Facebook’s emergence as the platform du jour for social networking and the Internet marketer made me start thinking about how to “do web analytics” on the Facebook platform.  Even Eric’s moved the Web 2.0 Measurement Group to Facebook.

Apparently Facebook is thinking about Web 2.0 measurement too.  A few days ago Facebook posted this blog entry.  Facebook claims to be measuring engagement based on touchpoints in applications across four areas:

• Canvas Page Views 
• Link Clicks in FBML
• Mock-Ajax Form Submission
• Click-to-Play Flash 

They are calculating a number of “Daily Active Users” from midnight to midnight each day by “putting” the touch points together.  Well that’s dandy, but it assumes all touch points are equal and biases the measurement toward applications that have more touchpoints… In fact, they’re only measuring engagement with applications in the most liberal of definitions.  Take a closer look, and I’d have to agree with Jeremiah Owyang that Facebook may have their terms confused.  These metrics measure Interaction.  Where are the frequency and time measures necessary for engagement?

Regardless it’s a good start for creators of Facebook applications!  But what’s a Facebook user to do?  Well I’ll tell you, my fine reader.

Over the last several months I’ve learned that there are several methods:

• Facebook’s API  and FQL.  The Facebook API is a REST interface, like Feedburner’s.  You can use it to add “social context” to a Facebook application using profile, friend, photo, and event data. Facebook’s FQL is a SQL-like language for querying the platform. Cool. Developers are psyched.
• FBML and Google Analytics. You can use the FBML Google Analytics tag to count the number of page views you’ve had on the canvas page of your Facebook application.  That’s one part of the FB engagement metric.
• Facebook Applications.  I’ve found a whole bunch of cool Facebook applications that provide unique ways of understanding the Facebook network.  Each application provides a slice of analytic-like functionality. 

Here are some of the social media analytics applications that I’ve been playing around with on Facebook:

• Friend Wheel.  Spin an interweaving mandela of ties to your all your nodes.  This cute app captures how all my friends related to each other, and also uses Web 2.0isms like “click to embiggen.”  Kewl.  Check it out:

• Friend Grid.  This app displays a little grid of your friend’s Facebook pictures.  It updates itself too.  That’s handy when you have a loose Facebook friending policy, like me.  Or like The Scobleizer who just today reached the limit of 4,999 friends on Facebook.  Because of Friend Grid I now know what some of my Facebook friends look like.  Heh.  Do you recognize any of these people?  I recognize most of them.

• Friend Sets.  Create multivariate syllogistic like visualizations about dimensions of your Facebook friends as you define them.  Don’t know what I mean. Check it out below (and note that I did not create any of these sets… they are ”presets” from the creators):

• Interactive Friend Graph.  This app is a cute little tool that provides a multinodal visualization of the ties that bind your Facebook friends.  You can rollover a circle to view the persons full name and an overlay mapping of their connections.  Then you can click on the circle to view their Facebook profile, send a message, poke, or add someone to friends.  Check it out:

• Socialistics.  By far this super cool application is my favorite for doing analytics on Facebook.  It gives you a bunch of insights into the relationships between your network including an intriguing amount of demographics and influence-based characteristics.  It can generate a multitude of tag cloud visualizations, pie charts, and assorted visualizations about things like gender, location, influence, relationships, and more.  Check out these cool screen captures below.

Here’s a distribution of the educational institutions of my Facebook friends, those Ivy leaguers:

Here’s a distribution of the political beliefs of my Facebook friends, those liberals:

Socialistics also has tag clouds of my Facebook friends.  The tag cloud on the right shows the popularity of friends within my personal network by name. The one on the left by picture. I’ve shrink it all to protect the identities of the guilty (mostly )

                

As Facebook goes more mainstream and social networking become more ubiquitous in the business world, we’re only going to see an increasing demand for tools that help measure activity, behavior, demographics, opinions, and influence on social networks.

While these applications aren’t enormously powerful and or very engaging for business purposes, they represent a widgety beginning of new type of new media analytics.  I’m excited to see how all this Web 2.0 social networking stuff will continue to play for out for “web analytics.” 

Integration of social networking analysis features into current offerings from web analytics vendors could take social media measurement into new exciting areas full of profitable revenue.  I envision many uses of social networking and social media analytics for online business:

• Helping companies realize new products.  Imagine the lessons to be learned from 227 groups with thousands of people discussing new product development.
• Identifying social trends impacting their business.  Anyone want to learn about social characteristics of those who believe in alternative energy and/or boycott Exxon, Citgo, and Shell? 
• Enabling larger enterprises to more proactively respond to the voice of the customer and manage risk.  Are companies like Walmart, Coke, McDonalds, and Nestle listening to the thousands of voices?
• New revenue models for behavioral marketing and targeting campaigns.  Maybe this is a long stretch, but could cost-per-target (CPT) ads be very far way?  It seems obvious to me to say that social network analytics will be used to target ads and offers in smaller batches focused on high-value niche segments that are typically hard to reach using mass media broadcasting techniques.  Facebook seems to already be investigating this niche. 

Do you Facebook or use other social networks?  Are you interested in analytics for social networks?  What do you think? 

I’ve been taking a look a deeper look at the trends in social networking and the analysis of social networks using nodes (such as taxonomy) and ties (such as clickstream data).   A few concepts from networking theory are intriguing me, and I figured I’d bring them up here to see if anyone has any thoughts:

• Betweenness. Identifies the degree to which a node in a social network is interrelated to another node.  Identifying degrees of betweenness in taxonomy nodes and combining with “normal” analytics data could enable the analyst to:
  • Detect nodes with the most betweenness to identify content that should be *automatically* served when a visitor interacts with a related taxonomy node (extending site optimization technologies)
  • Determine misappropriated editorial agenda and withering products by contrasting the “popularity” of nodes with the most or least betweenness.
• Clustering.   A concept used to express how visits relate to core taxonomy nodes could:
  • Provide a means for visualizing how visitor segments cluster around particular pages or nodes in a taxonomy
  • Enable the analyst to visualize the broad content themes that drive the most visits
• Density.  Certain bloggers and site pages tend to see larger numbers of repeat visitors, comments, or maximized time-based metrics when compared to other pages.  Can a metric for “content density” of a site be calculated?  Perhaps by crafting a equation from counting objects in a taxonomy node, value-scoring each object, and seeing which objects were interacted with most frequently?
• Influence.  The guideline is 99% lurk and 1% influence.  Can we gauge visit “influence” and visualize it from:
  • Pathing where visitors who have performed the most/least interactions and contributions ”go next” off-site.
  • Value scoring an “influence metric” for Interactions, Contributions, posts and comments, and off-site exit links in each visit, then adding up the values to calculate a new influence-based KPI measurement per visit.  Finally comparing the “influence metric” across all visits.

If you are still following me ( ), what I’m working at understanding and reconciling is whether social network analysis theory when combined with web analytics can illuminate the analyst with new ways for thinking about a web site. 

By combining a rules-based approach to processing this type of data, the possibility for automatic content targeting and the idea of a “living site” self-optimizing based on visitor interactions with taxonomy nodes or site objects becomes closer to reality.  The potential to use analytics data and social networking theory for building and realizing new combinations of product, content, and design becomes possible.  For example, I could create rules and logic commanding my CMS fill a “related topics” module or widget on a particular page with content from nodes that have the smallest amount of betweenness and the greatest density.

It’s clear that social networking impacts web analytics. Most major analytics vendors don’t seem to be thinking about applying (or how to apply) concepts from social networking.  I’m looking forward to vendors bringing social network theory into their technologies by perhaps combining, rules-based algorithms for site optimization with existing analytics data and new, open API’s (for example, Facebook’s new API or LinkedIn’s forthcoming API) to drive profitable revenue from new and existing channels.

Virtual Worlds and Web Analytics… Y’all ever play around with Second Life or World of Warcraft?  I have.  I think the concepts and worlds are very, very interesting and fun.  I find their messaging around the analytics of their user base even more entertaining though.  It’s like looking at ComScore and NNR for accurate web analytics data… really fascinating demographic stuff of questionable accuracy outside the frame of their audience panel and technology.  For example, I have three avatars, but have only downloaded one client. The trends are compelling though…

Some CMO’s I know won’t touch Second Life with a virtual ten foot, paisley, polygon pole.  Some finance folks I know laugh over beers about Linden Dollars.  Does that mean specific corporations become a central bank setting monetary policy subordinate to the central bank in the server’s home country?  How do International Fisher Relations apply when you have no interest rate?  My friends who have physical bodies say “virtual worlds are for when you have no friends in the real one.” Harsh criticisms, but they don’t negate the fact that something is happening and people are participating on some scale.  We’re all going to “do web analytics” on virtual worlds some day (maybe sooner than we think).

Where are the API’s for analytics data from these companies?  I believe Linden Labs announcing an analytics API would help push adoption by marketers forward and increase spend rates.  When I look at emerging technologies for 3D online collaboration, like OpenCroquet, I see the end of walled gardens like Second Life and WoW unless they open up the platform:

“Second Life doesn’t create a computational environment that belongs to its users - it uses a constrained computational environment (its servers) to capture “eyeballs” for a variety of schemes to derive revenue from them. With Croquet, users/developers may freely share, modify and view the source code (due to Croquet’s liberal license), the technology is not hosted on a single organization’s server (and hence governed by that organization as was the case with ViOS and now with Second Life), and it provides a complete professional programmer’s language (Smalltalk/Squeak), integrated development environment (IDE), and class library in every distributed, running participant’s copy (the programming development environment itself is simultaneously shareable and extensible). Croquet based worlds can also be updated while the system is live and running.”

Other online collaboration environments that would benefit from an open source of verifiable measurement include:

• Uni-verse.  An “open source Internet platform for multi-user, interactive, distributed, high-quality 3D graphics and audio for home, public and personal use.”
• Muse. A “software platform allowing organizations to create collaborative custom solutions that utilize rich media, 3D environments, and multi-user capabilities. Using Muse, developers can create immersive 3D environments that unite video and animation, audio, html, 3D models and much more.”
• Virtual Object System.  A “free and open platform for multiuser 3D virtual reality and interactive, collaborative 3D virtual spaces, and collaborative data systems in general.”

And the big guys and gals over at Microsoft and Sun are experimenting too (where’s Google and Yahoo? - do tell me!):

• Microsoft’s Task Gallery.  A “novel approach to bring existing, unmodified Windows applications into a running 3D virtual environment. The result is a working platform for experimentation in 3D user interfaces, in which the user retains all familiar productivity tools. This also allows for a smooth transition between traditional 2D interfaces and our new 3D territory.”
• Sun’s Looking Glass Project. A “Java technology and explores bringing a richer user experience to the desktop and applications via 3D windowing and visualization capabilities.”

Notice what all of these visionary ideas have in common: openness.  It’s only through open standards to key interfaces in these systems that we web analysts will be able to do what we do.  

So that beckons the rhetorical question, which web analytics tools right now could even work with extended data models for 3D virtual collaboration environments? 

I’m looking forward to how management at the following companies evolves their business models to focus on openness through analytics enabling their sustainable growth rate:

• There.com
• Multiverse
• Forterra Systems
• Whyville - ahh, my favorite question “why?”
• ProtonMedia
• Entropia Universe
• Habbo
• Areae

As Marshall Sponder forms the Web Analytics Association’s Social Media working group, I’m looking forward to hearing your voice on the phone calls.  Make sure you also read my good friend Eric Peterson’s take on some of this area as well.

This past Thursday evening I went to an event at Mistral in Boston hosted by an Internet consultancy named Molecular. Molecular began its web business in the mid-1990’s. They “did” Fidelity’s first website. That’s cool stuff in my book. Since then, they’ve done so much, and are now linked by Isobar.

My favorite part of a Molecular event is the opportunity to listen to smart people speaking about Internet innovation. Not to mention the fine food, good wine, and bright sommelier (who digs the first growth)… From a semiotic perspective, such a well-planned and engaging evening tells me a lot about Molecular as a company: focused, creative, organized, smart, connected, and successful.

So why I am telling you this on an analytics blog? Well, the evening’s topic was “user generated content”:

Today, technology has given customers control to determine what messages they will listen to and when they will listen – as well as a means to let their own voice be heard. This may be difficult, as it is much different than what we are accustomed, but denial of the customer voice will not make it go away - it’s only getting louder. Only marketers who can learn to adapt will remain successful in the jungles of untamable content.

Effective marketers must learn to utilize user-generated content to their benefit by creating authentic, positive, and valuable ways to engage customers in a “conversation” and incorporate their voice. During this provocative discussion, panelists will share their insight on this concept, its challenges, and benefits. These marketing experts will share their real world experiences and insight into such issues as managing, surviving, and spinning negative content, as well as maximizing the advantages of the positive.

UGC is powerful stuff. The mainstream internet and media meshing has made it unavoidable. Has what I said influenced your opinion about Molecular? Made you want to eat at Mistral the next time in Boston?

So how do you measure User Generated Content? That was the question I asked to the speakers from Reebok and TripAdvisor at Thursday’s party.

The good news is that both companies claim to use web analytics to measure UGC, and, like everyone it seems, looking to do it even better. That means making better use of existing data, deploying or upgrading technology, and/or extending their data model.

So I was thinking about making better use of existing data by working with and segmenting metrics and dimensions.

UGC dimensions could include:

• Event:
  • Post
  • Comment
  • Interaction (with types: play, pan, zoom, edit)
  • Contribution (with types: mashup and file)
• Visitor
• Persona

UGC metrics could include:

• Value scores.
• Counts of inbound/outbound links and new/return/repeat visitors.
• Search metrics, like organic search visits and visit rate.
• Time-based metrics, like total time online per visitor and average visit frequency and duration. 

When the web analyst creates this type of mental model for measuring UGC, selecting new technology or working with your geeks to extend the data model becomes more a lucid, focused activity.

For example, I could take a look at some cool UGC and:

• Value score events subordinate to the page view.
• Value score an engagement level of those events.
• Multiply the two together to generate a type of engagement metric.
• Identify the “Event Path” with the highest engagement.
• Identify the “visitor” or “visit” with the highest engagement. 

Then I could wield the extended data model in my analytics tool to identify the following online behavior and better understand my UGC during a period:

• Ratio of:
  • events:visitors
  • events:visits
  • contributions:visitors
  • contributions:cookies set
  • visitors:personas
  • comments:new posts
  • comments:existing posts
  • posts:visitors
  • comments:visitors
  • mashups:visitors
• Percent of:
  • high/medium/low contributing visitors
  • high/medium/low interacting visitors
  • high/medium/low engaged visitors
  • new posts
  • new comments
  • new mashups
• Number of:
  • events per page
  • interactions per contribution
  • comments per post
  • mashups created
  • linked posts
  • contributions per persona
  • visitors per persona
  • total events by post, comment, contribution, interaction

I know I could create other derivatives and use other metrics too. Events, like Interaction and Contribution, need more edification in future posts, but I think the beginnings of this model are clear.

The User Generated Content revolution doesn’t just affect Web business. It’s becoming part of modern capitalism whether you make sneakers or sell ads. This revolution is making web analytics an even more critical process in your value chain.