Blogs

Here's what happened when we took Watson Analytics for Social Media out for a spin

Social Media Manager, Internet of Things, IBM

Continually trying out new social media analytics, content management system (CMS) and reporting software is important when you work in social media marketing within any industry. Evaluating these tools is just part and parcel of being an expert in the social media realm.

In the course of my career, I’ve tried out what seems like every single piece of social media software there is under the sun: Curata, Expion, HootSuite, Percolate, SpredFast, Sprinklr, Sprout Social, Sysomos, TweetDeck and Zoho Social, to name a few. And while many of these products have their uses, few of them are unique; they all have competing counterparts of similar form and function.

Offering a quick list of disruptors

As a general rule, you don’t find many social media tools that stop you in your tracks. This statement is supported by the truth that I can remember vividly each time I discovered a product worthy of stopping in my tracks—a product that made me say, “Wow! This thing is amazing. Bravo!” Which products have generated that kind of reaction? Here are my 2015 picks: 

  • MeerKat: This tool hit the scene in February 2015, and got on the radar during the South by Southwest (SXSW) conference in March. MeerKat offers live-streaming, easy and interactive video integrated and distributed through Twitter. It’s brilliant; however, it was quickly supplanted by Twitter’s Periscope, which is the current go-to product. But MeerKat was there first, and that’s worth remembering.

And the trend continues in 2016, but the very first stop-in-your-tracks moment for me came from a surprising source—an internal email that went out to all IBM social business managers. It invited us to try out a new product: IBM Watson Analytics for Social Media.

Piquing my interest

IBM is not really a top-of-mind company when it comes to social media—yet. However, it’s been doing really interesting things to drive actionable data insights, such as acquiring The Weather Company and partnering with Twitter. These moves on the chessboard have to mean something, right?

In looking at the Watson Analytics landing page, clearly this tool incorporated real-time, cognitive, geospatial analytics into something actionable. And this particular version for social media was made for my kind of people. I signed up for an IBM ID and took Watson Analytics for Social Media out for a spin—a 15-minute spin, to be exact. Fifteen minutes is the usual amount of time I allot to figure out if a piece of software is worthy of consideration. I went into this trial with skepticism, but came out of it feeling pleasantly surprised by a number of key takeaways: 

  • It’s fast and easy to use with minimal onboarding.
  • It offers cognitive themes and topics that make it smart enough to help you tailor searches and obtain insights into what’s related.
  • It provides preconfigured visualizations for sentiment, geography, authors and demographics.
  • Its complex, unstructured social media data is presented quickly and in a consumable manner.
  • It allows you to ask questions in ways that are intuitive to you and get answers. The logic cuts through the need to type or click through dozens of filters.
  • It helps simplify the problem. For example, you usually look at a huge data set that is sort of unintelligible, and your first reaction is: “What the heck do I make of all this? Someone call our team of data scientists stat!” Watson Analytics for Social Media helps eliminate this reaction.

Reviewing my likes and providing a wish list

Applications for this software are easy to see and wide ranging. In politics, it enables seeing who is trending positively and negatively by state. For defense or early response issues, it can help identify when there’s been an incident. Retailers can look at what’s gaining popularity in any trending season—for example, identifying which colors are trendy, or determining whether people prefer gift cards instead of coupons based on department. 

As a marketer, for me Watson Analytics for Social Media offers a great way to report on share-of-voice and sentiment metrics. Honestly, it’s the best tool I’ve found so far. It’s a good, fast and simple way to identify influencers and trending topics. I enjoyed not having to fuss with any Boolean. And my first few queries were lightning fast. But later in the day, things did slow down some. I alerted the team supporting the product and received a helpful reply quickly. 

I admit that I was a little confused on the difference between topics and themes, and how to tailor topics specifically. The help documents were useful, but I wished they’d listed additional options. This area can definitely be a bit confusing and require more than 15 minutes to understand.

Including Twitter data is incredible, but I wish the sources of other data were clearer. From which boards is it pulling data? Where is all this data coming from? To find an answer to those questions, I learned that Watson Analytics for Social Media is analyzing—or will soon be capable of analyzing—data from many major areas of relevancy on the web. These areas include certain platforms that are often overlooked by other tools.

And for reporting, I’d like to see these visualizations exported. For now, screen captures are required, but reporting may be a future enhancement that’s already planned.

Seeing is believing

If you’d like to take a walkthrough of Watson Analytics for Social Media, watch a video depicting a colleague using it to solve a retailer problem concerning toys and merchandise planning in anticipation of the release of Star Wars: The Force Awakens.

Of course, you can always try out a complimentary trial version by clicking the Try it for Free button.

[Editor’s note: This is a brand new product, and a few days back, confirmation emails weren’t going through to certain email domains. Today, everything should be resolved, but if you should have any trouble, please leave a comment.]