Tracking Social Influence: Razorfish Files Patent For Social Media Action Tag

December 12th, 2008 by Garrick Schmitt    
Tags: , , , , , , , , , , , , , , ,

What Is The Value Of Social Influence?:

Razorfish has been trying to tackle a billion dollar problem: how do marketers identify the key influencers within a social network, and how do you really track and understand how social behavior works?

This problem is especially critical in a recessionary economy when every form of marketing is under scrutiny to prove its ROI.  We think we’ve cracked the code through a data tracking solution that’s going to sound kind of geeky to those of us who don’t measure data for a living, but essential to any marketer who cares about social media: the Generational Tag (AKA Incrementing Action Tag).

My colleagues in our Razorfish Seattle office filed a patent for the Generational Tag on September 30th, 2008 and we are sharing our findings publicly today.

The Data Problem:

Today social media apps (widgets, applications, viral media, etc.) cannot track or account for unique individuals other than those who download the application and those that download the application from a friend. Couple that with the lack of known quantitative methods for identifying key influencers within a social network in regards to a specific application and you’ve got a big problem for the industry. Some would say a billion-dollar problem, actually.

Core Questions We Tried To Answer:

  • What is the value of a key influencer?
  • How viral or how many generations has my social media application achieved?
  • What is the value of someone who receives a social application from a friend versus someone who receives it via media or a paid seeding strategy?

The Razorfish Incrementing Action Tag Solution:

Our solution is the Incrementing Action Tag which is a set of functions within a social media application that creates a variable that may be read externally based on where the user acquires the application.  When a user downloads the application from the original source, the Incrementing Action Tag notes the source and assigns the downloader the value of first generation (or one). When another downloader obtains the application from somewhere other than the original source (e.g. a friend, other website, etc.), the Incrementing Action Tag looks at the variable (or generation) assigned to the current source and increments it by one; thus making the next downloader generation two or other appropriate generation number.

The Incrementing Action Tag is thus able to identify (via a cookie and unique identifier and not through personally identifiable information) and track social media, identify how far removed (generation) cookies are from the original source of the social media, and identify key influencers (again no PII- see note above) of users of social media.

In essence, this technology enables our agency to create a system that allows us to value and reach key influencers across the Internet, regardless of property.

Here’s an example from a widget we recently tracked:

Successes Thus Far:

We’ve successfully used the action tag in three instances with three different clients.  We’ve seen as many as four generations of pass-along for these social media applications and are now looking  to begin using the Generational Tag on all social media applications so that we can build our knowledge of social media applications.

Thought this is still in its early days, we’ve definitely taken a big step forward towards tracking social influence across the web and maybe, just maybe, starting to crack a billion-dollar problem.

Congratulations to the Seattle team that made this all happen: Marc Sanford, Sandy Schlee, Steve Ebeling, Kelley Maves, David Niffin, Christopher Castle, Frank Kochenash, and Jesse Drogin.

For more on Razorfish’s efforts on Social Influence Marketing check out Shiv Singh’s work on the Going Social Now blog.


Del.icio.us     Digg     Technorati     Share on Facebook     Stumble Upon     Google Bookmarks     Furl     reddit

  1. 10 Responses to “Tracking Social Influence: Razorfish Files Patent For Social Media Action Tag”

  2. By Noah Carter on Dec 12, 2008 | Reply

    Awesome work!

  3. By Jordan Willms on Dec 12, 2008 | Reply

    Great work guys.

    http://www.sumolabs.com/blog/razorfish-files-patent-collecting-viral-distribution-metrics

    Hopefully I\’ve gotten it right. I am not sure whether you track generations from each source or just generations from the original source (Check out my post to understand where I am coming from).

    Jordan Willms
    sumolabs.com

  4. By Steve Furman on Dec 12, 2008 | Reply

    Very interesting yet amazingly simple. I assume this works only on applications/widgets. True? If so it does leave a lot of other social touchpoints and connections still untracked. Will continue to watch this concept evolve. Well done.

  5. By Jed Sundwall on Dec 23, 2008 | Reply

    This is a fantastic idea. I swear I advocated for building something very similar for Eventful\’s Demand widget while I worked there. I\’m glad to see that someone has finally built it and am looking forward to seeing what you guys learn from this.

    Excellent work!

  6. By Gina Carr on Feb 28, 2009 | Reply

    I’ve been watching Twitter.Grader.com and Facebook Grader. It seems you are taking this all many steps further. It is very complex to say the least.

    I’m anxious to watch the progress of trying to quantify social media ROI and identify the key influencers.

    May I put this up on my blog?

  7. By Dennis Yu on Mar 1, 2009 | Reply

    Guys,

    A viral tag that looks back at who influenced you– amazing. Now how will that apply to MLM? I would love to see the influence spread through multiple levels– quantified and visually displayed.

  8. By Jascha Kaykas-Wolff on Mar 2, 2009 | Reply

    This is very exciting progress; I\’d love to learn more especially as everyone in the marketing & analytics disciplines begin to drive the connection point between a service such as this, web channel operations, and campaign attribution generally. Practically speaking each small step at helping identify attribution only arms the marketer with access to more and more and better and better information to make their businesses better.

  9. By Garrick Schmitt on Mar 2, 2009 | Reply

    Thanks for the comments, Gina, Dennis and Jascha. We are continuing to evolve this and have several programs with clients up and running. The key takeway, for now for me, is the ability to measure “engagement” our interactions with a widget/app and its effect on the eventual business desired action (e.g. purchase). Stay tuned.

  10. By Dennis Lankes on Mar 17, 2009 | Reply

    So would this essentially be the equivalent of analytics for social media? How could you use this to track extended viewership and trends of online video audiences? Great post and great idea. I look forward to hearing more.

  11. By air jordan on Jun 29, 2009 | Reply

    Great work guys.

Post a Comment

This is a captcha-picture. It is used to prevent mass-access by robots. (see: www.captcha.net)

You must read and type the 5 chars within 0..9 and A..F, and submit the form.

  

Oh no, I cannot read this. Please, generate a