We’re often asked to describe how we determine our Social Share and Social Equity Index scores that drive our Social TV Ratings platform.
The answer is surprisingly simple. It’s best described by a genius with superior wordsmithing skills:
“The whole is greater than the sum of its parts.” ― Aristotle
It’s a thought we took to heart when developing the methodology behind our Social TV Ratings system.
The evolution of social measurement continues, but the stages of sophistication are significant.
Stage I: Effort
As with any emerging medium, success was simply being in the game. Successful social marketers were the folks who simply tried.Remember when the number of tweets was considered a legit indicator of someone’s Twitter skills? So, marketers focused on volume. The posts the better, right? Maaaaaybe. But if a tree falls and no one is there to hear it did it really fall?
Marketers needed a better KPI.
Stage II: Audience size
Marketers focused — rightfully so on building as big of an audience as possible. The bigger the audience, the better you are at social, right? Sure. But what if your followers aren’t really that interested in what you’re saying?
Still, we need a better KPI.
Stage III: Engagement
Ok. The more retweets, favorites, likes, shares, comments, the better. It shows that people are engaged and interested in your content. So, engagement is the ultimate metric, right? Sort of. What happens when “click holes” get in the way of duty? Consider the delicate dance performed by journalists in social. They’re tasked with growing audience by promoting “more viral” content while still serving the public interest as the fifth estate. Some of that content isn’t as clickable as Kim Kardashian breaking the internet.
So, we still need another metric.
It’s not that marketers are simple-minded dolts. Actually, it’s quite the opposite. All of these measures — and many more — are valid. But we need a good way to capture them in an easy to digest and actionable metric. One — or two if you will forgive us — score to rule them all. And that’s where we landed. Take those three major buckets — Effort (we call it Voice), Audience size, and Engagement — and blend them together.
Share Rocket’s Social TV Ratings platform provides broadcasters context about how they perform relative to their competition. Social Share measures your brand’s share of the total audience, voice and engagement in your defined market. This metric tells you how you’re doing over the course of the day, the week, the quarter. Social Equity Index is your long term performance score. This measures the social equity you have built over time and performance — relative to your peers. This tells you if you’re winning the war. We’ve written in detail about these before.
The details of our algorithm are proprietary. Insert Tom Cruise joke here. :) But here’s a high-level overview of how we combine those metrics.
How we combine the metrics
Indexing: The most important thing to understand about our methodology is that all of our scores are an index. The score a brand receives for their audience size is not simply the total of their audience. The brand receives a certain numbers of “points” for finishing first, second, third, fourth, etc. in a given metric category in order to normalize the scores for comparison with other categories (engagement vs. voice). It also allows us a more credible way of “combining” the metrics into one score that is meaningful. So your score is dependent upon how your brand performs in relation to peers and competitors in each area.
Weighting: Most indexes are weighted. We do the same. All of the key metrics are important, but some are more important than others. So, in the algorithm:
Engagement > Voice > Audience
Time decay: For our long term metric, SEI, we also have factors that weigh more heavily recent engagement. After all, how you performed three months ago is important, but it’s less important than how you are performing today.
Bonus metrics: Everyone likes bonus points, right? So do we. We like to reward brands that are “overperforming.” So, we’ve baked in a little “kicker” for those who are improving. Based on the historical data for each brand, we build a model that predicts it’s expected performance in each category (audience growth, posting frequency, engagement). If a brand outperforms that expected model in any of the categories, they receive the bonus.We’ve build a predictive custom model for each profile that we track and then compare actual social performance against that profiles expected performance in real-time. In short, we track how brands perform relative to their expected potential performance. We do not punish a brand for underperforming.
We collect and analyze millions of social engagements per hour each day. It’s all in an effort to better understand the whole social picture.