I just got back from a brief but intense conference with SAS. Followers of customer experience, social CRM, and other related topics will know SAS from … well, they might not know SAS very well at all. This is both understandable and a shame, but more on that in a moment.
On display at the the SAS Premiere Business Leadership conference were everything from customer stories of how high-powered analytical tools get them closer to their customers, all the way to computing clusters that eat terabytes of data in seconds and spit out detailed reports and predictive models in near real time. I got to watch a live demo of SAS Visual Analytics, a program that makes what used to take an advanced degree to create and turns it into a drag-and-drop tool with as much granularity and complexity as you want. It’s not just a dashboard, mind you—detailed reports with multiple variables and axes (the plural of axis, not the plural of chopping tool) can be built on the fly, several levels deep. The kicker? It’s all processed server-side so you can run it on a tablet as easily as on a PC.
Those servers aren’t necessarily budget-breakers either. SAS has been putting a lot of effort into working well in Hadoop clusters, inexpensive distributed computing environments operating under the Apache Hadoop open-source framework. There’s nothing to stop a company from running SAS on something like an Oracle grid (except maybe Oracle preferring you run Exalytics), but it doesn’t have to. Tremendous power is working its way into the hands of smaller and smaller businesses with each advancement; trickle-down theory doesn’t work in economics, but it works like crazy in computing.
SAS also talked bout products for information management, fraud detection, and hospitality/entertainment businesses (complete with their relevance to the average person), most of which are already in the wild or will be available by December.
An example of that relevance: Imagine a credit card company running multiple loyalty programs for cash back, discounts, etc., across multiple cards, each with a number of messages and contacts for customers. Customer X only wants a small number of contacts per day/week. Analytics can optimize and decide which messages to send for best results. It’s a very complex analysis that could take 4-6 hours. With the new system, running the same SAS interface, the company can do it in 2-4 minutes. Similarly, a retailer can manage pricing across thousands of SKUs and hundreds of shops on an individual level, rather than by region.
If you’re wondering at the amount of attention I’ve been devoting lately to high-end hardware, middleware, and enterprise software that the typical customer will never see or even know exists, let me explain. As much as I love the intimacy and immediacy of social CRM, customer experience, and all things SMB, it doesn’t happen ex nihilo. Businesses can track your preferences and history, make decisions about you as a customer, and provide compelling offers only because of advanced analytical engines running on powerful computing backbones. While it’s nice to imagine a team of elves personally interacting with you and evaluating your account, the fact remains that those elves are actually massive computing resources digesting millions of lines of database entries every second. Unstructured data—the kind that comes from social tools—is especially hard to manage, and the amount of it is growing at a rate that has storage vendors drooling.
Analytics, grid computing, in-memory operations—these are how the sausage is made, and the more advanced it gets the better the sausage tastes. Understanding what’s happening behind the CRM suite and the social platform makes me that much better at my job.
Besides, all that metal, silicon, and Big Data feeds a deep-seated geek desire in me for tech porn with MOAR POWAH! Petrol heads and overclockers know what I’m talking about. The possibility that all this face-melting power will make my life more science fiction than it already is sets me all a-quiver.