Recently I’ve noticed an increase in the number of articles/posts predicting that 2010 will be the year of measurement. When marketers using social media finally cross the Rubicon of Results.
While I am hopeful that this is true – that we’ll start measuring what matters, instead of simply counting things – I’m not sure we’re really there yet.
Social media measurement should be about actions, not simply awareness, and to matter, it needs to be tied to the measurements that matter in the business: sales and net revenue. Monitoring sentiment and counting friends, fans and followers is a step along the way to true results measurement. But it’s not the end game.
That said, understanding sentiment is a good first step, and useful for analyzing the impact of social media engagement on revenue. A sentiment score accounts for volume, so you can compare it week to week, regardless of changes in volume. You can then see if there are any trends in lead volume or sales that track consistently over time with shifts in sentiment. If you have an initiative to increase positive sentiment, you can track your results. And so on.
To calculate a simple sentiment score for a given time period, usually week or month:
(Number of positive comments – number of negative comments)/total number of comments, including neutral and mixed) * 100
Net positive mentions divided by total number of comments, which accounts for volume differences week to week. Times 100 to put the answer on a 100-point scale.
The scale runs 100 to -100, with zero the neutral point. The closer to 100 on the positive side, the more positive the sentiment. On the negative side, the more negative.
You don’t need any fancy tools. A spreadsheet will do. All you need is a method for capturing social media mentions across the channels you wish to measure (blogs, Facebook, Twitter etc.), a standard way to assign sentiment to each message and the counts for each sentiment.
