In this post:
Full Chart of Results, Why We Focused On Facebook Fan Volume vs. Twitter, But Twitter Isn’t Bad Either, Forecasts Are Only Going To Get Better
Can we have a do-over? Our theory was right, our sample was wrong. If we did it again, we’d hit it out of the park. And if we did it in 2012, our results would be even better. Gallup and all the other pollsters should be afraid, very afraid.
Since the September primaries, we’ve been toying with the theory that the size and growth of candidates’ Facebook and Twitter followings could predict election winners. To lightly indulge our curiosity, we started tracking the social network size of the top two candidates in eight key races—and then we made some predictions.
(Facebook fans and Twitter followers tracked between September 21, 2010 to October 26, 2010.)
Fortunately, we’re not the gambling type. Of the eight races, we only got four right—view full chart. But, all cards on the table, it’s probably because we didn’t track enough of the races.
According to a Facebook’s political team:
- More than 70% of the time, the candidate with the most fans ended up winning the race.
- Of 98 House races, 74% of candidates with the most Facebook fans won their contests.
- In the Senate, they found of their initial snapshot of 19 races, 81% of candidates with the most Facebook fans won their races.
We agree. We ended up basing our predictions on Facebook fan volume.
Why We Focused On Facebook Fan Volume
First, we noticed that this was the larger and still more aggressively growing number. Therefore we suspected it would be more telling. Of those we tracked, the average percent growth of Facebook supporters was 90 percent with 11 of the 16 candidates having more than 20,000 fans. The average growth of Twitter followers was just 24 percent with just two of the candidates having more than 20,000 thousand followers.
Then we realized how valuable a Facebook fan is. Liking a candidate on Facebook means making a public declaration of support. It’s the 2010 version of a bumper sticker. People use Facebook to make their preferences known so that their friends and network will hop on the bandwagon. Following a candidate on Twitter is a less social act. It doesn’t mean you’re voicing support. It just means you want to follow their announcements.
But then again, that’s not chump change, either. Social media scientist Dan Zarrella found that in 71% percent of races he tracked, the candidate with the most Twitter followers was ahead in the polls. And The New York Times, in a stunning chart, demonstrated how much more chatter was created by Republican candidates, addressed to Republican candidates and retweeted by followers, than was centered around the Democrats.
Forecasts Are Only Going To Get Better
The size of a candidate’s social network will become a better and better election prediction tool.
- First, more people will be on social networks. It only took about 6 months in 2010 for Facebook to gain 100 million more active users. And Twitter gets 1.8 million new users per week. This means the sample size will be larger. Plus, the demographics of the social network user are likely to increasing mirror that of voters as digital natives get older.
- Second, if the data is more than 70% accurate with all Facebook fans and Twitter followers counted, which includes people under 18, people in other countries and voting districts, how much more precise would the prediction be if demographic data was used as a filter? Imagine pulling out the stats for voters in particular constituencies, then narrowing it down further to analyze age groups.
Gallup, Rasmussen, Pew Research Center—be afraid, be very afraid. You may be the next industry to get disrupted by digital.
