Blogger: Larry Cannell
This week I had an interesting conversation with Tynt, a company that provides a unique method of measuring what people find interesting on websites. In a recent blog post Derek Ball, the CEO of Tynt, contradicted statements made by TechCrunch, which essentially said FaceBook was the most popular method people use to share information found on the web. Based on data gathered by Tynt’s service Ball concluded:
“Clearly, e-mail is by far and away the largest source of social sharing.”
Based on Tynt’s data, for every one person clicking a “Share This” button (to post a snippet on a social networking site like Facebook) 50 people (!!) grab content using copy/paste. A significant number of these copied pieces of content are shared via e-mail. The chart below from Tynt shows the breakdown of how people share information gleaned from a webpage:
But this only gives a partial picture. Just how effective are these links at generating visits back to the originating site?
While content shared via e-mail appears to be less effective at generating clicks back to a site (probably because it is being sent to and read by fewer people), it still drives almost half of the resulting visits.
Beyond these observations regarding the use of social media versus e-mail, I find the technology behind this fascinating. Like many web analytic products, Tynt’s approach involves embedding Javascript into a web page. Javascript is used to infer or sense what a user is doing and communicate this data back to a central analytics server. By tracking which links are clicked and where the user moves their mouse, an analytics solution infers what the user finds most interesting or engaging on a web page.
However, Tynt’s approach is different. They have discovered ways to hook into a browser’s copy/paste mechanisms. By doing so they can track what pieces of content are most engaging (because the user was compelled to copy them) and can also modify the text placed into the paste buffer. These text modifications enable Tynt to rewrite links to be redirected through a central server so clicks on these links can be tracked, as well as embedding other text to encourage the link to the clicked. For a simple example, this is how a Gmail composition window looked when I copied/pasted text from Ball’s blog post reference above (note the addition of “Read more”):
It will be interesting to see what opportunities these copy/paste hooks might enable and how the web analytics market responds to this deeper perspective on user behavior.
Comments