LRW Resources

Back to Resources
Digital Analytics

For Influencer Scoring: Take AIM

Posted On  September 21, 2014

On September 9th, we published a blog, discussing important methodological issues when measuring social media influence. This first blog was inspired by Greenbook’s August 19 blog by Ray Poynter in which he named the Top 25 Social Media Influencers in Market Research as determined by triangulating the output from several different approaches to measuring social media influence.

Subsequently we received many requests to publish our own Top 25 list, and we obliged by posting a second blog post with said list. Again, we received a flurry of requests from industry insiders asking for the specifics of our methodology. So, we’ll dig a little deeper here.

LRW’s Advanced Influence Mapping (AIM) is done using software originally developed by the US intelligence analysts to understand terror networks. LRW’s AIM uses advanced math — Eigenvector centrality techniques– to identify the most important nodes of an interconnected set of networks to obtain a multi-dimensional picture of social relationships.  We identify and rank the leaders of these networks on their ability to connect guide a discussion or promote an idea across the network.  The algorithms at play facilitate:

Finding the “relevant” network from the complex web of Social Media.

  1. Data are analyzed in aggregate to identify individuals with posts and engagement related to the topic(s) analyzed. A Stage 1 network is drawn between these individuals. In the case of this month’s market research, social media influencer identification exercise, we used #MRX and “market research” as our starting topics.
  2. Importantly, the network is allowed to expand beyond those authors using given topic words or hashtags. The network is expanded up to 6 times beyond the initial set of network connections to identify those who are associated with part of the network who may be using given similar hashtags or specific language related to the initial topic search. In the case of market research, think of it as 6 degrees of Ray Poynter.
  3. By conducting detailed network analyses, relationships are built around followers, following, retweets and replies. For example, a person who followed authors who frequently use the #MRX hashtag was deemed part of the network even if they had not authored themselves. This was distinguished from those with only a tertiary connection, as indicated by their following only an account or two within the space.
  4. The AIM approach weeds out passing mentions of market research likely to fall on deaf ears (e.g., Justin Bieber speaking on #MRX to his large number of followers) versus people who resonate broadly among those deeply interested in marketing research.

Scoring influence within the network.

Influencer index score is a measure of 1) the structure of connections which ultimately determine where and in which direction influence flows, 2) an influencer’s position in and reach throughout their network, 3) contextual relevance (like Google page rank),  and 4) engagement with their audience, taking into account things like the number of tweets, retweets and favorites.  #MathRocks

The tools are proven. The US Central Intelligence Agency, using the same tools as LRW uses, identified an individual as a key terror influencer.  This individual had only one follower and tweeted infrequently.  When this influencer posted, the conversation across the broader network changed. Regular “influence” measures would not have identified this terrorist leader.

Lenny Murphy hit the nail on the head with the introduction to Mr. Poynter’s piece, “Social media will only continue to be an evolving disruptive force in marketing, so the more we can get a handle on how it works and can be harnessed for marketing impact, the better we can do our jobs.”  We agree. Ready. AIM.

Written by Joan Cassidy
Joan oversees a team responsible for developing LRW’s brand through a combination of digital, content, and event marketing. She is a catalyst for the company’s thought leadership initiatives, including her role as the editor of the LRW blog and producer for sponsored events such as the LRW Client Symposium. Joan also secures speaking opportunities, media coverage and awards for LRW thought leaders. Her expertise lies at the intersection of marketing, research and technology. Joan started her career at LRW working as a hands on researcher working with clients such as Disney, McKinsey. She went on to head up Global Research, then E-Commerce and Marketing for the U.S. division for Jafra Cosmetics International. She returned to LRW, where she led process improvement and software development and served as interim CIO. Joan has been key to the development on some of LRW’s biggest client relationships and engagements as an expert in branding, customer experience, tracking, and technology. She received her B.A. in economics from UCLA.


Your email address will not be published. Required fields are marked *

Subscribe to Our Perspectives