Social media is a key marketing channel brands can leverage to build win-win brand-consumer relationships. Such win-win relationships deliver value to the consumer through trust and greater satisfaction with brands and value to the brand through greater profitability by effectively responding to consumer needs. For brands to be successful and build such win-win relationships on social media channels, data must be leveraged to drive strategy. This study aims to use social media data and data science techniques to systematically characterize Land O’Lakes, our client partner, Facebook’s posts. These characteristics are then to be used to provide novel insights Land O’Lakes can use to drive a successful social media strategy on Facebook and build win-win relationships with their nearly 210,000 followers on Facebook. Using concepts such as computer vision and natural language processing, we will characterize key aspects of each post such as the dominant color of the image, polarity of the headline and the business context in which the post relates to. Using these characteristics, we will highlight key trends that can be leveraged in Land O’Lakes strategic initiatives on Facebook. These features will also be leveraged to build a model that can classify the post based on engagement rate performance. By leveraging data science techniques to characterize posts in ways that have not yet be done at Land O’Lakes, this work will aim to generate novel insights that will alter the strategy of Land O’Lakes on Facebook to ensure their content is catered to the specific needs of their audience.