Software review The role of social networks in marketing

Abstract   Social network analysis is not new, but its business application in marketing is a relatively new area. This paper describes what social network analysis is and how it     is being applied to solving marketing problems around segmentation, targeting and campaign design. In particular it describes how the social network can be defined, the   role of the influencer and how this information can be used to improve marketing insight and communication effectiveness.

Journal of Database Marketing & Customer Strategy Management (2007) 15, 60–64. doi:10.1057/palgrave.dbm.3250070


In order to create marketing strategies suitable for our customers, we must first understand what the customer is influenced by during the decision-making process. This might be the most important way to learn how and where to correctly invest our marketing efforts. Recent research shows that more than 75 per cent of customers will consult a friend before deciding on the purchase of a certain product or service. But the main issue here is whether organisations know how to utilise this fact to their advantage.

In recent years, it has become evident that large organisations are beginning to appreciate the importance of word-of mouth marketing. We are still, however, nowhere near effectively utilising this information resource.

This paper describes what social networks are, what the best way of creating such  networks is, and how an organisation can utilise these networks in order to create effi cient marketing strategies for its customer base.


We will examine what would make a customer feel confi dent enough to purchase a certain product according to a survey conducted by eMarketer, in which each participant could choose multiple answers

  • A friend ’ s recommendation (76 per cent)
  • Previous experiences you had with this company (68 per cent)
  • A recommendation in a newspaper / magazine (22 per cent)
  • Advertisements (15 per cent)
  • The company ’ s website (8 per cent).

In other words, most of our customers will consult a friend prior to making a decision about a certain purchase.

Screen Shot 2016-03-01 at 10.38.15 PM

This type of promotion is called word- of-mouth marketing, and can take place between any two or more connected people, that is, via a social network.

Conversely, companies invest millions of pounds annually in an attempt to market their products, although most of them neglect to consider the influence of word-of-mouth. Moreover, even the companies who have already taken notice of this matter are usually doing so based on a gut feeling, rather than a statistically based analysis.

A good example is current word-of- mouth marketing tactics, such as viral campaigns. As the organisation does not, however, properly map the social network itself, it is very difficult for it to track and measure the results of such campaigns, and recognise its successes and its problems.

Consequently, as an initial step in the road to an optimal solution, we should map the social network accurately. But we must first understand what a social network is.


A social network is a collection of interconnected people.

Social networks comprise of points (people and potential customers) and connections between those points. These connections may be manifested in many different forms. Examples include

  • E-mail exchange
  • SMS exchange
  • Purchases
  • Telephone

Figure 1 illustrates how a social network is formed.

Each of us has a personal contacts list.

For instance, if we examine e-mail exchange, each e-mail I send will create a connection between me and the recipient   of that e-mail. That recipient can, in turn, forward that e-mail to his contact list, thus creating another connection between him and his recipients. Consequently, a network of personal connections is created or in its official title, a social network.


Now we know what a social network is. So what is the next step?

It is important to understand that the   first step towards a solution is our ability to identify the existence of a social network within our potential customer base.

Once we have identified the social network, we can move on to the next stage.

Identifying a social network

This is quite a tall order, but no longer an impossible one. There are quite a few technological tools developed for the sole purpose of efficiently and quickly identifying social networks, without having to invest any additional resources.

So now that we have identified the social network, what is the next step?

The second step is isolating those network members worth investing our marketing efforts in. In other words, out of the potential customer base, we need to determine who the opinion leaders are.

Identifying opinion leaders

Opinion leaders are network members regarded as having relevant knowledge, and who are probably the first ones to be consulted in regards to purchasing decisions.

Usually, most opinion leaders possess one or more of the following characteristics:

  • Part of a social network
  • Good communicators
  • Usually early adopters of products or services
  • Information

There are different technological tools that can help identify the opinion leaders among our customers.

Now that we have identified the opinion leaders and their connections within the social network, we can divert all of our marketing efforts to focus on those specific customers, assuming that they, in turn,

will spread the word to other network members. This way, we can reduce marketing costs and refocus our resources more effectively.


Word-of-mouth marketing is no novelty.

It is actually one of the earliest forms of marketing, going back as early as biblical times, when Eve suggested that Adam taste the apple, because it was very sweet.

Nowadays, there are a number of ways in which we can utilise word-of-mouth to effectively meet our marketing objectives:

  • We can use technological innovation to effectively detect social networks and opinion leaders.
  • It is a well-known fact in our world that the customer is in control, deciding for himself what the right product is, and when and how to buy it.Therefore, traditional marketing no longer suffices for answering our customers’ constantly changing
  • The rapidly evolving world of internet created a whole new game plan, for example, online forum debates, blogs, etc, which, in turn, produce new forms of word-of-mouth marketing.


Once we fully understand the social networks surrounding us and learn to identify the opinion leaders within those networks, we will be able to establish suitable marketing strategies that will spontaneously produce word-of-mouth marketing.

Additionally, we will also be able to allocate our financial resources towards strengthening connections with opinion leaders and recruit them as advocates for our business.


This case study is based on a fashion retailer in a European Market.


The client provided the analysis team with data from the loyalty scheme on 500,000 customers.

These data included:

  • Customer name and address
  • Date they joined the loyalty scheme
  • Summary purchase behaviour
  • Primary branch details
  • Purchase history at a product level
  • Coupon and voucher redemption data
  • Attendance at special events
  • Marketing contact history
  • Response to previous communication and promotional activities.

The data set covered

  • All customer included in the loyalty scheme both active and dormant
  • Transactional data since the scheme was introduced (nine years worth of history).

The team merged this customer data with relevant reference data on

  • Product and product groupings
  • Stores and store hierarchy
  • Branch and customer

The data were then refined and data quality problems removed.

The data were then processed through a social network analysis solution that was able to cope with the data volumes.

The basic stages were as follows

  • Extract social networks
  • Calibrate the connections between the parties in the network
  • Measure the flow of information/ behaviour through the network
  • Use the social network parameters to drive predictive analytics (eg churn).

The key output from the social network analysis was a set of attributes describing social effects for each individual. These included:

  • Total number of parties in the friends
  • Number of friends a customer has who are at risk of churn
  • Number of friends a customer has who are very loyal
  • How many opinion leaders in the customer network are influencing him/her to churn
  • How much cumulative influence (depending on connection quantity and strength) is the customer under to churn
  • How many of the customer’s friends have joined/left the network over any time period.

The following example illustrates the power of the social network analysis

  • A customer (Lucy) started shopping on her own in 1998
  • Within eight months she was shopping on a regular basis
  • Over the next 18 months, we identify eight other people who become part of her social network
  • The purchase behaviour of the network shows that the initial customer (Lucy) is the opinion leader
  • Three people in the network have the same second name
  • Over the next few years, we see a consistent purchase behaviour pattern across the customers (Lucy) network
  • The opinion leader stops purchasing
  • Within three months all other members of the social network stop

In order to understand what triggered the change in purchase behaviour of Lucy and the social network, a number of the members were contacted (including Lucy) and their purchase behaviour discussed.

It appeared that Lucy had purchased a blouse that had lost its colour in the wash. When she took the blouse to the store, the   sales assistant badly handled the situation and refused to make a refund. Ever more damaging she claimed with it was not defective and that Lucy had incorrectly washed the item.

Lucy has been so upset that she told all of her friends including those in the social network not to purchase at the chain any more.

As a consequence of this and other examples the retailer changed its refund policy.

This and other analysis show that opinion leaders can represent 7–20 per cent of the total customer population.

The retailer then went on to develop a range of marketing communication that focused on either the opinion leaders or the network as a group. These have proven to be very successful.


Social network analysis, although well proven in other disciplines is only starting to be applied with rigour to solve marketing problems. The initial results are proving to be valuable. As we see a growth in the use of this approach, I have no doubt that is will see the emergence of new marketing disciplines that focus on marketing to the social network and the influencers.