Subscribe to DSC Newsletter

Getting the Complete Picture from Your Brand’s Social Listening

Social listening has a nasty habit of being completely soft—all about the words without the context of volume and velocity of conversation—or completely quantitative with little information about what’s actually being said. Unfortunately, this leads to conclusions that don’t help you make better strategic decisions. Whichever way your listening leans, you can’t get a clear picture of a situation or your brand online without both the qualitative and the quantitative.

 

If your analytics are all about the numbers

Numbers are great. They’re concrete and factual. They have scale. And they really spice up your weekly or monthly report for the CMO. They make it very clear when something out of the ordinary occurred, and may even coincide with a campaign kickoff, product launch or big press release.  But what would your company’s leadership think if they learned that perhaps that spike represented lackluster response, or heavily mixed comments? What’s the real story in a spiked engagement or mention rate?

 

Using tools like Sprinklr, Radian6 or BrandWatch may give you some insights into what types of comments are in the spike, but reading a small number of individual posts won’t be totally representative. Watch out for some of these weaknesses in your quantitative analysis:

  • Not all topic spikes are made up of just one topic. To clearly see why a spike exists, you need to know all the key topics mentioned in it, and how many are truly relevant.
  • Spikes may be created by small groups. Even if comments on a post or mentions on a given day are double or triple the norm, this doesn’t mean unique engagement grew by the same amount. If it’s made up by only a few people, how influential are they?
  • Your spike could be made up of spam. One of the most common reason spikes occur in engagement or comments is an increased number of off-topic or irrelevant comments or mentions. If you don’t start with a clean dataset, your analysis could be extremely misleading.

 

When a company is very focused on quantitative indicators in social listening like the number of mentions in a certain timeframe or how many come from “influencers,” it’s difficult to determine what’s driving those conversations. For example, a former client—a consumer brand in the health industry—was hyper-focused on quantities of comments. They disregarded content analyses and sentiment in their reporting to hide that most of the comments were negative, or were accidental references at best. The report recipients didn’t learn about the brand’s sentiment issues until an independent review was conducted by its parent company’s digital agency.

 

Pro tip: Don’t buy into a purely quantitative report as a “first step” into reporting on social or digital marketing tactics. Without context around the conversations, you never know what problems or opportunities you’re missing.

 

If your analytics focus on word clouds and verbatims

Content analysis is a skill that is usually lacking in analytics reporting. It’s simple to count the likes, comments, mentions or shares on a post, but it’s a lot of extra work to sample them and analyze them for emotional sentiment, and assign them a category that is meaningful to your particular content strategy.

If your reports focus on qualitative elements such as word choice or sentiment, how much do you know about the size of one spike versus another? Did it last for hours or days? It’s critical to know if your program just gives positive results, or if those positive results represent growth in audience engagement and increased message penetration.

  • What is the content velocity? In analytics, time is a key element (one of many). It determines speed and quantity of engagement, as well as the timeframe they occurred. It provides context to sentiment by qualifying total numbers of positive and negative comments, as well as how many were received in a given period.
  • Difficulty identifying true influencers. Qualitative analysis allows you to identify those who speak at an expert level at a topic; but, without overall conversation quantities to provide context around their actual influence, it’s difficult to determine if a person is “just” an expert, or an influential one.
  • Comparing categories for sentiment and topic. If you perform content analyses on social conversations about your brand, how do you determine which categories of conversation you should pay the most attention to? Of course, the topic is key. How are the messages made up? What is the sentiment? How do the conversations compare to your brand’s messaging? However, you need to quantify those categories; if you have one very positive category that aligns very well with your outbound messaging pushes, but it only makes up three percent of conversations about the brand, does it really show message penetration? No, it doesn’t. If it makes up 40 percent, that’s a different story entirely.

 

Qualitative indicators such as sentiment, message repetition or hashtag adoption don’t tell the complete story either. I’ve worked with brands that were very concerned about how they were being discussed, but the quantitative elements were ignored. It didn’t matter to them if they had three mentions a month, or 300.

 

Consequently, they missed the fact that their brand was mentioned very little in general, and that no matter what the content of the conversations was, it didn’t represent any significant volume. The real takeaway that was missed is that their brand had minimal recognition or acceptance as a thought leader.

 

TL; DR: You don’t have to pick

To get actionable reports from your social listening activities, you don’t have to, nor should you, pick between purely qualitative or quantitative reporting. The yin and the yang of examining both will create a more actionable view, leading to stronger strategic decisions.

Views: 514

Tags: analytics, data, listening, marketing, reporting, social, twitter

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Videos

  • Add Videos
  • View All

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service