Research of WOM and influence - a warning
Guy Kawasaki points to an important article published in the December issue of the Journal of Advertising Research (JAR). The article is entitled Reconsidering Models of Influence: The Relationship between Consumer Social Networks and Word-of-Mouth Effectiveness. It looks at how influence works in wom, and whether common preconceptions of influence are valid.
Spookily, I gave a talk at the London WOM Forum yesterday, which was organised by WARC, the publisher of JAR. Copies of the journal containing the article were handed out to delegates.
The research suggests that a pyramidal model of influence, where the majority of influence is concentrated amongst relatively few, highly connected people, is incorrect. Instead, the paper says that "it is the moderately connected majority, not the much smaller number of highly connected people, who hold the greatest potential for influence.
It also says that, "contrary to conventional wisdom that points to ... blogging as indicators of influence," more altruistic behaviour, like product rating, carries more weight. The research concludes that influence is not an exclusive, top-down model. Instead influence is something we all share.
This article is important because it identifies some key assumptions being made in the market about influence and the way it works. Some of these assumptions are identified and challenged by the article: others are not.
Here are the assumptions I detected in the article, and thus probably quite common in the market. My list may not be exhaustive…
Assumption 1. We know what an influencer is. The paper cites ' influence' (or derivatives) 52 times but nowhere does it define it. This is really important, as the scope of influence frames all further discussion. It’s difficult to recognise or measure influence if you don’t know what you’re looking for. At Influencer50 we restrict influence to those effects that impact on a purchase decision.
Assumption 2. Influence is concentrated amongst an elite highly connect few. In fact there is a non sequeter here - that connectedness is somehow related to influence. The article challenges the view that influence is restricted to a small group, but it reinforces the view that influence is somehow correlated with connectedness. This is unproven (and I have big doubts, based on Influencer50 research).
Assumption 3. Influence is unidirectional. In fact, influence flows in all directions. You can observe this the next time you have a conversation, debate or argument.
Assumption 4. Influence online is a proxy for influence in the real world. The article is based on research sources from web site users.
Assumption 5. WOM happens just between consumers. It clearly doesn't - wom, and influence, occur throughout the supply chain, and outside it too. This has huge implications for the structure and measurement of word of mouth campaigns. It’s much easier (and more effective) if such campaigns are grounded within the supply chain.
We have never thought that influence lies with a few people. But it does cluster and there are certainly those with more influence than others. But who are those people? The only way I know is to research the market thoroughly and with a rigorous methodology and rationale - anything else is just guessing.
Guy Kawasaki seems to equate 'elite' with celebrity. In the book we say that celebrity is very unlikely to influence buying decisions, and there’s plenty of research to back this up. In the markets Influencer50 tracks – predominantly B2B markets – the closest we come to including celebrities is when notable management gurus or authors exert influence on purchase decision-makers. People like Geoffrey Moore or (ironically) Guy Kawasaki.
It’s good that influence is being researched, but we do need to be careful about exactly what we’re researching.
Spookily, I gave a talk at the London WOM Forum yesterday, which was organised by WARC, the publisher of JAR. Copies of the journal containing the article were handed out to delegates.
The research suggests that a pyramidal model of influence, where the majority of influence is concentrated amongst relatively few, highly connected people, is incorrect. Instead, the paper says that "it is the moderately connected majority, not the much smaller number of highly connected people, who hold the greatest potential for influence.
It also says that, "contrary to conventional wisdom that points to ... blogging as indicators of influence," more altruistic behaviour, like product rating, carries more weight. The research concludes that influence is not an exclusive, top-down model. Instead influence is something we all share.
This article is important because it identifies some key assumptions being made in the market about influence and the way it works. Some of these assumptions are identified and challenged by the article: others are not.
Here are the assumptions I detected in the article, and thus probably quite common in the market. My list may not be exhaustive…
Assumption 1. We know what an influencer is. The paper cites ' influence' (or derivatives) 52 times but nowhere does it define it. This is really important, as the scope of influence frames all further discussion. It’s difficult to recognise or measure influence if you don’t know what you’re looking for. At Influencer50 we restrict influence to those effects that impact on a purchase decision.
Assumption 2. Influence is concentrated amongst an elite highly connect few. In fact there is a non sequeter here - that connectedness is somehow related to influence. The article challenges the view that influence is restricted to a small group, but it reinforces the view that influence is somehow correlated with connectedness. This is unproven (and I have big doubts, based on Influencer50 research).
Assumption 3. Influence is unidirectional. In fact, influence flows in all directions. You can observe this the next time you have a conversation, debate or argument.
Assumption 4. Influence online is a proxy for influence in the real world. The article is based on research sources from web site users.
Assumption 5. WOM happens just between consumers. It clearly doesn't - wom, and influence, occur throughout the supply chain, and outside it too. This has huge implications for the structure and measurement of word of mouth campaigns. It’s much easier (and more effective) if such campaigns are grounded within the supply chain.
We have never thought that influence lies with a few people. But it does cluster and there are certainly those with more influence than others. But who are those people? The only way I know is to research the market thoroughly and with a rigorous methodology and rationale - anything else is just guessing.
Guy Kawasaki seems to equate 'elite' with celebrity. In the book we say that celebrity is very unlikely to influence buying decisions, and there’s plenty of research to back this up. In the markets Influencer50 tracks – predominantly B2B markets – the closest we come to including celebrities is when notable management gurus or authors exert influence on purchase decision-makers. People like Geoffrey Moore or (ironically) Guy Kawasaki.
It’s good that influence is being researched, but we do need to be careful about exactly what we’re researching.
Labels: influencers, WOM
posted by Duncan Brown at 9:23 am
3 Comments:
Based on your comments, I'm wondering if you found any fuzziness in the JAR's assumptions about what distinguishes "wisdom of crowds" influence from "word of mouth" influence.
Interesting question - thanks. I think there are some similarities and differences in the JAR's approach to WOM and that of James Surowiecki, namely:
Both agree that diversity in a community is a good thing.
Both challenge the idea that influence cascades from ‘super-influencers’ to the masses.
However, the JAR assumes that WOM influencers are highly connected (but then challenges this assumption). Surowiecki examines the assumption that experts are the best decision makers. So they are looking at different things. Of course, influence may be a product of connectedness and expertise, but other factors too. For example, at Influencer50 we track (among other things) ‘Closeness to decision’ which measures the likelihood that the influencers can impact at the point of decision.
JAR is also looking at the transfer of information and/or judgement within a community. Surowiecki examines whether this transfer of information leads to an optimal result. In fact, although I loved Surowiecki’s book, and I agree with most of its conclusions, it’s clear that in real life, crowds are still influenced by experts, lobbyists, journalists, regulators, and the rest of the influencer community. Whether this leads to sub-optimal decisions is a point for discussion.
In fact, Surowiecki argues that when there’s too much communication (via WOM or otherwise) a collective mindset prevails and quality of decision making suffers. As he writes, “the more influence a group’s members exert on each other, and the more personal contact they have with each other, the ess likely it is that the group’s decisions will be wise ones.” Which kinda argues WOM in general!
Does this make sense?
Pondering.
Post a Comment
<< Home