Social data science: How much are 1000 words really worth?

11th June 2013 2 Comments

We’ve all heard that a picture is worth 1,000 words. After seeing some of the visualization projects that our data science team was working on in anticipation of last week’s announcements regarding managed data sources and our Tableau partnership, I can confirm that the old adage holds true.

One of the projects that I thought demonstrated the true power of social data analysis and visualization explored the connection between Facebook Pages, Instagram and the release of some of this year’s biggest movies.

Visualizing a Movie’s Social Media Presence

Paul Gesiak, one of our lead data scientists looked at overall social media presence. In this chart, each dot represents aggregated likes for individual Facebook posts.

The first thing I noticed is that Sony’s After Earth, appears to have the biggest social presence, both in number of posts and in overall popularity. You’ll notice that Marvel/Disney’s Iron Man 3 seems to have the most popular single post but because we’ve aggregated all untitled timeline photos, this visualization represents multiple posts.


Detailed analysis reveals that After Earth received twice as many likes per post as any other movie’s Facebook page. After Earth also has the most likes overall and the most popular single post.

Paul also looked at Instagram posts that included a movie #hashtag and the results provides a significantly different perspective on presence. Iron Man 3 is a clear winner, with After Earth failing to make the top 5. It’s clear that Sony’s concerted effort to build awareness prior to the movie paid off on Facebook.


Consumer Interaction

After analyzing each movie’s overall presence, Paul began to review other data and discovered an important trend with implications for all studio marketing departments.

This image charts out Facebook posts by movie over the months leading up to the premiere. You’ll notice that with many of the movies, the interest in posts increases steadily until it spikes within a week or two of each movie premiere, after which there is a sharp drop off in interest.


But when we look on Instagram, interest is low until there is sharp spike during premiere week. In some cases, interest even continues to build for a few weeks as photos are posted on the premiere date and for some time afterwards. It’s fair to assume that lots of people are taking pictures during their movie night.


Based on these findings, it would appear that when launching a new movie, it may be best to focus pre-release efforts on Facebook and then use those efforts to drive post-show efforts on Instagram. An example? How about announcing an “at the movies” Instagram contest on Facebook?

While these findings are unique to the movie industry, imagine the trends you could uncover in your industry using social data & visualization tools like DataSift and Tableau. Other examples include visualizations for last week’s LeWeb social media dashboard and this week’s Tableau User’s Group in London. You can also join us for one of our upcoming webinars later this month:

Extracting Insights from Corporate Social Data Sources:

Facebook Pages, Instagram & Google+
Date: Thursday, June 20, 2013
Time: 9:00 am PT / 12:00 pm ET / 5:00 pm UK

Blending Social & Business Data for Better Intelligence:

Tableau & DataSift
Date: Thursday, June 27, 2013
Time: 9:00 am PT / 12:00 pm ET / 5:00 pm UK


Share This