Author Uses…Watson Analytics for Social Media

Blog Home > Author Uses…Watson Analytics for Social Media

Author Uses…Watson Analytics for Social Media

Author Intends To … Update Reader

We’re adding exciting new features to Watson Analytics for Social Media every day, and I want to update you on two of them.

The first is a big deal, and it makes your analysis much richer. We’ve added author behavior as an analytic output of all models. Let me explain: there’s a big difference between these two posts: “Teslas are cool” and “my Tesla is amazing.” In the first, the author is a person with an opinion. In the second, the author has identified him or herself as a person who has a Tesla (a pretty useful distinction for Tesla, don’t you think?)

This means you don’t have to continue to do something that, if you’re like me, you’ve been doing since you got started on Watson Analytics for Social Media: making themes for “cancelling” and “buying” and “using.” Now, all that work is done for you, on every model, every time you run it. You can now filter all charts by “users” et al, and you can also see the terminology consumers use on social media to express their current or intended actions (we call this field, “indicator that author…”).

indicatorwordcloud

With these three new segments, and their associated indicator text, your models are smarter, more accurate, and take less work to create. This author is (definitely) a user!

 

Author intends to…throw away the Post-It where she keeps track of how many documents she’s analyzed this month.

The second new feature available today in Watson Analytics for Social Media isn’t quite as sexy, but it’s incredibly useful: post estimates. Our users want to keep track of how large the project that they’re configuring is, and now they can.

estimates

Just before your data set is created, we’ll generate an estimate, so that you can keep tabs on your monthly usage and project size and never run out of data.

Estimate too high? Consider adding context terms to narrow your search.

Estimate lower than expected? Use topic suggestions to expand your search. Try adding relevant hashtags and alternate spellings or terms to your themes, or expanding the dates that you’re searching.

Whether you’ve been using Watson Analytics for Social Media for months, days, or you haven’t started yet, these features are available now, and I urge you to check them out today. I know that they’ll make Watson Analytics for Social Media even better at helping you answer strategic business questions that matter.

If you haven’t already, visit watsonanalytics.com to start a free trial of Watson Analytics for Social Media. If you have any questions please post them to the community, or message me directly at @allsnsmth.