How to Predict Trends with Social Media Monitoring

Social media monitoring enables you to hear conversations in real time in order to engage, measure, track, and learn from brand and industry mentions. But what if you took that a step further and used social media monitoring to share predictions for your community?

Sickweather provides predictions of illnesses based on public Tweets and Facebook mentions of sickness or illness. By capturing this data and analyzing the geographic location, they can pinpoint where illnesses are popping up. In fact, Sickweather's predictions based on social data helped declare an early start to last year’s US flu season, six weeks before the CDC.

I was inspired by the concept of helping your community by providing predictions and insight based on real social data. So I sat down with Sickweather's CEO and co-founder, Graham Dodge, to learn how they do it and what brands can learn from this approach to social media monitoring.

Great to meet you, Graham. What inspired you to start Sickweather?
It all started with a stomach bug. My first inclination was to go on Facebook and see who else had it. I found that others were suffering from the same illness and sharing it online. If I could look at this from a much bigger scale and look at it in terms of weather patterns, it could be very powerful. I teamed up with two others who gave gravity to the idea and helped get it off the ground. That was November 2011.

Since then, flu season calculated a 730% growth in site visits over 2011. We announced the flu season before the CDC last year, which is a tremendous case study for us. That story went national and was covered by the Today Show, National Geographic, and more. We track illness symptoms and previously we tracked 22 illnesses -- now we're watching 27. While that doesn't seem like a lot, it really is. The amount of data and bandwidth to monitor each illness is massive, so increasing it is a big deal.

Why is social media monitoring important for predictions?
Being able to predict major illnesses is huge. We focus on the healthcare vertical, but social media monitoring for predictions can extend beyond this.

What tips can you share about how to make predictions?

  • Look at historical data against real-time data to make predictions. The real-time piece is critical to forecast accurately.
  • When looking at the data, editorialize. Automation is great, but human eyes are critical.
  • Look at other data sets such as illness patterns against weather patterns. This gives you a road map to predict even further ahead. You can see where you're going to go. For instance, if a big event is happening in a big city, it will likely lead to in more illness in that area. The 2012 Super Bowl resulted in more illness in the Indianapolis area during the week of the Big Game. So don't analyze your data in a vacuum.
  • Look at sentiment analysis. Depending on how your audience feels about your brand or campaign can impact how your upcoming efforts will perform. For any vertical to make strategic buys for your brand you need to see what's happening now and the upcoming trends. For instance, if you're a cough syrup brand and you know that the East Rutherford, NJ area could be the next spot for illness after the 2014 Super Bowl, then geo-target your campaign to that area.

How can brands find unique ways to create great content from social media monitoring insight?
This data makes for great visual content such as infographics, maps, interactive APIs and more. Share stories and testimonials that stem from your predictions. Do "before and after" blog posts to show what you predicted and where results landed.

Do you have a great story to share from your community?
A mother Tweeted that because of our forecasting, she was able to get medication for her child before they caught the illness. The child had a compromising immune system and this proactive approach helped immensely. What we're doing is not just sharing content and information, but we're making the world a healthy place.

Thanks for your time today, Graham! If you have any questions for Graham or the Sickweather team, Tweet them at @sickweather.

Graham Dodge, CEO and Co-Founder, Sickweather

Sickweather LLC is based in Baltimore, MD. In 2012, Sickweather was honored among "100 Brilliant Companies" by Entrepreneur Magazine, and has been featured on the Today Show, NPR, CTV News, BBC, MSNBC, Al Jazeera, Mashable, Discovery Channel, and Fast Company. Learn more about Sickweather here.
Comments are closed.