Whether you love it or hate it, there’s no escaping it. Americans open up social media apps on their phone 17 times a day – more than once an hour, assuming you get a full night’s sleep.
Given how much time people spend on it, it’s no surprise that many companies view social media as a data goldmine. The sample size is high – as of 2016, 78% of Americans have a social networking profile. Twitter alone has 313 million monthly active users as of Q2 2016. It’s also much cheaper (or even free) to access that data compared to traditional data collection techniques like surveys, panels, and focus groups.
Many companies see these statistics and immediately begin devoting time and resources to analyzing all the social media data they can get their hands on. Surely, they think, there must be valuable insights in here.
We certainly won’t deny that there can be value in social media data. However, the hard truth – and one that more and more of the companies we work with are beginning to realize – is that it doesn’t contain the wealth of insights expected. Instead, companies are dealing with truths like:
There’s less data available to scrape than you might imagine.
Many social media outlets are tightening their privacy settings in response to pressure from their users. As an example, we frequently get asked whether Facebook posts are a good data source. We have to explain that only a tiny percentage of posts and data on Facebook are publicly available.
A surprisingly large percentage of social media users are…. bots. This is especially true of Twitter, which is also the most readily accessible social media data source.
And even if you can find enough data from real users…
People probably aren’t talking about you as much as you think.
Companies who have asked us to analyze their Twitter data are invariably surprised when the total volume of tweets about their brand/company is much lower than expected. No, sorry, the Twittersphere is not talking about what they think about the new layout in your company’s dressing rooms (true example).
Insights are extremely hard to come by, and businesses find it difficult to take action on what they see in social media posts.
Tweets are general by nature, due to limits on length. Also, unless customers are tweeting directly at your company, you are most likely not the intended audience – so any mentions of your brand will likely be in passing or of a very general nature. (“Cheeto time!”, “Can’t wait for drinks tonight – gotta pick up some Sam Adams,” etc.)
What are you, as a market researcher, going to do with that information? What business action can you take as a result of analyzing this? For most companies, the answer is “nothing.”
So, how should social media data be used, if at all?
We would argue that companies should move away from trying to use social media as a replacement for surveys and product reviews. Instead, think about that data – particularly people tweeting at your company helpline handle – as just another channel in your contact center. (And if your company doesn’t have a support account on Twitter, consider implementing one. Almost 40% of US adults have contacted a company via Twitter, and that percentage has almost certainly increased since that report was released.)
Look at all the data you’re collecting from your support channels, social media included, and ask yourself: What types of comments are you receiving via social media? How, if at all, do they differ from what you’re hearing from other channels in your contact center? Using social media data to augment your customer support is far more likely to yield actionable findings compared to scraping sites and hoping to uncover a deep insight.