Algorithms, and, more recently, social media sentiment, have been billed as the one-step panacea for all of life’s problems. Tweak the algorithm just right and you get the perfect romantic partner. Read the social media sentiment correctly and you can become rich by predicting the ebb and flow of equities markets.
If only it were that easy.
Writing in Sunday’s New York Times, social psychologists Eli J. Finkel and Benjamin R. Karney question claims by dating sites like eHarmony, which promise us happily-ever-after with someone who has been “prescreened for deep compatibility with you across 29 dimensions.” And maybe its high time we applied the same degree of skepticism to all the social media prophets promising fail-proof predictions based on close scrutiny of Twitter’s API.
This isn’t to say things like a compatibility test or social media sentiment can’t help us make better choices. But as Finkel and Karney note, “things like communication patterns, problem-solving tendencies and sexual compatibility…are crucial for predicting the success or failure of relationships.
“Likewise, dating sites don’t take into account the environment surrounding the relationship: factors like job loss, financial strain, infertility and illness,” they wrote. “Another major problem with the algorithms of dating sites is that the information that they do collect – about individual characteristics – accounts for only a tiny slice of what makes two people suited for a long-term relationship.”
Extending that argument to social media, the lesson is don’t believe everything “they” tweet. If you did, Ron Paul would be assured a victory in November’s presidential election, Apple’s iPhone 4s was a disaster and Jon Bon Jovi is dead. Social media sentiment is one of those things that, when it’s right, the degree of accuracy is chilling, but when it’s wrong, it can be a very pointed swing-and-a-miss.
I’m endlessly fascinated by the potential of the predicative capabilities of social media sentiment and spend a good chunk of my time at ReadWriteWeb talking to people about it and covering it. I’ve been particularly interested in talking with people who are looking at the election and the equities markets with social media sentiment, but the quickest way to lose my interest is to tell me that you’re analysis is showing an assured victory for candidate X in state primary Y or that Ticker Symbol XYZ is going to see its share price rise, solely because social media chatter about the stock is on the upswing.
The people in the growing field of social media sentiment who are worth listening to are the first to concede that social media doesn’t hold all the answers, and that there are other factors to be included in any predictive model.
“None of this suggests that online dating is any worse a method of meeting potential romantic partners than meeting in a bar or on the subway,” Finkel and Karney wrote to conclude their op/ed piece. “But it’s no better either.”
Frankly, it’s too soon to tell if analyzing social media sentiment is a better or worse way of predicting what may happen next. And that’s the most important factor to keep in mind when someone tells you otherwise.