How people's attention flows on the web
30 March 2013
We didn’t see any statistically significant relationship between our buzz and our short-term sales.
The above is a quote from Eric Schimdt of Coca-Cola at a conference last week. From my own experience this is a completely correct assessment of the situation around how brands are mentioned on the web.
Unfortunately as Coca-Cola has by now no doubt spent significant sums on investing in ‘social media’ and is one of the ‘most popular’ brands on various social platforms this is tantamount to heresy. Hence Wendy Clark, also of Coca-Cola, playing the ‘social is part of an integrated approach’ card by claiming “we’ve been able to track closed-loop sales from site exposure to in-store purchase”.
This recalls the McKinsey consumer decision journey where apparently awareness, familiarity, consideration and purchase all lead to loyalty. Of course no data is given to support any of this.
What this shows is how large businesses still lack an understanding of how people talk about things on the web. In this context ‘buzz’ implies excited discussion that is relevant in some way to the brand. It is no such thing. ‘Buzz’ is simply a keyword-matched mention that could be anything and is most likely being used in a context of very little relevance to a brand’s marketing activity.
‘In one 2010 study where Coke pulled out more than 1,000 social-media messages randomly and had human raters compare them to automated sentiment analysis by one vendor, there were widespread differences. “When we say it’s positive, the machine about 21% of the time says it’s negative,” he said. “That can cause some problems in our understanding” of how buzz impacts sales.’
Expecting an algorithm to give you context is foolish. It does not work. Inaccurately counting mentions as positive or negative tells you nothing. ‘Buzz’ cannot be directly linked to sales. Take out the spam and what you have is a background hum. If you are one of the most heavily marketed brands on the planet then that amounts to a severe case of tinnitus.