Social Networks and Influence Analytics at the Web Analytics Day


Dave Rhee’s Influence Analytics for Social Networks presentation

“Wandering” Dave Rhee here – reporting on last Friday’s second Web Analytics Day organized in Brussels by OX2, for which I had the honor of delivering a presentation about one of my favorite subjects, Social Networks, and more specifically, analytics which can help determine the influence of individuals within those networks.

Overall, the day went really well, and not just because the venue was well-chosen and food and cocktails were great, but because we had a very large proportion of actual web analytics practitioners present, and just enough of a vendor presence to help answer the questions which inevitably arose. Although a few people had to leave in the afternoon, the vast majority chose to stay until the very end (a good success metric!), and even though the day was full of information, the attendees were so engaged that we usually ran out of time before running out of questions.

Microsoft’s worldwide first look at Gatineau was certainly a surprise highlight for most attendees, and emphasized how much Microsoft values its European audience for Web Analytics tools, as well as how highly regarded OX2 is in the global web analytics community.

Watch this space for video highlights of all the presentations, including Stephan Loerke’s opening keynote from the World Federation of Advertisers, and Eric T. Peterson’s usual brilliance in summarizing difficult concepts in a way that makes perfect sense to the rest of us. And of course, our own Aurélie Pols, whom you know from this blog, provided an overview of running a web analytics implementation project smoothly, given her extensive experience in helping OX2 Clients avoid common problems with their own sites.

I led off my own presentation in the afternoon with the worst slide of the day: a very dense table comparing several media, including press releases, surveys, blogs, audio and video podcasts, discussion message boards, social networks like Facebook or Myspace, and immersive environments like Second Life. The comparison attributes included style, “coolness factor,” ability to control your message, the value of feedback you can collect, the level of intimacy and trust possible, what resources are required to commit to creation and maintenance, the primary risk factors, and finally, the best use for that type of medium.

Fortunately, the bulk of my talk considered the typical and ideal lifecycles of discussion boards (like the Web Analytics Forum, which I moderate), as I believe they have the highest long-term potential to cultivate a true community within your customer base — the sort of fans that give you lasting influence in the marketplace, and who help you create value and spread the word in a way that most “viral” campaigns usually fail to do.

We discussed which metrics were easily measurable (like numbers of members, numbers of connections, and the volume of posts exchanged), and which ones were truly worth measuring (like influence and trust) — and which ones you might have to settle for as a compromise until the right tools are developed (like network degree, centrality, and other social network analytics concepts).

More importantly, we covered the necessity of segmenting your audience when you look at those metrics (e.g., best customers v. new prospects, or “platinum” members v. “gold” or “silver”), as well as measuring trends over time (comparisons v. baselines, predictions, and prior time periods like last month or year). KPIs were mentioned briefly, such as number of members entering v. those leaving in a given month, or number of active members who post replies (answering questions) v. the total number of experienced members in the community.
Moderators or community managers have a specific role to play, from defining the purpose of the community (which part of the sales funnel do you wish to focus on: awareness, consideration, purchase, or customer support?), to setting policy in a fair way, to helping control flare-ups in a constructive manner rather than in one that destroys your hard-earned reputation or perceived integrity with one careless stroke.

How can you use metrics and analytics to predict which moderator interventions are necessary, and which ones likely aren’t? Especially if you run multiple separate social networks (e.g., multiple brands, or multiple country sites each with their own personality) without having to read hundreds of messages a day?

As this seems to be a practice area experiencing rapidly growing global demand, we at OX2 will begin to offer workshops covering these topics, and more, as we integrate this area with our existing Web Analytics practice.

Slides of the presentations will be made available for attendees later, but in the meantime, please don’t hesitate to contact us if you have questions about your own organization’s social network projects that we can help with, from blogs, to discussion boards, to a presence on Facebook, Myspace, or SecondLife.

%d bloggers like this: