Well, not all eyes. Some are going to be eagerly monitoring Twitter and Facebook feeds, in the hopes that the will of the people will reveal itself in hashtags and retweets. Social media has been used to successfully predict elections, World Cup soccer matches and even virus movement patterns.
Is it all a fluke? Or can social media really predict the future? And, if so, how can we use it to foresee the unforeseeable?
A digital crystal ball
Those of you who watched the 2010 World Cup might remember Paul the Octopus. This psychic cephalopod was able to predict Germany’s World Cup and Euro matches with an 85% success rate – not bad for an inky invertebrate.
We’ve come a long way since then. Now, people are using advanced analytics to predict the results of major tournaments. Bing was able to get full marks predicting the knockout stages of the 2014 World Cup (and has an equally impressive record when it comes to the Scottish referendum, the Oscars and the Cricket World Cup
Meanwhile, some companies are using social media to anticipate where the stock market is heading. Twitter itself sells data directly to banks and other financial service companies pertaining to sentiment analysis around various stock options. And Foursquare has accurately made a number of major stock and sales predictions
It’s a whole new world of possibilities out there. But it’s not as simple as running a quick algorithm and always getting the right results. Anyone looking to Twitter to foresee the recent England-Iceland shocker would have been disappointed. So what do companies need to keep in mind before looking to the social media crystal ball?
The end justifies the data
In order to make accurate data analysis, you need to start with understanding the final outcome required. By understanding what risk or result you’re looking to predict, you can better determine what data you need.
If you know that your final objective rests upon sentiment analysis of any kind – say creating a new service around the needs of your customers – you know that social media should be your first port of call. But your objective will also drive the other sources you use to gather data, as well as how much weight it gets.
Data, data everywhere
Social media is only one piece of the data puzzle. The more complex a situation and the greater the number of outcomes it has, the more data you need to be able to make accurate predictions. A robust data management strategy cuts across all of digital, and understands the context into which each different source of data falls.
Let’s take any one of the myriad healthcare initiatives out there as an example. Using social media mentions to track the spread of the flu might yield good results. But take that and add data gleaned from wearables and other IoT devices, and suddenly the accuracy of the results skyrocket.
Know your limitations and work around them
In data science, there is a saying: “What you don’t know, you don’t know.” And that’s the big thing to remember when using analytical tools to predict what might happen. We know that as machine learning
advances, artificial intelligences will better adapt themselves to be more accurate with every new iteration. But that’s not going to eliminate the need for strategic and creative thinkers overseeing the process, ensuring that the right data is being collected for the right purpose, and making the right decisions. Predictive analytics are powerful, but only if applied properly.