This post originally appeared on Our Social Times. I will be discussing subject in their free webinar ‘How to Create an Integrated Social Business,’ on 7th May at 2pm BST (9am EDT) with Jon Bird of American Airlines and Philip Sheldrake of Euler Partners. Register for free here.

Somewhere between 87% and 98% of companies now have a presence on social media sites, depending on whose statistics you choose to believe. Whichever figure is more accurate, it’s fairly clear that if your company doesn’t use at least one of Facebook, Twitter, YouTube, Pinterest, etc., it is in an ever-decreasing minority. But just because you’re using social media for business, that doesn’t make you a “social business”.

Far fewer companies have applied the principles of social networking throughout their business; in many cases the Facebook and Twitter presence is just a Social Façade, a marketing layer that aims to disguise that in the rest of the company it’s the same old anti-social business as usual.

Of course, there’s absolutely nothing wrong with using social media as a marketing channel. But to see social solely for this purpose means missing out on much wider potential benefits. A report by McKinsey Global Institute in July 2012 claims that “while 72% of companies use social technologies in some way, very few are anywhere near to achieving the full potential benefit.”

To begin to take full advantage of social technologies, it’s important to recognise the three main types of social network, and understand how to engage with the right audience in the right place. Used together, social media, company-managed customer communities, and internal-facing employee social networks can form an integrated social business strategy that turns your company into a true “social business”. That sometimes sounds like rather vague and unattainable goal, so here are five pragmatic reasons why you need an integrated social business strategy.

1. Create more meaningful customer relationships
Public social networks are a great place for making contact with customers. With 200 million users on Twitter and a billion on Facebook, the reach of these services is immense so you would be crazy not to have a presence here. But they’re not great places to have more productive conversations with customers – there’s only some much depth you can go into in 140 characters.

A good example of this is Best Buy’s use of Twitter for pre-sales support. It provides fast, short answers, but needs to redirect more complex discussions to other channels. If you read through the stream of replies sent from @Twelpforce, typically every 4th or 5th response directs a customer towards a traditional email or phone CRM channel. Twelpforce is an effective, but very thin social layer. Integrating it into a Best Buy-managed customer community would enable deeper engagement, and more meaningful customer relationships.

2. Integrate Social and CRM for more consistent response
Customers of many companies have realised that if they complain loudly and publicly on social media they get a faster response. Indeed, some companies seem to be proud of their responsiveness on social media compared with traditional CRM, without thinking this through to the logical conclusion. Setting up a social media team as a rapid-response CRM team is clearly not sustainable – instead social and CRM need to coherently integrated, giving the same speed and quality of service whichever channel the customers uses.

For most companies, the level of integration between their Facebook page and their CRM system is very poor, so is it any wonder that irritated customers hijack the comments threads of the latest faux-cheerful marketing posts to complain? It’s perhaps a little unfair to single out any particular example of this when so many companies are guilty of it, but Three UK’s Facebook page provides as good an illustration as any you are likely to find.

3. Make your employees more efficient
While you’re establishing a more open, collaborative relationship with your customers, it seems rather unfair if your employees are still stuck with email and old-fashioned intranets as their main communication mechanisms. Unfair, and inefficient; the McKinsey report mentioned earlier estimates that use of social technologies inside the company can increase productivity of knowledge workers by 20-25% by reducing the time spent handling emails and searching for information. Indeed, McKinsey estimate that potential value of social inside the company is double that of the external value.

4. Learn how to be social
Employee social networks not only make the workforce more productive, they teach employees how to work in an online social environment. The list of social media disasters caused by inappropriate messages from employees grows ever-longer by the day, and while it’s easy to blame employee incompetence for this, the truth is that if you don’t regularly work in a online social environment, it can be easy to misjudge the tone or content of messages you send. Using a social network for communication with your colleagues gives invaluable experience that makes you a better communicator with customers.

5. Connect your supply chain
Perhaps the least explored area of social business is in connecting the company’s network of supplier and partner organizations. Very few businesses are entirely self-sufficient, so communication with other companies is essential. Yet business-to-business social networking is still in its infancy, with email still used as the lowest common denominator for communication. Establishing cross-company, private social networks can apply the productivity benefits noted by McKinsey to the wider supply chain.

Image credit: Networked Insights

The first part of this series described approaches for using social graphs to illustrate the way members of enterprise social networks comment on each others’ content. All the examples in part 1 used very small networks to describe these concepts. Let us now apply these to larger networks, and see how different graph layouts can highlight different aspects of the network.

The graph below shows comments between members of a 200-user network during one month. The graph layout is determined by the Fruchterman-Reingold algorithm, a force-directed algorithm which encourages closely related nodes to be plotted near each other. The effect of this is that the best-connected members of the network gravitate to the centre of the graph, and the least-connected to the edges.

This immediately highlights which members are engaging well, and which are completely disconnected from other members. However, as the network grows, this layout becomes increasingly poor at identifying clusters in the network which represent groups of members who are working closely together. The graph below shows exactly the same data, with a different layout algorithm which aims to identify these clusters.

Here we see one central cluster of usage, with several smaller clusters. However, the clusters remain fairly closely connected to each other, as one might expect in a small company. In larger companies, the clusters are typically more distinctly separated, as shown in the graph below.

It also becomes very obvious from a layout like this how some clusters are too heavily dependent on single members to hold them together.

Another interesting way to lay out the graph is based on geographical location. This highlights the communication across regions in a globally-distributed organization. The graph below plots all the members of the network from the first example based on their primary office location (centred on zero line of longitude).

Of course, the problem with such a layout is that it plots members at the same location on top of each other. This is a useful reminder that the most scientifically-correct graph is not necessarily the most useful. If we force members at the same location slightly apart from each other, we get a much clearer picture of inter-region communication.

What we lose in geographical accuracy, we gain in insight into network behaviour. Here we can quickly see three major centres of activity and three smaller, less active locations. We also see very strong cross-region links, suggesting the network has been successful in connecting a geographically-distributed workforce.

I am often asked which layout is best. The answer, of course, is that it depends on what you want to know. Social graphs can show a wide variety of different relationships – all of the examples covered here and in part 1 have focused on comments between members but it is also useful to visualize relationships such as members viewing other members’ content and members assigning tasks to other members. Members can also be grouped together into departments to see the connectivity between departments within an organisation.

One of the best ways of understanding precisely how your enterprise social network is being used is to visualize the activity using a social graph.  This two part series will look at ways presenting the activity and connections in your social network to illustrate where it is working well, and where further work is required.

The term social graph was popularized by Facebook’s use of it in 2007, but the principles of illustrating networks in this way is nothing new. The basic principle is that members of the network are represented by a nodes or vertices, and the connections between members are represented by edges. So the connections between 9 members of a network may look something like this:

This is useful, because it shows who is well connected (Molly) and who is less so (Ginny and Percy). But it doesn’t tell us much more than that, because it doesn’t describe the nature of the relationships between each member. Indeed, this is the major failing of LinkedIn’s InMaps, which although beautiful, are of relatively little use because my connection with someone I have worked with for 10 years is represented in exactly the same way as someone I met once. An enterprise social network has much richer information about the relationships between members, so let us consider how best to illustrate this.

Firstly, we need to decide the type of relationship we want to represent. Most people initially think of showing who follows who, but I find that this is little more useful than InMaps; following relationships often bear very little relation to how people actually communicate. I find that a far more meaningful relationship to show is the number of comments exchanged between members, as this more accurately represents the extent to which members are collaborating.

It is easy to extract this comment activity from a Clearvale network using the API. These can then be analysed in a variety of different tools – in the examples here, I am using Gephi. Detailed analysis of the data is something our consulting team helps customers with through our Social Engagement Analysis Service.

In the graph below, the weight of the lines between members indicate the number of comments they have given to and received from other members.

We immediately see the most active relationships in the network (Molly/George, Molly/Ron). We also see that some seemingly well-connected members (Bill) are actually weakly connected with several members, but not strongly connected with any.

Of course, not all relationships are equal, so it is valuable to distinguish between comments created and comments received. You may notice that in the previous graphs there are two lines between some nodes. This is because it is a directed graph, with incoming and outgoing indicated separately.

In the next graph, each node is coloured based on the number of comments created – green represents many comments, red represents few, yellow and orange somewhere in between. The edges are coloured to match the source node, making it easier to see the direction of the comments.

Here we see that although Molly is the centre of the network, the vast majority of her comments are directed towards Ron. We also see that it is George who is the most activer commenter, and Bill’s weak connection with the rest of the network becomes more apparent.

Finally, we can change the size of the nodes to represent the number of comments received.

Now, at a glance, we can see the rather more complex nature of relationships between network members. George creates a lot of comments but receives no more than Ginny or Percy. Molly receives a lot of comments, but creates few except on Ron’s content. We can use this deeper understanding of network member behaviour to help formulate our network adoption strategy.

In part two of this series, we will look at how these principles can be applied to much larger networks, and how different layout techniques can be used to highlight communication between  departments and regions within an organisation.