At any given moment, vital conversations are taking place online. The modern consumer shares experiences and opinions on Twitter, blogs, and interest-based communities. These conversations are undeniably beginning to wield huge amounts of power.

Tap into this essential resource, the role of community manager has emerged as part of the social media value chain. This job requires listening to conversations across the web, tracking sentiment determining what people are saying and sharing, and channeling company responses to certain comments or questions.

In the beginning, community management typically entailed one person equipped with state of the art “listening tools.” Community managers diligently tracked keyword and brand mentions — and often ended up staring at an endless stream of updates. In many cases, brands receive thousands of mentions per day — far too many mentions for one person to follow. And so, a company hires another community manager, and another and the cycle continues.

While this is going on, conversations are amplifying on the social web at a breakneck pace. Over the next 30 minutes, approximately 22,000 blog posts will go live; 2.3 million tweets will fly; and Facebook users will post 2.77 million status updates plus 41 million pieces of content. Factor in the steady stream of online articles from traditional media outlets, and an untold number of niche sites and other networks.

Companies can try to keep pace with the sheer number of voices on the social web by adding more staff. Or, they can look for a smarter way to tackle the problem by giving community managers a way to identify which voices are the most important to their brand, market, and company. These “influencers” or “opinion leaders” are the individuals who carry most weight in any given market or topic area. They can impact particular communities, shape opinions, and even incite offline behavior from other consumers, like a boycott.

There are tools that measure a raw influence score, gauging the importance of the person behind a blog, tweet, or comment. However, here’s where it can get tricky: These scores are generic — meaning one score is assigned to an individual — and that’s not how influence works.

Influence scoring is only possible when considered in the context of topics. For example, Joe is a frequent blogger and contributor in the world of commodity futures. When it comes to talking about silver prices, he’s regarded as an expert. However, he rarely or never talks about cooking tools, gardening, etc. It’s clear that Joe’s level of influence is markedly different when discussing commodities trading vs. car engines. So, how’s it possible that a single influence score could represent Joe’s influence across countless markets and interest communities out there? It isn’t.