This is the second post in a series dedicated to explaining what social networks are, how they work, and why interests and relationships mapping are useful for brands and digital marketers.
Marketers who analyse online conversations are usually interested in consumer behavior. In particular, they usually wonder things like “why has this specific user been mentioned?”,“Why did the user mention the brand here?” and “Why have these users, interested in this specific topic (food, sport, fashion, movie, automotive…), decided to interact on this medium?”.
We at ROIALTY have observed that the individual attributes of the actors involved in the social interactions (gender, age, education, social/economic status) are significant elements in defining a social network, therefore they have to be monitored in order to achieve a complete mapping of the social network.
Therefore, analysing a brand, a product or a particular topic on social networks means focusing on social interactions and the relationships among the users that are interacting with these media.
Understanding who are the most “important” (this adjective has a broader meaning here, according to the object of the analysis) users in a conversation will disclose to marketers a wide array of information (target, copy, format) that can be used in their communication strategies. This is the objective of Centrality Metrics: indicators identifying the “vertices” in a network.
The value of this kind of investigation is clearer if we give a practical example.
A common analysis among marketers is the “trending topic”, i.e. what is the most discussed topic by a target audience. This action should also include the semantic analysis of the online conversations, in other words, marketers should also know whether their target audience is expressing positive, negative or neutral opinions about the object of the conversation. This analysis is often made during live tweeting or social media events in which centrality metrics are used to understand who the most important network users are.
But this investigation gives just a superficial understanding of what is happening around a topic: if a number of users with a few links in the network have a good opinion about a product, but another node, well linked to the entirety of the network, expresses a negative opinion about it, you will not be able to forecast what is going to happen next…
How can this bias be avoided? A simple and effective measure of an actor’s centrality and power potential is his degree. The actor degree refers to the “nodes’ numbers” linked to it by a relation or tie. Degrees can be input (“in-degree”), in case of interactions received by the node, and output (“out-degree”) when the interaction originates from the node. Using this analysis, marketers will understand who are the most linked users in the target network.
Being well-linked is the first step towards being influential: if this is true in real life, why shouldn’t it be so online? This is the field of community detection.
Detecting communities deepens the understanding of a community towards a topic or brand, detecting in which way an actor/node has more influence in his own network. The influence is mapped according to the interactions in-degree (according to the language of Social Network Analysis, this is the number of interactions that a node receives via input). Again, a practical example is helpful: if the ROIALTY Twitter profile receives more re-tweets, mentions or replies than another account with the same number of links, this means that the influence of ROIALTY in the social interactions is stronger: this is the ABC of influencer marketing.
Detecting who the influencers are in a network and what matters to them (in terms of interests and behaviours) dramatically enhances the likelihood that a brand will be included in a conversation, as they will be able to tailor targeting, content and calls to action for the specific “vertices” of its audiences.
Other sophisticated metrics are “closeness centrality” and “betweeness centrality”. These two metrics synthesize the characteristics of the path that interactions follow from the node of origin to the node of destination.
Closeness centrality describes the extent of influence of a specific node on the network, while betweeness centrality emphasises how many individuals choose to (or have to) go through a specific node (called a “bridge”) in order to reach another node, in an attempt to minimize the number of steps. These metrics are typically used to monitor how an actor controls a particular piece of information: we are talking about SNA (Social Network Analysis).
These are the main centrality metrics
ROIALTY customer segmentation platform has an efficient annotation system categorizing influence, interests and interactions and allowing a thorough profiling of user clusters.
In this way, brands can personalise their digital marketing activities according to real influencers’ characteristics and social behaviour. Achieving brand awareness, customer loyalty, brand reputation and lead generation is simpler and quicker: a value added to your digital marketing strategy.