At Altimeter, we’ve been measuring the state of social business for several years now.

Each year brings new shifts. Some are surprising, others not so much. This year’s theme, in “The 2015 State of Social Business,” by Ed Terpening, the shift is from scaling to integrating. That makes sense from a data perspective too.

I have to admit I had an ambivalent reaction to this particular chart, which shows the relative maturity of various aspects of social business today.

Screen Shot 2015-07-29 at 10.17.05 AM
Social engagement–the ability for an organization to interact digitally with communities at scale–is unsurprisingly first, at 72 percent maturity (those figures are self-reported, by the way.)

Coming up next is social analytics, with 63 percent of organizations surveyed reporting that their programs are mature. But when you unpack these numbers, it suggest a bit of a different narrative. All of these functions–from social selling to governance to employee recruitment–must be measurable, and must have relevant and credible KPIs to demonstrate performance. Do those exist in these organizations?

It’s hard to know, as these capabilities are themselves maturing, and social identity–the ability to match social posts with individuals–has only achieved maturity in about a quarter of the organizations we surveyed.

So what exactly are these social analytics measuring?

In my own work with brands and social technology companies, the answer is highly variable, but there are some consistent themes. Engagement is pretty measurable, but the outcome of engagement is much harder. Social customer service metrics in many organizations have matured to the point that social service levels aren’t too different from service levels in “traditional” channels. Event/sponsorship activation works when there are ways to attribute outcomes to those programs. But we still struggle mightily with the dimming effect of last-click attribution on the actual, meaningful outcomes we all want to see: revenue generation, cost reduction, time-to-hire, etc.

What this chart suggests to me is that we are still at a point when, even though analytics have supposedly matured, the actual criteria for business value (impact of social on sales, on activations, on recruitment, on acquisition, on churn) may still be cloudy for many companies.

The other point–not in scope for this research but a related theme–is the extent to which social data is being tied with other data streams. Anecdotally, I’m hearing far more evidence of this in 2015. Most companies I speak with are looking at social data in context of other business-critical data, but norming remains a challenge. And so social analytics tends to revert to the mean, which in this case means counting volumes rather than outcomes.

For the most part, I agree with what Forrester and others have said–that social analytics need to become more predictive. So I look at this data as a bit of context for brands, and a challenge for analytics companies to take up: we need to focus less on volumes and more on holding our own feet to the fire on what executives really care about: real business indicators and outcomes.