A CMO Storyline-Hadoop and Cognos Sentiment Analytics












  • The sentiment analytics dashbord was built by NES.  It was designed to illustrate how a “big data” solution, e.g., Hadoop or the IBM Puredata Data Warehouse Appliance Family of Products, can be integrated with a Business Intelligence/Visualization solution, e.g., IBM Cognos, to derive business insight from an “unstructured” social media data source, in this example Twitter. (I apologize that the graphic is not displying well in this blog)



  • CMO comes into his office in the morning.  He wants to see what is being said about his company on twitter.  His main concern today is with what people are saying about his company’s products, how a recent promotion is impacting sales and what general sentiment is on a recent corporate litigation.  The CMO also has interest in understanding at what point in the day (in this case tweet volume is measured in minutes) people are conversing about his products.  He finds this information feed helpful to determine when to tweet out offers about products so they hit the most people possible
  • As the CMO looks at the dashboard, what he is seeing is data being harvested from twitter, filtered and analyzed.  Looking at the graphs from left to right, you will see that he set the solution parameters to determine
1.)What the most active topics or hashtags that are associated with his company’s brand.
2.) We then analyze topics related to the CMO’s company to see what subject matter people are tweeting about the most.  As you can see, the subject legal/litigation is the most active topic on social media.  As we mentioned, this is logical given that there has been a recent lawsuit against his company.
3.) Focus then shifts away from legal to products.  What are people saying about the company’s products?  As you can see there are 5 products primarily being tweeted about.  Again this information has been derived by analyzing tweets, filtering out content to determine the subject of the conversation, comparing that information against a semantic dictionary to determine whether the tweet is positive or negative.  The graph reflects how many tweets are being generated about each product and what percentage of the conversations are positive or negative.
4.)  From the conversations about products, the CMO then wants to see how the social sentiment on his company and products are impacting the company’s sales. He also recently released a promotional offer on their soundlink product.  He wants to see if social sentiment, or tweet volume has a correlation to product sales. (you will see that there actually is a lot of chatter about the soundlink product, but the sales have not increased in a commensurate fashion w tweet volume)  The next chart marries the social sentiment information with data from the company’s CRM and ERP systems to show the sentiment trends in the data and how/if they are impacting recent product sales. (*Twitter data joined with static sales metrics as an example of what would be implemented in production)
5.) Lastly, the general tweet traffic about the company has been analyzed hour by hour.  This will show the CMO the trend as to when people are socializing most actively about his products, thus giving him information about when best to tweet out promotional offers to help drive sales
  • The CMO dashboard is an NES created analytics solution that sits on top of a big data platform (Hadoop).  Even if social sentiment is not a topic that is pertinent to your organization, this storyline was designed to give you a feeling for how large volumes of data from “non traditional” sources can be, aggregated, filtered and analyzed to make informed business decisions.
  • NES is a data and analytics consulting firm.  We have been focusing on data centric technology for 34 years.  Our solutions and services are designed to fit organizations of all sizes.  For more information regarding NES, please visit our website at www.nesystems.com, or contact me using the information below.
***Information for this example was gathered from Twitter around Bose Corporation
Liam O’Heir
Vice President

This entry was posted in Big Data, Business Analytics, IBM Cognos, Uncategorized. Bookmark the permalink. Comments are closed, but you can leave a trackback: Trackback URL.
  • Upcoming Events

    There are no upcoming events at this time.
  • About NES

    NES has been in business for 33 years and we’ve always been focused on data. But now, it’s about the liberation of data - with solutions like IBM Netezza, Big Insights and Cognos, we are educating and empowering our customers to make better, more informed decisions about their business and ultimately, to achieve their goals.

Log in