Last month we took a look at how people were celebrating Mother’s Day, so it’s only fitting that we analyze activity around Father’s Day this month.
At a first glance, we found that Father’s Day is a significantly less social holiday than Mother’s Day. The week prior to Mother’s Day had three times as many social mentions as the week prior to Father’s Day. Continue →
The protests in Turkey have undoubtedly caused a flurry of online activity. Our data allows us to see how people are sharing URLs in relation to the protests (for example, through Facebook, Twitter, and email), and how the protests affect behavior online. Here’s what the data has shown.
First, the volume of the events jumped in the end of May. Continue →
People spend an increasing portion of their time online, so it’s not surprising that their daily rhythms (i.e. work schedule, commute, etc) are reflected in their online activity.
Our presence on 14 million websites worldwide gives us a unique perspective on this behavior, and the opportunity to use data science to help our publishers understand their users. Here’s an example comparing mobile vs. desktop users by hour over a typical week.
With the weather already heating up, it only feels appropriate that Memorial Day weekend, the unofficial start of summer, is right around the corner. We decided to take a look at what social content users are sharing the most, and where they are sharing it, to get a better idea of how Americans will be spending their three-day weekend.
Last year we took a look at how people were preparing for Mother’s Day, and this year we thought we’d bring you a new take on the holiday.
We analyzed the difference between the types of people who bought Mom a gift a couple of weeks ago, versus people who still haven’t bought anything for her yet. This information’s based on shoppers looking for Mother’s Day gifts online in the past few weeks!
It is that time again for America’s favorite sporting event, the Super Bowl. The ads, the halftime show, and of course the game all come together for a 4+ hour event that is celebrated with friends, family and food.
As we did last year we took a look at our data to see which brands are getting the most pre-game buzz and to see if the same data can help determine who is going to win the game. Continue →
Given how social continues to evolve quickly, I thought it would be interesting to look at several representative publishers leveraging our social platform and assess what’s happening in the social space overall. To keep things simple, I’ve selected a large publisher (based on uniques) in three verticals. Internally, our deeper analyses have shown each to be representative of the aggregate of the vertical with respect to social behaviors of their users.
Through our AddThis social platform, we empower publishers to provide and track all kinds of social activity on a site – from copy and paste of the URLs and text in page to posting to social services to printing/emailing. We have provided these social services for five years and are now installed by publishers on 14+ million active domains touching 1.3B unique users monthly. We are 10x our closest social tools competitor and, as a result, have an exceptional view into web-wide behaviors.
In this analysis, we assessed:
Shares – a user sending content to their friends / social graph (via sharing button, email or IM)
Clicks – a user shares, then another user that sees that share clicks on the link (or, if you prefer, a clickback)
Profile of Action – a breakdown of service or share type of the social behavior.
Here are some topline insights from the analysis…
-News sites see the highest virality (as measured by clicks over shares), then e-commerce and food. News sites tend to have short shelf life, highly viral content, such as articles around Frankenstorm, crimes or Olympic victories.
-The e-commerce site exhibited lower virality than the news site, but greater proportion of emailing and printing. We didn’t reflect it in the analysis but the shelf life of the content was longer. This is likely because this site had more expensive, considered purchases. The food site exhibited greater pinning but the lowest virality overall. The content shared were recipes and exhibited the longest shelf life.
What is interesting is not only how much users are sharing but how they share, and how the preferred mechanism for sharing changes by vertical. Alexis Madrigal of The Atlantic touched on this in his piece called Dark Social: We Have the Whole History of the Web Wrong. While his title implies something ominous, it’s a scientific reference to dark matter being all around us. In our analysis, users are clearly aware of how public/private they are, why they share and what social brands or tools satisfies their need(s) based on their current task. Facebook is the mental anchor for social, but user behavior varies, and is both locked-in and evolving.
Learned behavior – The simple act of copying and pasting an URL drives significant traffic across all categories. It also remains one of the most impactful sharing mechanisms (as measured by virality, e.g., URLs shared by copy/paste exhibit the highest clickback). This behavior is different by medium (mobile vs PC) but is ingrained into consumer behavior. The profile of copy/pasters varies greatly (it’s exhibited, for example, by casual up thru heavy users) and has become a habitual behavior that isn’t changing over time.
Evolving – Pinterest’s growth has been extraordinary. As Pinterest adoption grows across categories of publishers, we are seeing specific verticals adapt faster to quickly changing user behavior. Pinterest has been working to enable easy pinning, and AddThis rolled a pinning tool back in July. We continue to see social behavior evolving, and Pinterest won’t be the last. Sites and technologies that solve a pain point or make actions easier will be adopted. Innovative companies like Sulia recognize that social is maturing and are attempting to build mass niche social communities.
There are interesting implications of this changing consumer behavior. On the one hand, consumer habits are hard to change. On the other, social activity is maturing and fragmenting rapidly. If you look at AddThis global stats, Facebook represents ~50 percent of social activity to 3rd-party social services. However, when you include social activity such as copy/paste of the URL, the percentage decreases significantly.
Stay tuned, we’ll touch more on questions this raises next . . .
As we approach what is traditionally the most anticipated shopping day of the year, marketers are looking to social sharing habits of shoppers for more and better clues into what items they are sharing, what deals they are searching for, what networks they are participating in and who they are talking to.
So, if you aren’t already camping out in front of a Best Buy, here’s a fun Black Friday infographic made using AddThis data. Stats include top shopping personas, popular products and the retailers who stand to win big:
Click image to enlarge.
By now you’ve certainly heard the news that President Barack Obama will spend another four years in the White House. The AddThis network was extremely active as election results were flowing in and when the official announcement was made. See what activity looked like in what was the most social presidential election ever.
A Close Prediction
Prior to the election, we predicted how results might turn out based on online social activity – velocity and quantity of mentions and sentiment. The preliminary swing state prediction was extremely close – see this example in Virginia. Below is a comparison of our predictions to actual results as reported by our friends at NBCNews.com.
AddThis Prediction Data
Actual Virginia Results from NBCNews.com
The Announcement
Social activity related to Obama really started picking up as the media began reporting his victory. The social mentions graph below – showing shows how often a term appears in AddThis shares, social referrals and searches – shows that social activity related mostly to Romney early on, giving way to Obama as the results became more conclusive.
Click to enlarge.
The Social Breakdown
Last but not least, our service data shows that Facebook and Twitter were the top mechanisms for sharing articles about the election last night, with an even split between Facebook Like and Facebook shares. A social election indeed.
After 2.5 years and over $2B spent by both candidates, we have arrived at election day. Pundits from both sides are predicting comfortable victories, but what does the data say?
If online behavior is any indication, neither candidate should feel all that comfortable and the margin of victory will be close. Check out the infographic we put together based on AddThis network data since the conventions ended.
Stay tuned as we’ll wrap up the election coverage tomorrow and see how close our data matches the results. Most importantly, if you are a US Citizen exercise your right to go vote today!
