At this week’s Social Media World Forum Europe I joined a panel to debate the merits of measuring engagement in social media activities. It was a friendly panel, with few areas of harsh disagreement, and we all agreed that we’ve not yet cracked the search for meaningful social media ROI metrics.
Here’s my take on things. All marketers are on a journey to identify metrics that can truly prove if social media marketing programmes can work or not. I think this journey looks a bit like this [click to enlarge]:
In the early days of social networks, crude measurements like ‘Fans’ and ‘Followers’ ruled the roost in the land grab to establish a brand’s presence on the social web. But we quickly realised that raw follower numbers alone are meaningless as they can mask a multitude of problems like the relative value of different types of followers.
So we graduated to measuring engagement, looking to indicators of actual involvement with our brand to assess the value of our social media marketing efforts. But doing this just shone a light on another problem: engagement is not the end, it’s just a means to the end. Just because we can measure when someone ‘Likes’ something our brand does online, it doesn’t mean they will become a source of revenue. Let’s not forget, this isn’t a new phenomenon; TV ads have had ‘fans’ for generations but we had no means of counting them yet our marketing was none the worse for this.
And so now we must enter a more enlightened stage in our journey. Just as we’ve done with every emerging media format in the past, we need to combine data sources from multiple touch-points and adopt sophisticated econometric modelling approaches to help us make sense of the data.
It’s this stage in the journey that gets me really excited. The social web affords us more rich, real-time consumer data than we’ve ever dreamed possible. Collecting this data is simple and cheap. Combining these data, however, with other indicators of brand health like propensity to purchase research, customer feedback, price elasticity models and so on is extremely difficult.
But modelling of this kind will soon become far easier and affordable for many brands as the tools we use to monitor online behaviours evolve to connect more seamlessly with other sources of data within our businesses. Our ability to derive value from these rich data sets, however, will still hinge on the skills of talented researchers and data miners who can manipulate the data to unlock the consumer insights hidden within.
This is an exciting time to be a marketer if you have an appetite for learning and the good instinct to become a great data miner.
Thanks to everyone at SMWF for their feedback on the debate. It was fun!