The IBM Study is Wrong. Twitter & Facebook Rocked Black Friday/Cyber Monday

Rob Bailey
29th November 2012 5 Comments

Earlier this week, IBM published a report that claimed that social traffic from sites like Twitter and Facebook only had a negligible effect on traffic to online retailers. Considering the massive flood of Tweets, Facebook Likes, and bitly clicks that all of us at DataSift saw about these major shopping events in our feeds, I was surprised by IBM’s findings and think they may have given marketers the very wrong impression that social traffic wasn’t important for retailers. Given the amount of press the study got, we decided to do a deep analysis of our data to see if our results were different.  We posted our summary Black Friday infographic here, and I’d like to provide a little more context.

What did we find?

  1. There Was a Massive Amount of Social Traffic
    We did a deep search on social sources, with a particular focus on Twitter and Facebook, the two most popular social networks that companies use for social marketing campaigns.  We analyzed a variety of keywords and hashtag terms and the numbers are staggering.  During the five-day period between Thanksgiving and Cyber Monday, we recorded more than 50M Tweets relating to Black Friday/Cyber Monday, with peaks of almost 100 Tweets per second.  We were also amazed at the distribution of the social activity.Sentiment by candidate
  2. Offline Retailers Did Really Well
    From what we could tell there were several million posts worth of coupons posted and shared across the social networks.  We were amazed to see how frequently they were for mostly offline retailers like Radio Shack.  Radio Shack even beat both Apple and Amazon in terms of virality around coupons.  As you can see from the infographic,  a lot of other “brick and mortar” retailers like WalMart did well. We think it’s likely that IBM massively underestimated the impact of social on retailers by only focusing on online retailers.  Indeed many of the coupons that we saw getting shared were for offline retailers.
  3. The Social Activity Was Geographically Distributed
    While there was a lot of social activity from states that are more commonly considered “wired” states like New York and California, we were amazed at the traffic that came from other states. Texas beat New York for social activity around Black Friday/Cyber Monday and Ohio, Florida, Pennsylvania and Virginia also had very impressive volumes of traffic. Even New Jersey almost cracked the Top 10, despite all of the ongoing fallout from the tragic devastation around Hurricane Sandy.

Social data provides an incredibly powerful source of information for companies to learn about their customers. Crucial to making use of this is working with a powerful platform that gives you accurate pictures of the relevant data you need quickly.

Rob Bailey

Written by Rob Bailey

Rob Bailey is the CEO of DataSift.

5 Responses to "The IBM Study is Wrong. Twitter & Facebook Rocked Black Friday/Cyber Monday"

  1. [...] the latest version of that argument, via social media analytics startup DataSift, which says that the IBM study is “wrong”, [...]

  2. [...] the latest version of that argument, via social media analytics start-up DataSift, which says that the IBM study is [...]

  3. [...] the latest version of that argument, via social media analytics start-up DataSift, which says that the IBM study is [...]

  4. [...] that DataSift responded with The IBM Study is Wrong. Twitter & Facebook Rocked Black Friday/Cyber Monday. DataSift said, “We analyzed a variety of keywords and hashtag terms and the numbers are [...]

  5. [...] flood of tweets, Facebook likes, and Bit.ly clicks that all of us at DataSift saw,” wrote Rob Bailey, chief executive of this U.K.-based social media data mining firm, “I was surprised by [...]

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>