Let me guess: your company has an outstanding social media presence. You do social listening every week. You have an 80% Share of Voice. You have a ton of retweets that showcase your “reach.” Sound familiar?

I know from experience that once you peel back the covers, take a look at how you’re getting these numbers, and, more importantly, how you’re interpreting them, that it’ very easy to pinpoint problems and inaccuracies. A lot of companies I’ve worked with seem to approach social listening as an attempt to showcase how great they are. Very few are using it for what it should be used for–to gain market intelligence and to better understand your customers.

With all the data and platforms available today, it’s easy to forget a simple rule: crap in, crap out. Meaning, the input into your social listening platform is the most important piece. If you put a bunch of junk into it, you’re going to get a bunch of junk out. It may look nice in a pie chart or executive presentation, but for those with a trained eye, it means nothing.

When creating your keyword inputs for your social listening platform, don’t just use branded terms. Meaning, don’t just create a list of keywords that are variants of your own branded terminology. For example, if you’re an athletic apparel brand, would you rather listen to a “real” conversation in a gym or a conversation in your own corporate building?

A smarter approach is to focus your social listening around keywords and categories that are unbranded. You want to get the pulse of what’s happening related to the topics you care about. This will give you a better sense of how the market is discussing these topics, not how you want them to talk about topics, using your branded language. You don’t want to just listen to yourself!

For example, think how easy it is to skew your Share of Voice metric. If you just listen for your own branded terms, I wouldn’t be surprised to see that your Share of Voice is 50% or higher. That’s absurd! Do you really think you’re dominating half of the conversations happening in the market? If you were listening for the non-branded conversations, you’d probably expect to see something closer to 10-15% (this obviously is entirely dependent upon your industry and a many other factors).

So make sure to use a critical eye when looking at your social listening output. A few recommendations:

  • Share of Voice – As mentioned above, make sure to use unbranded keywords as the input to your listening. Collaborate with your Paid/Organic Search teams to get a list of terms. This will give you a much more realistic view.
  • Impressions – Be careful with this one. Many times I see a report claiming that a Tweet “reached” 10,000 accounts and generated 25,000 impressions. Sounds great, but it’s not really true. Don’t assume that all of your followers actually saw your Tweet. Retweets from influencers are a much better metric to focus on.
  • Owned social channels – Make sure that your owned content is not skewing the results. It’s best to create a separate report for this. You don’t want your employees’ social activity, along with your own published content, to skew your results.

Apply some critical thinking to your social listening efforts to get a more realistic picture of how you’re performing.