Posts Tagged ‘network theory

changes in media (broadcasting to mass participation, towers to platforms, scarcity to abundance) are changing the whole cultural landscape. I would argue that this new landscape affects the context within which we act, even if our actions are entirely offline. It engenders a growing expectation of, and desire for, individual creativity

 Amy Twigger Holroyd

Simply, I have been trained to see the world in terms of what I can post to the Internet. I’ve learned to live and present a life that is “likeable” … The tail of Facebook documentation has come to wag the dog of lived experience

Nathan Jurgenson

“Reviews containing spelling and grammatical errors consistently result in suboptimal outcomes, like lower sales or lower response rates… the effect is very systematic.”

Panagiotis Ipeirotis

In a recent study of Amazon product reviews – Panagiotis Ipeirotis (Associate Professor of  New York University’s Stern School of Business) analysed how different language affects web users’ perceptions of products. He found that poor language and wording of reviews resulted in less sales of a product, but that the exact reason is not yet known.


This finding took me back to Levi and Kurzban’s theory of cascading benefits in social networks:

Essentially ‘cascading benefits’ works like this:

  • If I behave in a way beneficial to someone else, this will in turn benefit their connections
  • I unconsciously know this, which means I am more inclined to make friends with people who are friends with my friends
  • These ‘tertiary benefits’ I receive by benefiting my connections come to me more consistently if I am part of a dense network
  • The two factors above lead to clustering in social networks

A potential connection between the Amazon review data and the theory of cascading benefits seems to me to be this – Amazon reviewers look to readers like socially connected ‘friends’ of products. If these social connections have poor writing skills, even in delivering positive reviews, we may believe that buying the product brings us into the same social circle, or product/consumer cluster. We avoid this, as we fear the cluster, and product at its centre, to be socially undesirable – and that undesirable social effects will cascade towards us as a result.

The anthropological backdrop to ‘cascading benefits’ theory is that our innate instinct as species living in close-knit communities leads us to believe a favour done to a friend will eventually benefit us. Let’s reverse Levi and Kurzban’s theory and see if it fits- substituting ‘benefits’ for ‘contamination’:

So ‘cascading contamination’ works like this:

  • If I behave in a socially undesirable way which contaminates someone else by association, this will in turn contaminate their connections
  • I unconsciously know this, which means I am less inclined make friends with those I can see are being contaminated by social undesirables
  • The ‘tertiary damage’ I might receive by having a new, ‘contaminated’ social connection would come to me more consistently if I am part of a dense network
  • The factors above lead to my avoiding what I perceive (consciously or unconsciously) to be socially contaminated people or products – even given other indicators of high quality for them

Thus the way in which a product is talked about affects peoples’ attitude towards it – however positive the sentiment expressed.

This is worth considering in multiple social media marketing scenarios – for example running a Facebook page promotion bringing in floods of new fans who have a different socio-cultural use of language to existing fans / your target customer. In this case, securing hundreds of thousands of active (but off-putting) ‘fans’ is the last thing the brand needs to actually engender trust and purchases.

Facebook image from Pacific54

one of Twitter’s great benefits was that it made the relationship between two users asymmetrical… Your network can be pruned to be large, small, broadcast, narrowcast — precisely you can choose who to follow without finding yourself overwhelmed

Bobbie Johnson

The New York Times project cascade shows how networks online may be seen from the perspective of a news item or issue in relation to its temporality.

Brands, political campaigns, etc. can all benefit by being able to see how and where mentions and sharing propagates on Twitter in real time. As mentioned on the video below, “ultimately this tool may be useful to any entity that uses social media to initiate, conduct or encourage conversation.”

Visualisation of ‘cascades’ could indeed improve the navigability and usability of existing influence and network analysis tools such as those based on openAmplify. Not to mention make for more impressive moving image social media reporting for those needing to wow clients / stakeholders…

The ‘issue network’ is a fusion of online deliberation and network theory – highly valuable for the exploration of social, political and commercial influence.

A socio-technical theory, it draws from science and technology studies (Actor Network Theory – ANT) and critical media theory (Habermas’s public sphere, and Chantal Mouffe’s agonistic pluralism) – and has mainly been developed by Noortje Marres.

Her concept enables us to conceive how connections form online between various passive and active, powerful and less powerful political actors, who may or may not agree. (As opposed to a ‘social network’ which implies positive, friendly connections between actors).

‘Issue networks’ – then – are formed around subject areas – linking what is said online, (by individuals, groups and institutions) to influence and decision-making – meaning they can well underpin research for business, governmental or NGO purposes.

Marrying quantitative and qualitative approaches

An issue network approach is particularly valuable for the ease with which it enables the researcher or communications planner to marry qualitative and quantitative information.

Begin with a node. Where are the links to? How many are there and how much traffic flies between them? How much can be statistically gaged about influence from this? Then what is happening at each end? Qualitative methods, such as ethnography, can drill into the social dynamics of participation at each node.

Issue networks, traced and visualized

From an academic perspective, the issue network has been used to underpin research carried out by Marres and others. But its value for practice may be clearly seen by this influence map – developed by Salter Baxter. Although not apparently based at all on Marres’ issue network, it follows the same principles by revealing a network of actors online in relation to the issue of palm oil.

Greenpeace’s campaign against the use of unsustainable palm oil led to a damaged reputation and loss of business, reflected by a dip in its share price, for Sinar Mas. The influence map shows how specialist media, blogs and some mainstream media were connected to, and affected this issue. The size of the icons relates to volume of mentions and/or level of influence. According to Salter Baxter:

The web is a powerful tool to track progression of an issue and stakeholder mood. It can be used to identify where and when risks to brand, reputations and company operations are peaking. Clever dovetailing of sustainability strategy and communications strategy can then be deployed to address those risks and find opportunities.

Salter Baxter have developed their own tool (S:TAR) to track issues and produce influence maps such as the one above. A much simpler tool, which can be used free, and looks at connections identified via search engines, is the touchgraph SEO navigator. A far more complex network visualiser, which is again free, is the issue crawler.

This blog is about utilizing and optimizing the social web for business, pleasure and social change

Enter your email address to follow this blog and receive notifications of new posts by email.

Creative Commons License
This work is licenced under a Creative Commons Licence.

The views in this blog do not reflect that of my employer