These days, social media monitoring and listening is almost becoming standard for most brands. But if you’re a large brand with millions of customers, actually doing efficient and effective social analytics on all that data can be a challenge. How do you handle an avalanche of tweets, messages and more, and analyze it effectively?
Let’s take a look at three keys to maximizing the ROI of your social media monitoring and analytics efforts:
1: Use all your data. The first thing companies must realize about social media analytics is that it should not take place in a bubble. Instead, it’s crucial to cross-reference social media sources with surveys, emails, phone calls, your website, review sites and even competitors’ websites. With advanced analytics, this data-gathering should take place frequently too – once an hour, if possible, as opposed to once a day.
By taking such an ongoing and integrated approach to social media analytics, companies are better equipped to determine how important different pieces of social data are. You need to know whether something that springs up on social media is an isolated problem or a systemic one. If you cross-reference a negative tweet with other sources, like call logs, you can tell whether or not it needs to be addressed immediately or at all.
Considering the half-life of a tweet is less than five minutes, an isolated negative tweet can likely be ignored. On the flip side, a plethora of negative tweets about the same issue can actually serve as an early warning system; thanks to proactive social media analytics, you can resolve the issue before it gets to the call center, where it will be far more expensive.
2. Filter out the noise. Of course, using all your data is easier said than done. That’s the paradox of advanced analytics – companies have been told they need to be customer-centric and listen to everything. But now that they’ve started doing that, they’re drowning in data.
The solution isn’t to stop listening – it’s to find ways to focus on what matters and filter out what doesn’t. Once again, cross-referencing social media data with other sources is one way to do that. But even before you get to that stage, you need to filter out the noise that comes in your social monitoring alone. All social listening tools have some element of spam detection to filter out the obvious junk, but you also need to be able to tag social media data that lacks content value.
That might mean a link with no commentary, a tweet with just one word, or even commentary that isn’t operational. For example, a tweet that says “Going to Wendy’s for lunch” doesn’t add much value to your social media analytics or customer interactions. On the other hand, a tweet like “Had a Wendy’s triple for lunch, best burger in the business,” is very operational and therefore much more important.
3. Consider context. Another key to successful social media analytics is to make sure you are integrating the unique characteristics of social language, from hashtags to emoticons to social-specific slang to industry-specific wording.
One aspect of this is a robust natural language processing platform that handles social media jargon. For example, the word “sick” is often negative, but it can be positive on social media in certain contexts. Your social media analytics system must be able to handle those nuances. Along the same lines, analytics that can handle emoticons and hashtags are becoming increasingly important for truly comprehensive social media analytics.
Another aspect of context is the acknowledgement that social media posts do not exist in a vacuum. Instead, they are usually part of a conversation – and your social media analytics must take this into consideration as well.
For example, let’s say Person A tweets about how she just bought L’Oreal’s newest lipstick shade, then Person B retweets it, adding something like “That’s awesome! So jealous.” Your social media analytics platform needs to understand that this is a conversation in order to understand what “that” is referring to and properly assess the sentiment being expressed. Without a nuanced social media platform, that detail will be overlooked.
The Bottom Line
Once you’re got these three fundamentals down, your social media analytics are ready to lead to action and improved ROI. You’ll be able to find the answers to key questions, like: Does this person appear to be a risk for churn? Is Person C an influencer in my market segment – and is he or she talking about my competitors? And you’ll be able to use those answers to focus on the social media conversations and interactions that will have the most impact on your brand and business.
In fact, all the keys we just mentioned mean nothing if you don’t take what you’ve learned and then do something with it. By integrating your data, filtering out the noise and taking context into consideration, you’ll be well on your way to making all that gathering and analyzing of social media data more than worth it.
About the Author: As leader of the Clarabridge marketing team, Susan Ganeshan defines the brand, leads the charge for educational, useful content, and enables both Clarabridge and its partners to promote and deliver on the promise of customer centricity. During her 25-year career, Susan has worked with organizations such as newBrandAnalytics, webMethods, Software AG, Deloitte Consulting, and Checkfree.