Social Media Marketing

Can Social Media Predict Obama’s Town Hall Answers?

Companies involved in social influence metrics believe they can predict the future based on what people waffle on about on sites like Twitter. Given our experience with Klout, we’re pretty skeptical to say the least.

However being a fluffy, we’re always open to new ideas and we’re also open to being proved wrong. With that in mind, we decided to carry out an experiment.

We asked TwitSprout (a new social media analytics startup) to see if they could predict what questions Obama will answer at the 2PM EST Twitter Town Hall meeting, based on what’s being said on Twitter.

Guest post by Dan Holowack (@dHolowack)  Co-Founder at TwitSprout

Dan Holowack

Dan Holowack

 

TwitSprouts Predictions

Obama Town Hall ExperimentResponding live to a nation of eager and inquisitive Tweeps is no small task. Is Obama prepared to step in front of the cameras at 2PM EST for Twitter Town Hall (#AskObama) and respond to the most pressing concerns of the American people? Perhaps he has already chosen the questions, rehearsed the answers. Could the entire event be scripted — and would we ever know?

TwitSprout has been crunching the numbers on over 30,000 tweets to predict which questions Obama will answer. President Obama could select 1 of 3 strategies:

1. Politics as usual: Obama selects tweets that align perfectly with his campaign platform. The Twitter event becomes more of a PR stunt than a genuine effort to speak directly to the people and answer the toughest questions.

We’ll leave this to the political experts, let’s move on.

2. Retweets! The retweet is a powerful action and a perfect indicator of shared interest. TwitSprout has analyzed both retweets and RTs to determine the most widespread questions across Twitter.

3. A popularity contest (Klout): The President might aim to engage with those who have high Klout. Perhaps by answering questions posed by the most influential tweeters they will be so delighted that they use their “great powers” to spread a positive message. We certainly hope this is not the case, but our predictions acknowledge this possibility as well.

TwitSprouts Magic Formula

The truth is, there is no magic formula. It still takes time and human eyes to gather meaningful insights from data. Separating the real questions from other promotions and garbage is the first challenge. Human language is still difficult for computers to understand. We have captured over 30,000 tweets with #AskObama and still thousands are posted each hour.

TwitSprout Obama Info Graphic

The initial heuristics indicate that just under 60% of the tweets are indeed genuine questions. After grouping the questions, we apply filters to identify and group similar tweets – these are typically quotes or RTs. TwitSprout then gathers retweet data on every question and profile information about the publisher (including Total Followers and Klout score). Finally, a member of the team works through the results manually, grouping questions into higher-level categories.

TwitSprouts Results

TwitSprout has identified the most popular tweets for the President based on Retweets and Klout. Here are their predictions…

1. Most Retweets

If all that matters is quantity, these are the tweets that have gathered the most retweets, quotes and RTs.

#1 (with 2,907 retweets): “Would you consider legalizing marijuana to increase revenue and save tax dollars by freeing up crowded prisons, court rooms?”

#2 (with 1,567 retweets): “You’ve said many times that the Bush Cuts for the 2% Should Expire. Can you promise to let them in 2012?”

#3 (with 750 retweets): “Mr. President, why should you not be held responsible for your silly prediction that unemployment would stay below 8%?”

2. Most Klout

If we consider only those tweets posed by users with the highest Klout, we see the following results.

#1 “Tech and knowledge industries are thriving, yet jobs discussion always centers on manufacturing. Why not be realistic about jobs?”

#2 “Mr. President, why should you not be held responsible for your silly prediction that unemployment would stay below 8%?”

#3: “Why do we have 1.5 million fewer jobs than we did before the stimulus when the # of ATMs is unchanged?”

Top Categories

We also grouped the final results by category to highlight the shared interests. There is of course some subjectivity involved, but the results are not surprising based on what we have seen in the data and recent events.

#1 Legalization of Marijuana (4,911 total tweets and retweets)

#2 Jobs (2,024 total tweets and retweets)

#3 Taxes (1,800 total tweets and retweets)

#4 Economy and Debt Ceiling (442 total tweets and retweets)

We have created a word cloud to help visualize question topics.

Our team will be watching the Twitter Town Hall event closely as we continue to collect and analyze tweets. We are very curious to see how the President will focus on specific topics and tweets of interest – by retweets and Klout.

As the TwitSprout team gleans new insights the Obama Dashboard  will be refreshed with the latest and greatest.

End of TwitSprout Guest Post


Flufflylinks.com would like to thank Twitsprout for their interesting post. We will also plan a follow-up post to analyse TwitSprouts predictions.

Don’t worry we will be merciless, if indeed TwitSprout fail to make any predictions. Maybe you can help us?

Stay tuned.

9/7/2011 Update:

1. Fluffy Links post: Twitter Town Hall Results

2. Fluffy Links follow-up guest post: Adrian Petrescu Co-Founder & CTO at TwitSprout on Why politicians can’t ignore Twitter and Social Metrics


Twitter Reaction 

Joe Fernandez @JoeFernandez
San Francisco

CEO and Cofounder of Klout. Just trying to get enough beans to make a burrito
@ I thought was great. what about you?
@JoeFernandez
Joe Fernandez
@ #2 & #3 in RT's were answered, but so were q's on higher ed, veterans and immigration. The last 3 didn't make any of ur lists
@AssignmentDesk1
Dorrine Mendoza
The power of social media! RT @: The Obama Town Hall Twitter Experiment http://t.co/HmKcQCr #AskObama
@Artang
Maral Artang

 


You might like these also


Tags: , , , , , , , , , , , ,