Why Measure Social Media

:
A Brief Point of View by Christopher Berry

Christopher Berry, VP Measurement Science, Syncapse Corp.

You can’t improve what you can’t measure. You won’t improve what you won’t measure. That second sentence is harsher, granted, but true. Measurement is essential to learn from past experiences and to improve the effectiveness of anything.

We learn through cause and effect.
In many organizations we are rewarded because of cause and effect. How can we ever hope to know the effect social media has unless we measure it? In sum, you won’t improve what you won’t measure, since the decision to do otherwise is a strategic decision that explicitly precludes real improvement over time. I’m incented to view that as an unacceptable decision, naturally. I do head up a Measurement Science group after all. Measurement is the act of gathering salient data. Science is how we go about learning and isolating cause and effect. Both measurement and science in social media are complex. We acknowledge that complexity. And yet, we’re confident we can march into it and emerge on the other side with an elegant model.

What is the source of this complexity?
Take what you know about the problems with Nielsen TV Ratings, search engine marketing, web analytics, direct marketing and data mining. Mash them together to get an idea of how complex social media measurement is.

Take, for instance, sentiment analysis.
We’ve been trained over the years to think of all customer buzz as being grouped into three groups: positive, neutral, and negative. I’m dissatisfied with the accuracy of such metrics. If you take 200 statements and ask 3 perfectly rational people to code them by hand, you’ll get 3 different versions of what they regard as an accurate count. Each person reported his or her own truth. It is very easy to dismiss such subjective metrics especially when they don’t mesh with your own version of truth. What real learning can come from that? The science part can be equally complex. Perhaps the relationship between business objectives and sentiment is non-linear. What if a certain degree ofnegative sentiment from a handful of detractors is necessary to cause the polarization around your brand? Is such polarization beneficial to the objective? If so, how much is necessary before it becomes detrimental? Every person, whether they’re aware of it or not, has a working mental model of how social media works with multiple channels - shaped by the organization they’re from and their personal experiences. Somebody from a strong TV background will perceive TV being the cause of social media buzz. Somebody from a strong data mining background will perceive social media as causing higher rates of retention. Somebody from a strong ecommerce background will always be unimpressed by the direct conversion single session conversion from Twitter. Somebody from a strong measurement science background will perceive social media as being all of the above: it’s just a matter of which model of the world will enable our partner to succeed and improve.

The names and logos for SocialTalk, SocialSync, SocialSupply and Syncapse are registered trademarks of Syncapse Corp. All text is copyright ©2009-2010 Syncapse Corp. All rights reserved. www.syncapse.com

A measurement strategy is required given the inherent complexity of an organizations goal structure and of social media itself. We must start with what is important to you and the organization. What is it that you’re trying to achieve? Reach? Acquisition? Awareness? Deeper engagement? Conversion? From there, we align an elegant model with that goal and measure what’s important to that goal. We use that understanding of cause and effect to improve the effectiveness of social media over time. Since social media is continuous, it lends itself particularly well to such improvements. My point of view is that the purpose of measurement isn’t simply to generate reports. It is to gain an understanding of salience, cause, and effect. It’s about generating answers to the question: knowing what we know, what is possible next?

Christopher Berry,
VP Measurement Science, Syncapse Corp

The names and logos for SocialTalk, SocialSync, SocialSupply and Syncapse are registered trademarks of Syncapse Corp. All text is copyright ©2009-2010 Syncapse Corp. All rights reserved. www.syncapse.com

Master your semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master your semester with Scribd & The New York Times

Cancel anytime.