Direct and relationship marketers (such as my company) make marketing decisions based on what we know about our customers and potential customers. We use demographic, behavioural attitudinal, and geographic data to better target our audience -- age, sex, education level, income, zip code, past purchases etc. According to a recent study by Shawndra Hill, Foster Provost of NYU's Stern School and Chris Volinsky of AT&T Labs Research there's another important variable that we should be considering: Who is connected to whom? Their research found that consumers are 3 to 5 times more likely to buy a company's product if they are "network neighbors" with existing customers.
Following an in-market test for a telecommunications company, the authors of the paper said: "We provide strong evidence that whether and how well a consumer is linked to existing customers is a powerful characteristic on which to base direct marketing decisions. Our results indicate that a firm can benefit from the use of social networks to predict the likelihood of purchasing."
The availability of information concerning links between consumers will allow companies to do "network targeting,". While traditional marketing research and data mining techniques can create detailed profiles (or personas) of potential customers, it doesn't reveal the social connections that exist between those consumers.
Now if we can just clear up the privacy questions….Download the full paper - "Network-Based Marketing: Identifying Likely Adopters via Consumer Networks".
Thursday, 25 January 2007
Subscribe to:
Post Comments (Atom)

1 comments:
In the Dynamics of Viral Marketing this and the investigation of propagation of recommendations is reviewed. Using a sample size of over 4 million people with over half a million recommendations for products in the popular DVD / Music arena being tracked, it provides insight into the fascinating world of word of mouth.
Within their studies they find recommendation chains do not grow very large ( we have a greater impact with our friends than others within the same community ) highlighting the long tail of products outside the top positions making up significant percentage of sales ( Amazon having between 20 – 40 % of sales falling outside of it’s top 100,000 books ).
Interestingly as more recommendations between two similar individuals the likely hood of these being actioned upon actually decreases, therefore exhibiting the characteristics of Spam.
Some figures from the report –
DVD’s reach a recommendation saturation point at 10 incoming referrals , with this figure changing as these two people start exchanging more recommendations the numbers drops to 5.
This reinforcing the findings of people can become a spammer if they send too many recommendations within the community group.
Figures differ for the ticket value and type of item ( fiction books VS technical manuals ) in terms of influence the recommendation has, therefore a person who recommends only a few products in the correct categories will be a respected peer within the community.
100 products amount for only 11.4% of all sales, top 1000 products amount for 27% of total sales through recommendation system.
Some products such as fiction books only require one recommendation, further recommendations have no influence and therefore become spam.
Providing excessive incentives for people to recommend products can weaken the credibility of community they are trying to strengthen.
One of the many key observations is around the value of the items - you will notice a fiction book recommendation the value is derived from time taken to read and therefore make a recommendation - a technical manual however would be on the value of your standing within your peer group as to the value of the information within the book
Post a Comment