• by Mike Petit , June 23, 2010

You would think that, given the technology available, and the mature metrics in place for almost all other marketing channels, our industry would already have a more defined metric around engagement. The truth is, deciding how to define and measure engagement is still a big problem facing online marketers today.

The proliferation of listening platforms is a direct result of the need to define consumers' interactions with brands. However, most listening platforms and buzz monitoring technologies are far behind the curve when it comes to measuring engagement. The mediocre, and often counterproductive, results speak for themselves.

Since these tools measure the amount of chatter on a given topic, they often define engagement according to frequency -- how often something is mentioned. However when human beings are engaged with a topic, they don't necessarily express it by simply repeating words. In fact, often the opposite is true. Take a blog comment like, "Tiger, Tiger, Tiger. All day and night, Tiger. Can't we move on to something truly important, like Finance Reform?" Traditional metrics would define this person as "highly engaged with Tiger Woods," when in fact, this person is clearly stating that he can't stand the constant Tiger chatter and is more engaged with finance reform (though he does not repeat the term "Finance Reform" more than once).

Another consequence of relying upon frequency and other simplistic, statistical measurements of engagement is that such techniques don't take account how people really express what they care about. When people care, they often express engagement by offering opinions, asking for or giving advice, or showing intent to do something. For example, "I really like the new healthcare bill and I'm going to write to my senator to show my support". Or, "I know you're thinking of buying a car. You should really consider a hybrid."

Brands who want to monitor and measure how people think want insight into the meaning of conversations that are happening and not solely how many conversations are happening. As an example, Kate Gosselin's recent stint on Dancing with the Stars sparked quite a bit of chatter across the social stream. If you're measuring on frequency and keywords alone you would think this meant that Kate Gosselin was one of the more popular contestants. However, it's not until you dive deeper into the conversations that one would realize the chatter was around her performance and people's desire for her to be voted off. Unless you look closely at the adjectives or descriptive language being used around a given topic, you're bound to come away with an incorrect read on the consumer's relationship to that topic.

We now have more sophisticated technology solutions at our fingertips that will allow us to measure engagement more accurately. I would offer that a better way to measure engagement is to focus on the elements of language that we use to communicate that engagement, among them polarity, requests for and offers of guidance, and expressions of intention and their associated decisiveness. Human communication is extremely complicated and unless your approach is to mimic the way humans think, anything else will ultimately yield compromised results.

As an industry thriving during the explosion of social media, we need to raise the expectations on how to best measure engagement. If we don't, the amount of data will only continue to grow and marketers will miss out on the opportunity to utilize that treasure trove of directly-expressed consumer attitude and intent. Without insight into what people are really saying we'll never be able to understand what people care about - and finding what people care about is the only way for marketers to determine what path people will take.