Ad Week Required Viewing: Text Mining Undresses Advertising

My colleague Max Kalehoff and I thought this would be an ideal Brand Association Map (BAM) to pull in anticipation of Advertising Week, which starts next Monday. We're going to use it as context for a webinar on Friday entitled "Nielsen BuzzMetric's Guide to Advertising Week."

The map is instructive, if not a bit unsettling, but first some context.  The BAM Map uses advanced text-mining algorithms to plot the most important language, attributes, issues and themes about a brand in concentric circles, derived from consumer expression in blogs, boards, ratings sites and other forms of consumer-generated media (CGM). With the brand or keyword represented in the bull’s eye, the resulting visualization is intended to help marketers to quickly grasp how brand identity, reputation and equity mesh with core assumptions. Not surprisingly, BAM outputs are often severely at odds with desired brand equities, benefit language, and beyond. But that's also why they are so cool and useful.

In this case, the terms “false”, “deceptive” and “misleading” -- all highly associated with conversation related to advertising -- are quite instructive.  They appear to reflect overt skepticism around advertising, and the data may well give pause for marketers to rethink the role of trust and authenticity -- even transparency -- in advertising.  It's also instructive to think about the positive signals, or the forms of advertising references (e.g. TV versus website).  Here are some of the types of questions a marketer might probe using the BAM map in this context:

• What’s the value/cost of consumer perceptions?
• Are these credits or debits? What triggers a credit? What triggers a debit?
• What’s the cost of messaging through the noise of distrust and disbelief?
• What’s the efficiency of riding momentum in the conversation?
• Is advertising the only solution to influence perceptions toward marketing?
• If the BAM were constructed with only advertiser or industry conversation, as an input would it look differently?

Sign up here for the webinar.  To round out our commentary, we've invited Erik Rabasca of Omnicom's PhD.

(Data Source: General Blogs September 1, 2006 – September 12, 2007)

Comments

[Max Kalehoff]

Hey, Pete! Great to see you today. Thanks for the attention. Have any other suggestions for brand association mapping?
Max

[Max Kalehoff]

Hey, Pete! Great to see you today. Thanks for the attention. Have any other suggestions for brand association mapping?
Max