Making Content Marketing Easier With Data Science David DuPuis January 26th, 2016 The information age has led to an explosion of public conversations occurring all across the internet. These conversations contain vital information about your marketplace, but without significant resources to sift through massive volumes of data to track these conversations, important learnings about customer perceptions, competitor activities, and influencers in your field will be missed.The ScribbleLive-Insights platform processes millions of pieces of news, blog, and social media content each day allowing users to uncover and understand the most relevant conversations in their industry. While the machinery and infrastructure enabling topic-relevant insights to be synthesized from millions of pieces of content is quite complex, the intuition behind the approach we use to quantify attention is quite simple.ScribbleLive-Insights Vs. Social MonitoringThe best way to understand the capabilities of ScribbleLive-Insights is to compare it to social monitoring tools. Insights and social monitoring tools both try to quantify attention. For social monitoring tools, this is done by summarizing basic metrics such as the number of retweets or the number of people following a company. This information can be interesting, but is unlikely to provide deep understanding of the conversations relevant to the user. ScribbleLive Insights goes beyond social monitoring in 2 ways:First, it considers conversations wherever they happen, be it on social media, news articles, or the blogosphere.Second, Insights will be able to flag conversations that are related to user's interest without the user having to specify (or even know about) these items beforehand. For example, consider a ScribbleLive user tracking the activities of a competitor. If one of these competitors releases a new product with a feature that is being widely praised on social media and in the blogosphere, Insights is able to a) alert the user to the increased attention being paid to the competitor and b) inform the user of the specific feature that is garnering attention.ScribbleLive-Insights does this by filtering millions of documents down to those containing keywords of interest (like a competitor's name), incorporating free text analytics to extract relevant themes, influencers, and entities (such as products, events, features, etc). Then we score these items based on the amount of attention each receives.What is "Scoring"?Reams of conversations relating to the interests of Insights users occur every day. Scoring is how Insights surfaces only the most impactful bits of information from these conversations. Scoring today is based off volume of mentions as well as the quality of those mentions. All else equal, an item with more mentions will have a higher score than an item with fewer mentions. And an item with higher quality mentions will have a higher score than an item with low quality mentions. Many things affect the quality of a mention such as the publication in which the mention occurs and the influentialness of the speaker of a mention. A mention by a prominent national politician in national newspaper garners more attention, and results in a higher score, than a mention in a local newspaper by a local politician. Once the quality and volume of all items are determined, attention scores are generated by the aggregation of the attention paid to each of the items.What’s Next?While our scoring algorithm allows for easily digestible, topic-relevant findings from millions of sources of content, we’re always looking to improve Insight’s capabilities, and we’ve just begun research on our next generation scoring algorithm. With the next iteration, we’re aiming to:Incorporate social proof (like the number of followers) into our assessment of the quality and impact of an individual’s mention.Increase the transparency of the mechanics of our data processing engine.Optimize to allow Insights to process more documents from new social networking sites.Enable a suite of new product functionalities.Want to see Insights in action? We dove into the data to help create the fourth annual CMO Report in partnership with Forbes and LinkedIn, and published a detailed report on the top influencers at CES 2016. We also used Insights data to round up the top 5 content marketing influencers at the end of 2015 so you know who to follow, learn from, and connect with in 2016.