Share with Friends

The Right and Wrong Way to Approach Social Media Data

The business value of social media insights has been proven time and time again. They can tell you why ride-sharing-application users are switching from Uber to Lyft. They can also predict consumers’ holiday shopping habits and what the hottest tech gift will be. And social media insights can give you valuable feedback from customers when they really love or really hate your product—like Chick-fil-A discovered when it changed its barbecue sauce.

Without the ability to draw conclusions from social media insights, companies cannot act on the data to regain customer loyalty, make accurate sales forecasts or improve products.

Before you can unlock the value of social media data and act on it, you first need to know how to approach it. Only then can social media data be used effectively to drive performance. Below are three principles to keep in mind before diving into social media analytics.

Keep in mind that people—not machines—produce insights:

However, the machines do empower the people. It’s important to use humans and computers for what each is good at. Computers are great at scaling and speeding up the social media analytics process, and humans are good at understanding language, such as sarcasm or jokes. When you combine machine learning algorithms for social media analysis with human understanding, you have the potential to uncover deep insights quickly and answer the most complex of questions, such as: Why do people buy my brand over my competitors? How is my product being used? What are my biggest business opportunities?

Start with a question, then find the data—not the other way around:

You can only get so much value out of a set of data. You do not glean a lot of insight when you take a data set and see what comes out of it, without any parameters. Instead, ask a question. Know what you want to get from the data and what you are trying to achieve. Once you have a question in mind, then you can find the data to answer that question.

Focus on identifying patterns in the data:

While data about how a campaign performs is valuable, it’s much more valuable to tap into data to gain insights and seek opportunities to create better campaigns. Find trends in the data and measure how that data changes, either over time or among different audience segments. This way, you can sharpen your targeting or improve your messaging to be more effective.

When you embed these principles into your social media analysis processes, you have a greater opportunity to optimise different aspects of your marketing, including the creative. For example, imagine you’ve just run a wildly well-received campaign. You want to replicate that success for your next campaign, but how? You can’t just run the same one, so you need to ask yourself, “What made it so successful?” Next you need to get the data that will answer your question.

Data from social media analysis can tell you what people are discussing most from that last campaign and how they feel about it. Beyond positive, neutral and negative sentiment, you can uncover whether they feel happy, surprised, angry, fearful, disgusted or saddened by the content.

You could also use image analysis to gauge what people are expressing when they share visual content about your brand or campaign. Are they sharing pictures of themselves with friends while enjoying your product? Are they using your product in an unusual manner?

Machine-learning technology makes it possible to analyze different elements in images, which can then dictate your approach and the elements you should include in your own creative.

Another way to use data to improve your creative is to use the social media data as the creative.

The latest Apple ads for its iPad Pro feature social media as the creative: real tweets from Twitter users comparing PCs to iPads. Apple could have taken it a step further, dug into the iPad versus PC conversation and featured interesting data in the ad creative. For example, it could have looked into the question of what embarrassing “fails” users faced. If it recognised a pattern that users have admitted on Twitter to having tried to tap, swipe or pinch-zoom a non-touchscreen laptop screen, that figure could’ve played out in an iPad ad: “50,000 of you admit you have accidentally swiped your non-touchscreen devices. Many more of you have done it but won’t admit it. Get on board with iPad.”

The application of social media data opens many opportunities to be smarter and more creative in business. Ask the right questions and use the right data, and the possibilities will be as endless as any of your social media feeds.

Ben Cockerell is the director of global marketing for Crimson Hexagon, a leading provider of social media intelligence software.