Sentiment Analysis: Extracting Meaningful Information While Not Oversimplifying Your Data
This week, I moderated another Social Media Today webinar as part of their Best Thinker webinar series, this time on the topic of: Sentiment Analysis: Extracting Meaningful Information While Not Oversimplifying Your Data. This webinar was sponsored by Tracx and featured: Munish Gupta (@guptamun) is the Director of Marketing Analytics at Dell and Eric Berkowitz (@tracx) is VP Solutions Engineering at Tracx. We discussed the history of sentiment analysis and where we are heading in the future.
Here are 3 of the key takeaways:
- Sentiment Analysis moves to Sentity Analysis - The new way to do sentiment analysis is not just to look broadly at words in your stream, but dig deeper in the post to learn which each entity really means - hence the term Sentity Analysis. It's a combination of sentiment and natural language processing (NLP).
- Social Net Advocacy (SNA) a new social metric - Dell tracks over 25,000 posts/comments per day about every facet of their business so they started to look at how they could group them into streams that parallel their business and then look at the brand advocacy scores with each on a real time basis. They even go so far as to correlate them with their net promoter scores (NPS) and they are finding a strong correlation.
- The power of discovery - Both speakers talked about how important real time discovery was to their marketing. Within hours, and sometimes within minutes, of launching a new campaign they were receiving real time feedback from the social sphere.
To get a copy of the slides or listen to the replay please click here. You can also scan the highlights of this webinar on Twitter by reading the following Storify: https://storify.com/54ecc876fb613dde5636a338
The next webinar is Boldly Take Your Content Where No One Has Gone Before: Tactics for Better Content and Competition Analysis. Be sure to sign up for it or view the schedule of other upcoming webinars.
Follow Paul Dunay on Twitter