Rival Insights

Abstract

competitive business environments, businesses need a deep look into their customers, competitors, as well as their market. These businesses require a systematic way to collect, monitor and analyze all of the online data generated by their own customers as well as their competitor’s, on social media sites and blogs, in order to find trends, performances, customer bio, potential customers’ location and many more insights directed towards that particular market/industry. These sites have a huge amount of text available, where people communicate their truthful opinions, recommendations, responses about the products/services that they have witnessed or used. So, rather than conducting surveys or distributing questionnaires among customers, businesses can easily interpret valuable insights obtained by analyzing the social media content. These insights will be shown on a single platform through our web-based solution. Current working systems provide basic sentiment analysis based reports to show what their customers are saying and how their businesses are performing. We are using text-mining algorithms inside our social-media competitor analytics platform, that will provide analysis reports of businesses and customers (within a specific market) for enhancing market intelligence strategies, and pointing-out some reasons why business are leading or lagging at a certain time. These results will accurately help them more in planning their marketing campaigns, product research and reputation management. Our proposed system will work with unstructured content, freely available on major social media sites like facebook, twitter, reddit, wordpress, tumblr and some other blogs. Some of these sites require specific page/group credentials for data to be accessed, and after a lot of usage, many data will be loaded in our servers from different markets that our system will require a faster means of analyzing texts in real-time. This will require a distributed system that will execute different tasks among multiple nodes, which can be achieved using hadoop and spark. This will lead to extracting even more insights that we have extracted in the current version of our system.