Academy Awards Wrap Up

February 27, 2012 | written by Kori Hill | Data Science

It was Hollywood’s biggest night and AddThis was along for the ride, tweeting data tidbits and measuring audience sentiment for some of the evening’s most stand out moments! Here’s a quick wrap up of what our network told us during the 84th Annual Academy Awards.

First up, prior to the event, we pulled some data on the nominated films and actors to see if public sentiment could predict the results. Here’s what we found:

Hugo and The Artist had strongest pre-Oscar buzz building off the momentum they seized at the Golden Globes and Critics Choice Awards

Best Actor winner Jean DuJardin didn’t see the highest spike, but he did see a good sustained amount of interest. It seems there is a correlation between consistent interest versus spikes in interest and winning the award. Consistency counts!

Our geographic map showing movie interest by state is especially telling. Not surprisingly, Hugo, which won 5 awards last night, saw massive interest across the board. And The Help reigned in the South, a movie which takes place in Mississippi.

As for audience sentiment, we saw many different emotions from users during the telecast. Here are the highlights:

Billy Crystal, whose monologue was poorly received, had 42% negative sentiment, but recovered as the night went along and ended up being viewed 47% positive by the end of the evening.
Rooney Mara saw a 4x increase in positive interest during her appearance on the red carpet.
Sacha Baron Cohen saw 90% positive sentiment after dumping “ashes” on Ryan Seacrest during the E! Pre-show.
Mr. Seacrest’s unfortunate luck didn’t just mess up his Burberry suit – he had a response of 68% negative sentiment after the stunt.
Overwhelming positive sentiment (97%) for Ellen Degeneres in the JCPenney ads.

We also saw some very interesting stats regarding service usage:

Proof Twitter is the service of choice for real-time events: Chatter about the Oscars on Twitter is outpacing FB by 45%
Sharing to Pinterest increased 150% during the red carpet show! Lots of fashion sharing going on.
Sharing via iPad was up 20% last night over a normal Sunday night.
Over 1.45M different domains received traffic yesterday with people looking for information about the Oscars.

If you like this data and are interested in following more stats from our network, be sure to follow @AddThis or the hashtag #ClearspringData or subscribe to this RSS Feed.

Web Intents Design Push Recap

February 27, 2012 | written by Jim Lane | Design, Travel

This weekend I attended the first Design Push in Brighton, UK. At this inaugural get-together a group of 20 or so designers and technologists focused on the problem of implementing Web Intents – see more background in my earlier blog post. It was a pleasure to not only meet other presenters from Google and Mozilla, but to also collaborate with a number of UK-based freelance user experience professionals for the day.

After an introduction by Danny Hope, Paul Kinlan from Google kicked things off by presenting a working demo of an image editing application built with Web Intents. Then, as part of my presentation about AddThis I had the opportunity to share with the team what bloggers and other larger publishers care about with regards to sharing tools and functionality. After my talk, Shane Caraveo from Mozilla provided a comparison of the original F1 sharing tool developed by the Mozilla team, and a more recent client-side interface they’ve been working on. Finally, Glenn Jones, one of the event organizers and founder of Madgex, posed some important questions that would help frame the recurring themes of the day:

How important is “social proof” – a propensity to share because a share count or a list of friends who also shared is displayed
How important is recognizability of social icons like Facebook or Twitter, compared to a button that simply says “share”
How should we balance the automated discovery of available “services” such as Facebook with explicit customization of sharing tools by publishers and “favorites” selected by their visitors?
How much of Web Intents belongs in the browser, versus on the page?

After our presentations and a survey of participants, we distilled everyone’s questions and ideas into several topics for breakout groups, including the interplay between the browser and the page, the linguistics of verbs and objects, and how to achieve critical mass of Web Intents.

Each group of 2-5 people met for several hours and then presented to the team at the end of the day. My team developed personas for different types of publishers and service operators to help the team understand ways that Web Intents could be beneficial, but also the kinds of questions and concerns each might have.

All told it was a productive get-together, and a great opportunity to meet a number of smart, passionate people. I’d like to thank Glenn, Danny and Andy again for inviting us to attend. While I was flying back to the United States they were working diligently to compile the day’s efforts in a wiki for public use and ongoing work. Look for additional details by following #pushwebintents.

Data Showing Rick Santorum Pulling Away from the Pack

February 24, 2012 | written by Greg Cypes | Data Science

Frequent blog readers know that we have been covering the election and other real-time events analyzing hundreds of millions of events a day across our network. We have been tracking the number of searches being done for each candidate and the data clearly shows Rick Santorum pulling away from the pack in the key Super Tuesday states over the last 30 days. It appears that searches are a good indicator of voter behavior as national and local polls show Santorum opening a 6 point lead. It will be interesting to see if the search data we shared earlier in January that showed Mitt Romney building momentum in ~~South Carolina~~ Florida prior to his win there, continues.

If elections or politics are not your thing, this Sunday, we will be live tweeting data from the Red Carpet at the Oscars. To follow any of our data science news, follow @AddThis or the hashtag #ClearspringData or subscribe to this RSS Feed.

Web Intents Design Push in Brighton, UK

February 21, 2012 | written by Jim Lane | Design, Travel

As Director of User Experience for AddThis I’m fortunate to have the opportunity to participate in design-related events and collaborate with other creative professionals. This week I’ll be attending the Web Intents Design Push in Brighton, UK with design and development leads from Mozilla, Google Chrome, and the W3C.

A Design Push brings designers and technologists together to collaborate on a variety of technologies. For this first Design Push, we’ll be working on Web Intents, which make it easier for site operators to offer visitors a more personalized way to interact with content.

Derived from a similar specification used by Android mobile phones, Web Intents determine what kind of actions (“share”, “save”, etc.) a user might want to perform on an object (a page, a picture), and which methods that particular user prefers. The group will focus on the various user experience and technical challenges specific to these kinds of interactions.

As we’ve built AddThis and OExchange, we’ve pursued similar goals in order to provide social tools that meet the needs that publishers have said they care most about:

Easy installation and customization of in-page tools
Performance, reliability, and security of any 3rd party tool
Increased traffic and audience via optimization and personalization
Support for rich embedded experiences offered by services like Facebook and Twitter such as Like and Follow
Comprehensive tracking and reporting
Opportunities for promotion and monetization

I’m looking forward to exploring ways that Web Intents can help publishers achieve these kinds of goals. Also, I’d like to extend our thanks to event organizers Glenn Jones, Andy Dennis, and Danny Hope for the invitation. If you’d like to follow the day’s activities on Twitter, we’ll be using the hashtag #pushwebintents.

Word of Whitney Houston’s Death Spreads ‘Round the World

February 12, 2012 | written by susan | Data Science

The world was shocked when news of musical superstar Whitney Houston’s death at 48 broke last night. Online activity spiked as people looked for details of what was first announced on Twitter and later verified by mainstream media. Interest in the singer who had 11 number one hits in the 80s and 90s spiked at nearly 1M% over normal activity.

The loss was felt worldwide. Interest in the US star was greatest in North America, and folks in Australia, the UK and Brazil were actively interested, as well.

In fact, most frequent searches related to the incident were in English, French, German and Italian. This word cloud shows the most frequently searches just after the incident.

Tragic though her death may be at such a young age, it’s clear the woman behind “I Want to Dance with Somebody” and “I Will Always Love You” affected hearts (and ears) internationally.

Phetric: Persistent Metrics for PHP applications

February 7, 2012 | written by Aaron Jorbin | Engineering

This is part two of three in a series on how we monitor application metrics on AddThis.com. See part 1.

Last month we open sourced two internally developed pieces of our metrics stack. Drew Stephens wrote about the first, MetricCatcher, which enables non-java applications to utilize Coda Hale’s Metrics Package. The second piece, which we use on AddThis.com to monitor internal application metrics, is Phetric.

AddThis.com attracts millions of page views each month across multiple servers on a complex site to which we are constantly adding new features and enhancements. In order to monitor our metrics during this process, we gather metrics with Phetric and MetricCatcher, sending them on to Ganglia and Graphite for recording and viewing.

Using Phetric

To initialize Phetric, first you must include Sender.php and call Phetric_Sender::init() which takes the following four arguments (the final two of which are optional):

The host where we want to send our metrics
The port where MetricCatcher is listening
A string you want prepend to all metric names
A boolean for if you want to send each metric as it comes in rather then waiting for the end. (off by default)

All metrics are sent (as JSON) using UDP, which is a non-blocking operation; sending metrics won’t ever hang your application. Phetric hooks into php’s shutdown event to send the metrics after your application code has finished. If you are debugging or running extremely long running application code, setting the auto flush boolean to true will mean you send all the metrics as they come in.

Phetric supports everything offered by Coda Hale’s metric Package except for health check. Take a look at the Phetric Readme to see how to implement each metric.

Verifying your metrics

There are two options for verifying the metrics that you are sending as you develop. The first is to point Phetric at an install of MetricCatcher and tail its logs. If you aren’t using MetricCatcher, you can trivially see the metrics that Phetric emits using Netcat. I have this function in my .bashrc to make testing easy:

catcher(){
    while true;
    do
         nc -w 1 -l -u 1420;
    done;
}

In Conclusion

Grab Phetric on Github, it’s available under the MIT License. If you want to extend it please submit a pull request. Have ideas or questions? Open an issue on Github.

If you haven’t already, check out Drew Stephens previous post on MetricCatcher, where our Phetric metrics end up. Also, check back here soon for the final part of this series.

Social Activity during the Super Bowl Up 143% YoY! What was the Internet Buzzing About?

February 6, 2012 | written by Kori Hill | Data Science, Fun

If our pre-game analysis was any indication of fan-base, the majority of the nation is happy with last night’s outcome. We analyzed millions of data points to bring you some highlights of the social activity that took place while the New York Giants took on the New England Patriots in Super Bowl XLVI. Activity was at an all-time high and amazingly, mobile sharing was up 500% from last year. Check out some of the online behavior that the game, commercials and performances inspired.

Bridgestone, Pepsi and Budweiser saw huge spikes in interest in their brand during the game. Anhueser-Busch ran four different ads, but you can see the final “Here Weego” spot had the biggest impact.

Interest is based on an analytics of related shares, clicks and searches:

Click for larger version.

As for the performances, perhaps we shouldn’t be surprised that M.I.A.’s much talked about obscene gesture had the most sustained interest throughout the 4th quarter and even after the game. Cee Lo wins for generating overall most interest.

Click for larger version.

Doritos, Budweiser and Teleflora were the big winners in terms of percentage lift in brand interest. Below you can see the lift in clicks, shares and searches for each brand relative to the day before Super Bowl.

Click for larger version.

Looks like Canada was more interested in Madonna than the US! This intensity map illustrates the scale of interest in Madonna due to her flashy halftime show.

Click for larger version.

And, finally, this map shows a geographic break down of interest in Budweiser brands in the US. Bud’s got some big fans in Colorado it seems. Maybe the folks in WV and KY need a little elbow in the ribs.

Click for larger version.

Some other great stats are around where all the activity was taking place:

Sharing via mobile was up 500% from last year
Overall sharing was up 143% from last year
Shares to and Clicks from Twitter was up 116% from last year
Shares to and Clicks from Tumblr was up 615% from last year
Shares to and Clicks from Facebook was up 18% from last year

Shares to and Clicks from Email were down 16% from last year

It seems Internet users are drifting away from email for much of their social correspondence and interacting more and more on social networks… and on mobile. Food for thought!

If you have any questions about our data or would like more information about AddThis and Clearspring, please feel free to contact us. We’d love to hear from you!

Social buzz has Giants favored in Super Bowl XLVI

February 5, 2012 | written by Kori Hill | Data Science

This is the second Big Game match-up for these two teams in recent years. The New York Giants beat the New England Patriots 17-14 in 2008 for Super Bowl XLII. But are citizens of the great United States hoping for an even score?

Not according to our data! We analyzed the search keywords, sharing behaviors and other social data points of millions of Internet users and found that the Giants are causing the most activity:

We’re also looking forward to analyzing our data for all the Super Bowl commercials. The big game is one of the most obvious illustrations of offline media affecting online behavior. Leading up to today, Honda saw a great spike in interest over the Ferris Bueller ad campaign this past week:

Can’t wait to see the other big winners.