Predict Social Success

Unmetric is helping social media marketers predict the impact of their posts.


The social media intelligence platform has launched a new tool called Predict, which allows marketers to input details about a planned piece of social content to see an estimate of the engagement the post will receive.


For example, marketers can input details including the post's structure, tone, intent and context. The machine learning algorithm then provides an estimate of engagement the post will receive based on the number of likes, comments and shares. Then, Predict provides recommendations for ways to boost the post's engagement and delivers new predictive results based on those edits.


"Building Predict from the ground up was a complex but rewarding project," said Aswani Yeraguntla, director, predictive and business analytics for Unmetric. "To start, we identified various elements of social media content in each sector. Then, our analysts tagged thousands of representative posts that formed the training set for building Predict. We explored multiple machine learning algorithms that accurately capture the relationship between different content elements and performance in terms of likes, comments and shares. These algorithms are at the core of our prediction technology."


It is important to note that Predict is being offered as a custom tool for individual enterprise brands leveraging the Unmetric platform. Moreover, each build is based on the brand's specific sector, desired competitors and relevant tags.


"Predictive analytics technology is the next evolution in enterprise software, and marketers are leveraging data to improve customer experiences and business outcomes in ways not previously possible," said Lux Narayan, chief executive officer of Unmetric. "We developed Predict to help brands stay ahead of the curve with their data-driven marketing strategies and achieve greater content results by bringing actionable quantitative evidence into the process of content creation."