular with US CIOs within cloud computing, for example. These insights can then inform the content they create and publish to increase their influence, impact and inbound pipeline. Conversations around the ‘Wealth Management’ industry on LinkedIn and the share of voice for a brand’s product or service can be benchmarked, helping marketers to identify how to drive broader awareness and advocacy around their company.


Our customers have already highlighted how using LinkedIn Engagement Insights is like going from seeing everything in one dimension to a full 3D experience. The multidimensional analysis re-invents marketers’ campaign planning and delivery and we’re looking forward to working with LinkedIn on advancing the use of privacy-first data to inform business decisions worldwide. You can read LinkedIn’s blog post on the announcement here.

What our customers are saying

“We’re excited by Datasift’s partnership with LinkedIn because it solves a key problem in social insight – in that you get lots of words but don’t know a great deal about who read, interacted with or shared them,” Sameer Modha, Partner, Customer Data Strategy at Mindshare. “Being able to slice content and interactions using LinkedIn’s vast range of CV-quality data, combined with DataSift’s technology, we see lots of uses for LinkedIn Engagement Insights going into 2017.”

“I’ve worked with many social APIs and I’ve not seen anything as sophisticated as LinkedIn Engagement Insights,” said Noah King, VP, Director of Social Media, Socialyse. “The way the data is structured means it’s like going from seeing everything in one dimension to a full 3D experience. Even with this level of intricacy and the huge volumes of data it returns from a single API call, LinkedIn Engagement Insights is easy to use and set to radically change how the network’s data informs business decisions.”

  • Toby Beresford

    I still think there is value in the surface metrics for dashboards – they can help individuals and businesses self optimise. Good to hear you are getting some data out but without the identity data it’s unfortunately not something we will be able to use within our risefuse social selling success tracking programs.

Share This