Social media has become an essential tool for connecting businesses with potential customers. In previous posts we have demonstrated that having an understanding of the underlying structure of a social graph facilitates better, more authentic communication with your fan base. As a new member of the Hearsay Social Data Team, I’m excited to apply my neuroscience background in brain networks to this endeavor. The goal here is to give you a quick lesson in network science in order to demonstrate the value that Hearsay Social brings to companies who aim to optimize the balance between global and local social media communication.
Networks are everywhere. From social networks to search engines, we are all familiar with technology that takes advantage of very basic properties of complex systems. But Facebook or Google did not invent networks; any complex system can be modeled as a network where nodes are the elements of the system and edges represent the interactions between them.
For example, LinkedIn has a snazzy tool for visualizing your professional network. In my network you can see clusters that separate into people associated with schools I’ve attended, cities where I’ve lived, and the company where I currently work. This type of organization is probably quite intuitive to most of us, but what may be more surprising is that this clustering and pattern of connectivity is prevalent throughout both natural and man-made networks.
It’s a small world after all
Imagine that you are trying to tell a secret to every person at a noisy restaurant (that’s a normal thing to do, right?). You could tell one person, then they tell their neighbor, and so on, which would take a long time but would eventually get the job done. Alternatively, you could randomly select people in the room in hopes that they’ll disseminate the information, but it’s unlikely to reach everyone. A solution that lies between these two extremes is one that takes advantage of the layout of the restaurant: go from table to table telling small groups of people. In this way you can retain the integrity of the message while still reaching a large number of people.
Biological and man-made systems alike evolve to optimize the balance between having a constrained amount of physical space (not everything can be connected to everything) and a need for rapid communication. In network science, this balance is quantified in a much-studied network property called ‘small-worldness’. Small-world networks have many localized connections supporting clusters of related processing coupled with relatively few connections between these clusters. These networks can take many forms, from the metabolic structure of bacteria, to flight patterns across the globe; they all share the same underlying small-world property.
How is the brain like a social network?
Since my background is in neuroscience, I’m accustomed to finding structure in seemingly random networks of brain activity. Coming from this perspective, the complicated interactions amongst fans on a Facebook page or professional contacts in a LinkedIn network is strikingly similar to the way different parts of the brain communicate.
Similar to a large social network with clusters of friend groups, the brain is made up of many smaller regions specialized for processing certain types of information. For example, the occipital lobe (green cluster) selectively processes visual input from the outside world. There are many sub-regions of the occipital lobe that are even further specialized to process more complex aspects of visual stimuli (i.e. motion, color, etc). Once the information has been processed locally through highly connected clusters of nodes (or neurons), the output can be sent through relatively long-range connections to other regions of the brain. If there is a decision to be made (throw a baseball), the frontal cortex (yellow cluster) incorporates this with other pre-processed information it has obtained from other local clusters in the brain and sends the appropriate output signals to motor cortex.
How can we use this understanding to take advantage of social media?
Social media has emerged as a natural way to connect to consumers on a more personal level than mass email marketing or other bulk communication techniques. By having a local business page, you are more likely to engage in a meaningful way with fans of your page and provide content that is specific to the users. However, having a single Facebook page, Twitter account, etc. will only allow for communication with a subset of all potential customers. This is the tough problem that Hearsay Social aims to solve: how can a large organization possibly manage the huge number of local pages required to reach the largest number of potential customers?
Above is a network visualization of an organization that has a single corporate page, hundreds of local pages and thousands of recently engaged fans. The small black circles represent individual fans that have engaged with any of the Facebook pages in the organization and the larger colored circles represent each Facebook page, the larger the circle, the larger the fan base. Thus, the large green circle in the middle is the corporate page for this organization, which has the most fans. The edges in this graph represent any engagement between the fan and the Facebook page (i.e. commented or liked a post). Each cluster in the graph has a different color and represents small communities of fans on a particular local Facebook page. As you can see, the highly clustered pattern of connectivity enables the organization as a whole to disseminate information and respond on a personal level with a massive number of individuals.
A critical lesson from the brain and other complex networks that share a small-world architecture is that the most efficient way to transfer information is to structure the organization into smaller groups that have many individual connections but are connected sparsely to each other. Hearsay Social is a platform that epitomizes this organization. Namely, by enabling local branches to connect with potential customers while still remaining compliant with the corporate brand, Hearsay Social, much like the human brain, becomes a complex infrastructure supporting quick and efficient communication.
- Guimera R, Amaral LA (2005). Functional cartography of complex metabolic networks. Nature. 433: 895-900.
- Wang et al (2010). Graph-based network analysis of resting-state functional MRI. Front Syst Neurosci. 4: 1-14.
- Newman & Girvan (2004). Finding and evaluating community structure in networks. Phys Rev. 69 (026113): 1-15.
- Vertes et al (2011). Topological isomorphisms of human brain and financial market networks. Front Syst Neurosci. 5 (75): 1-12.