Getting a Handle on Social Influence
By Edwin Margulies
Social Engagement for Customer Care is one of the fastest-growing disciplines in the customer service space. There are all kinds of metrics we use to quantify the importance of a social post and social authors. Amongst them is topic relevancy and sentiment. But perhaps the least understood or leveraged is social influence. Let's take a closer look at influence and see what can be done to make it more useful.
What is Social Influence?
In the context of social engagement for customer care, social influence is an attempt to score or rate how influential an author is. Influence scoring is available via social feed aggregators and also companies such as Klout, for example. Numerous algorithms are used to calculate an influence score.
Some of the things that have a bearing on a social influence score are: a) the number of followers, connections, or "friends" a person has; b) the number of posts a person has; c) the frequency of posts; d) the number of views or hits on a blog or blog landing page; e) number of "likes" or "pins."
Curb Your Enthusiasm
Well, don't get too excited just yet. Yes, you can use social influence scores as a data point you need in making outreach decisions, but it's just a data point. First, let's take a look at some things you should consider:
1. Influence Gaming. Social networking users can "game" their influence by exploiting "junk follows." A junk follow is a throw-away follow based on marketers, scammers, and robots that stalk you just to get their own follows for commercial reasons. "I am following you now, so follow me back and I will show you how to get one million followers." By indiscriminately following these solicitations, which turn propagate other reciprocal follows, you can "game" or goose-up your influence ratings.
2. Domain Expertise. Most influence scoring algorithms do not take into account the author's actual domain expertise. That is to say that influence can be "watered down" if the author in question is posting on random subjects and is not recognized as a thought leader on a certain subject. This would include the celebrity quotient - wherein people have a lot of followers because they are a celebrity, not because they are an expert on a certain subject. Clearly, it is important to know if someone tweeting about your brand is a celebrity with a high influence score, but it is also important to consider the impact of their tweets. For example, if a tweet from a celebrity is not being seen by the constituency you are targeting - which may be an subject matter expert constituency, the celebrity tweet may not be as important.
3. Relevancy. The influence of an individual may not be so important if their mention of your brand, hashtag, etc. is out of context from a customer care perspective. If your name or brand is simply mentioned, and there is no sentiment or intent attached to it, it does not really matter how influential the poster is.
Is that it - Social Influence?
Not by a long shot. Consider the fact that in a customer care setting, influence goes beyond the number of followers someone has. It goes deeper. Your company has a relationship with this author. That makes the author a customer. You are probably tracking the loyalty, buying patterns, revenue spend, and history of that customer. Such is the stuff of a "Corporate Influence" score.
Here, you can use your own algorithm to determine how important a customer is to your enterprise. Modern social engagement for customer care systems, especially if they are CRM-integrated, can take advantage of Corporate Influence scores.
Imagine being able to see public influence and corporate influence scores side-by-side and sorted by author. This juxtaposition can help social care agents to make quick decisions as posts by that customers are made.
Better still, let's Automate that
Modern Social Engagement for Customer Care platforms will also allow you to define rules by which automated filtering and dispositioning can occur. For example, you might want to set up a rule that selects posts based on the following criteria:
1. Public Influence of over 20 points
2. Corporate Influence of over 15 points
3. "Not Happy" sentiment-wise
4. In the "Lemon Product" trending cluster
Now, you can set up next best actions or automatic dispositions based on these criteria. The list of how you can set up rules using influence is endless. Tweaking it for your enterprise can be done pretty easily based on the observations your social care team is making and also of course your established business rules.
Public social influence scoring can be an effective tool in social care outreach. But it is much more effective when combined with other author and customer attributes such as corporate influence, clusters, and sentiment. It's a good idea to explore ways to hook-in customer scoring based on your brand's relationship with the customer so you can go deeper and establish more impactful outreach.