As you might have seen from my recent tweetfest, I’m in Seattle at the SAS Global Forum. The reason, other than my need for frequent-flyer miles, is to learn about the analytics company’s new Social Media Analytics product.
“SAS invited me to their SAS Global Forum user event as their guest to attend the launch of SAS Social Media Analytics. They paid my airfare, hotel and conference registration fees and gave me access to the product for evaluation.” [Their words, but I accept and endorse them.] In other words, this.
I have said previously that a company that develops a truly effective social media analytics package that includes sentiment and modeling in depth, and can tie it into CRM, has essentially created a license to print money in today’s social CRM-focused world. I haven’t seen enough of SAS Social Media Analytics (SMA) to say if it achieves this, but the demos put me in a favorable frame of mind. Analytics (has? have?) come to my social media world, and this is a Good Thing.
SMA is more than a dashboard or reporting engine. It gives the user live interactive access to conversations about the brand. The view is not static, but can be tracked over time, against multiple sentiment components. The data models are subject to updates and new instructions, so what you capture can be sliced and re-sliced as needed. This human angle—user input refining the model—is a big deal to me. It prevents SMA from being a black box.
SMA is a slightly misleading name, in my opinion. It’s media analytics, which includes social media. I’m not faulting them on the name, mind you; social media are harder to track because each piece evolves with use. One could argue, though, that all media today are social media, since everything that’s published seems to end up on the Web with comments and links.
SMA doesn’t come cheap. While SAS is describing SMA as an “on-demand” application, there is an initial investment in data gathering and modeling, and a fee of $5,000 to $15,000 per month. I’ve overheard SMA described as “an enterprise-class Radian6,” and that’s probably a fair estimate. Radian6 appears to be more focused on engagement (which is VERY important) while SAS is playing to its strength in analysis, but both companies have capabilities that mirror the other. The way I see it, if you can afford to spend SAS money and get value from that expenditure, you probably should migrate from Radian6. It’s not just a question of money, though; I’m sure there are some massive businesses that need exactly what Radian6 provides, no more and no less. SAS has a reputation for brute-force analytics power (emphasized with last night’s demo of a multiple-terabyte process run in two minutes), and that’s got to be worth the price tag for a lot of businesses as well.
There are some things that still need to be answered for me, hopefully with an in-depth demonstration. For one, I don’t know how quickly SMA responds to new rules and model parameters. Would I need to back away from the workspace to change keywords and sources, then start over? Or can I play fast and loose, tweaking the factors as I go?
For another, almost everything we’ve seen today is about internal analysis of what happening in the socialverse. There hasn’t been much emphasis on the engagement portion, or on closing the loop and reiterating the feedback process. It looks like the customer is still “out there,” rather than at the core of the business process. To be fair, this is an analytics product, so I shouldn’t expect something else. Still, some more examples of how SMA can have an effect over time on the customer sentiment it monitors would not go amiss. My interest is social CRM, not merely social media—the customer and the opinion-maker need to be right up front. Capturing the voice of the customer is good, but listening to it and then capturing the ear of the customer with your response is better.
Overall, though, my first impression is that SASSMA is a promising product that arrives at the right time. I’ll be keeping my eye on this and providing you with updates as needed.