I have had the privilege of working with ShareThis, the leading sharing widget on the web, in partnership with SMG to conduct the largest analysis of sharing patterns ever published. Not based on surveys, but based on actual digital patterns of sharing across 7 billion pages of content viewing in March 2011 that have the ShareThis widget. This is the digital switch that we talked about at the Research Transformation initiative at the ARF; start with organically occurring digital streams, rather than the survey project as your source of insights.
In doing so, the facts are quite different from some of the musings you hear at conferences from speakers who do not have access to such data or who rely on surveys and anecdotal case histories. Here are some of the big points.
- Sharing is bigger than fans, friends, and followers. Facebook accounted for 38% of inbound traffic driven by sharing activity. E-mail was second at 17% but for some content categories, it was nearly equal to Facebook (business and investments, for example).
- Sharing is big. It accounts for nearly half of the referral traffic that search accounts for.
- Sharing is about scale not virality. There is actually very little content sharing and in-bound traffic that occurs past the first generation (meaning, someone clicks on a link that was shared.)
- Sharing is about relevance. 80% of those sharing a link only did so in one of the 27 content categories that ShareThis warehouses. 50% of the in-bound traffic comes from those who only share links from one content category. these two findings taken together (3, and 4) mean that the Duncan Watts hypothesis (that anyone can be an influencer in a given subject area that is relevant to them) is closer to the truth that the Gladwell concept of influencers.
- When someone does share multiple categories, there are categories that cluster together. This is important to know because the marketing opportunity for sharing is that it offers scale at finding audiences that are based on moments of relevance. Hence, for example, knowing that Arts and Entertainment and video games tend to be “co-shared”, helps to build audience size for marketers interested in either category. Similarly, parenting, health and pets go together.
For marketers, these new insights into sharing behavior are very important. They direct marketers away from finding magic influencers towards finding audiences who care about a given subject at a moment in time, at scale. Speaking about scale, behavioral targeting based on abandoned carts or those who are searching for certain words are fine, but are limited in scale. Finding audiences based on sharing behavior (sharing, clicking links, etc.) adds important scale and also, is somewhat up funnel at the place where branding messages work best, when brand consideration sets are still being formed.
Here are three ways you can get more information on this analysis:
- Contact me at firstname.lastname@example.org
- Read the Techcrunch article
- View the Internet Presentation on slideshare