5 Tips for Analyzing Coronavirus Conversations on Social Media

Even with the best crystal ball on the market, no one could have predicted a pandemic would completely upend the way of life for people everywhere. 

The Coronavirus, and the disease COVID-19 have rapidly changed people’s experiences- how they consume, feel, buy and live their lives. 

How can you understand how this has impacted people like your customers and adapt in an appropriate way to help and serve them best?

While not a crystal ball, social media analytics platforms like social listening can help you track the millions of conversations about the coronavirus and how people are experiencing it.

Our research team leveraged this technology to discover tons of crucial consumer insights related to the crisis and gather it all on our coronavirus social insights page. 

In this article, we will discuss some key tips and considerations to keep in mind as this situation unfolds. Here are 5 tips for tracking coronavirus conversations on social media:

1. Separate General Public Conversations From News Publications

The first step in learning how people experience the current events is parsing out their online conversations from news sources to understand which issues, topics and concerns actual people have. 

There are three large buckets of social conversation that you may want to take into consideration during research:

General Consumer Conversation	Firsthand Consumer Conversation (I / me / we / our statements)	News

When comparing how consumers talk about certain topics related to the coronavirus vs. news sources, key insights, as well as some notable differences surface.

Consumer Conversation Themes	News Source Conversation Themes

Consumers discussed various topics like the pandemic response, staying safe, supporting those involved and social distancing. Over the month of March, these topics grew in conversation:

• Current response to pandemic - 108% increase in conversation

• How to help support and help fight - 90%

• What's happening today/tomorrow - 80%

• Social distancing - 71%

Meanwhile, News-created posts were more likely to mention these terms:

• “Global Pandemic”

• “Death Toll and Testing”

• “Breaking news on deaths”

• “Resources to help”

2. Understand The Noise

The next important tactic when monitoring social conversations about coronavirus is making sure you understand and break through the noise. 

It is easy for social media users to quickie retweet or share a post. Depending on the research topic, you may or may not want to include that data, so understanding what is trending the most is important. 

Below, Coronavirus conversation has been split into conversation that is exclusly organic versus conversation that is only retweeted or shared. Content that is retweeted or shared is 14% more negative that content that is organically written. 

Retweets and shares can also spike or decline conversation around a topic. The below graphic shows the variability of conversations that include retweets and shares. These variability can easily skew data to reflect point in time shares of tweets or posts. 

Pinpointing the conversation for the week of March 12th shows many retweets around Tom Hanks announcing he contracted COVID-19  and sporting events being cancelled left and right. By understanding the noise being generated via retweets and shares, you can start to understand what is making waves online and if you want that included in your analysis.

 3. Look at Organic Unbiased Conversations

Conversely, it could also reveal pivotal data points by removing content like retweets from your research. Why? It helps you focus on the organic conversation from consumers happening without influence or bias. 

By removing conversations from retweets, shares, replies, etc. and only looking at organic posts, you’ll see how people discuss issues naturally without reference to another topic or conversation from someone or somewhere else. This is a good approach if you want to understand individual concerns rather than breakout topics.

Take a look at the conversations mentioning “social distancing”. Notice the conversation topics change dramatically when comparing data of retweets versus organic conversation. Retweets contain topics that have been influenced by popular posts. Organic posts occur naturally from consumers without the influence of another post or share. 

The organic conversations (in purple) have more pertinent insight about individual people you’re targeting. Dig into that data to learn more about what these people are saying about their experiences.

By only looking at the organic conversation here, you’re getting people’s first impression thoughts and how they’re naturally expressing themselves without being influenced by other online voices.

4. Don’t Make Assumptions

This is no time for guessing. You need data and evidence to understand people’s experiences. 

The approach we’ve taken to all of our analysis is to really hone in on how people express themselves and talk about their experiences first hand.

If we examine consumer conversations about the coronavirus crisis, using the “I/me/we” statements we mentioned earlier and looking for the leading topics, phrases and themes, we see what people are talking about.

Furthermore, we see how the conversation is evolving over time as the situation changes:

Looking at these conversations and identifying the standout themes helps you understand what consumers are interested in, what their top concerns are, and how they describe their own experiences.

5. Understand and take into consideration the evolving conversation

During this uncertain time, the one thing that is certain is that the situation is ever-changing. What was true last week is no longer the case. As the times change at record speeds (is 2020 over yet?) so does the conversation online.

How consumers were originally talking about their daily lives and the Coronavirus has changed. 

As day-to-days transformed into social distancing lives for consumers, different hashtags and ways of talking about their experiences surged online. Hashtags such as #FlattenTheCurve, #StayHome, #StayHomeChallenge, #CoronaVirusLockdown saw a rise after the original hashtags such as #COVID or #coronavirus first took hold. 

To have effective analysis of the conversation, the analysis needs to follow the changes in how consumers are talking. You need data around people’s specific experiences and topics related to this situation. 

You’ll want to ask questions like:

• How do people talk about their lives in quarantine? Are they working from home more? Are they homeschooling kids? 

• How do people buy and consume? Are they ordering delivery? Doing curbside takeout? Do they get their groceries shipped in a box?

• How did people respond when sports were canceled? 

• What are people saying about a government stimulus bill? 

• How are people discussing layoffs and local business struggles? 

• How do people express emotions? Are they anxious, expressing more fear, trying to stay positive?

We aren't talking about a single point-in-time experience. We are trying to understand how consumers are feeling about the virus and those different particular point-in-time experiences (ie delivery on a specific day) specific to their own personal experiences. 

The conversation is going to evolve. So, you need to keep that in mind in any analysis you do during the coronavirus era. 

The answers to the above questions can be found using social listening. Do you have a unique situation that you need data like this? We’re offering free insights. Read below to learn more.

Having the tools and resources to serve customers in a helpful, ethical, efficient and respectful manner is more important than ever before. 

During this time, our team at Infegy is offering 1-2 customized coronavirus insights in real-time inside our social listening platform for you to leverage. This isn’t a sales pitch. We believe this is a critical moment for all, and we want to be here to help your team do right by your customers. Just click here to let us know your questions and reserve time with our team.