As influence becomes a coveted metric among those who measure and analyze social media data, many are beginning to segment content creators into one of two groups: those with influence and those without. But considering that our ability to measure influence with true precision is still relatively limited, this may not really be the most effective to way to think about influence. Rather we may be better served by breaking influence into its component parts: reach and relevance, and then examining the individual business purpose for wanting to know influence.
Reach is about numbers. How many fans and followers does an individual have across all of his or her online properties? This is often the easiest way to measure influence, because it is objective and public metric. Some will say that this metric doesn’t always tell the right story since “tribes” are influenced by very specific people. However, it can still be a relevant metric, especially when we think about social customer service within retail and consumer products/services. It doesn’t matter if Justin Bieber has zero relevance to the pizza business and is not an influencer to that industry. If he says that he hates Dominos, that’s going to have an impact on the brand, and customer service professionals with access to social monitoring and social CRM tools need to pay attention to the absolute reach that each person has online.
Relevance is about topics and interest graphs. This is a little harder to measure, but there are already a number of vendor tools that are trying to crack that nut as well. We’re starting to see Klout scores incorporated into many of the leading social media monitoring tools, and a combination of social mining with topic level influence is going to make this type of analysis important for brands trying to influence the influencer. Unlike reach metrics around influence which are essential for real-time response and crisis avoidance, relevance metrics can help a brand quickly figure out who is influential to their particular brand or industry and then target that person for outreach. A food blogger who could potentially rave about the new Dominos pizza sauce is a much better candidate to engage with than a B-list celebrity who never talks about food but technically may have more fans and followers.
Linking reach and relevance clearly allows brands to have a better understanding of influence for customer service, outreach and trending analysis. But because the methodology of determining accurate relevance is still in its infancy, it may be a bit premature for brands to automatically combine reach with relevance. Rather, we encourage organizations to look at both reach and relevance individually, and to bring together reach metrics with relevance data and interest graphs when appropriate.
To illustrate the danger of combining these metrics: a food blogger who talks about pizza with 5000 followers is obviously a better candidate for outreach than a food blogger with 500 followers. But the reality is that we don’t know if we have captured all of the individuals who are relevant to Dominos by focusing just on food bloggers who write content about pizza. Data analysis of social conversations is still generally keyword based, and there can be large gaps in the types of relevant influencers that automated tools bring back.
We recommend using both methods of measuring influence, but using them separately and as appropriate. Reach for customer service. Relevance for influencer engagement and trending. It’s not always black and white, but it can help focus analysis to those metrics that will have the most impact for your specific business goals.