Jim Goodnight founded SAS with three college friends in 1976, and has run the company as CEO, and has run it as CEO ever since. He’s earned the right to a vision, which he described in his opening address to the 2019 SAS Global Forum as using analytics to change the world.
It quickly became clear, in the general and breakout sessions which followed, that two imperatives are implied by the vision. FIrst, to change torrents of blind data into actionable insights — “Transform a world of data into a world of intelligence,” as COO/CTO Oliver Schabenberger put it — and second, to make those insights invisible beyond the rarified world of data science.
Schabenberger’s message was clear. Digital transformation has to extend beyond the data lab; it needs to create value for an organization; and organizations need to see tangible, short-term results. The road map from vast data lakes to practical outcomes has a clear sign pointing the way: “automation.” And automation at scale, of course, has to be supported by AI. “We’re not adding to the buzz about AI,” said Schabenberger. “We’re making it real.”
That’s a claim which is not entirely unfamiliar, and in due course we’ll see why. But first I wanted to mention a disarmingly simple demonstration, presented on the main stage, which elegantly summarized where SAS is going. We’re all familiar with recommendation engines, but can any of us build one? (If you answered yes, because you have Python and a dozen other coding languages at your finger-tips, this was the conference for you.)
Seriously, using the SAS platform, it seems any one of us can put an elementary recommendation engine together — in about two minutes. But the eye-opener was the possibility, within the platform, of toggling between full-on-coding, low code, and practically no code. Data mavens can hit the knitting patterns, and customize the engine to their hearts’ content. More competent users can piece together snippets of existing code, and make changes where necessary. People like me can just drag and drop the basic elements. Accessibility, then, combined with the flexibility to get much more complex. Plus visibility: the ability for the white coats to drill deeper and understand the reasons the automation (the AI) surfaces particular options.
That other imperative though — digital transformation to change the world (which, as Schabenberger pointed out, is essential made of bits and bytes these days — means that SAS has a very broad range of analytics offerings. We saw breathtaking 3D visualizations of tumors revealed by CAT scans. We were moved by the use of analytics by Hanover County to implement a continuous risk assessment for child protection services. The list went on: financial institutions, human resources; as I said, bits and bytes everywhere.
Which led me to ask, what about the marketers? The main reason I was here was the persistent recognition of SAS, not just as a leader in data and analytics, but as major player in the marketing tech space. Perhaps it’s no surprise that SAS (and Pega) are leagues ahead of Salesforce and Adobe when it comes to the specific skill of Real Time Interaction Management (Forrester). But it’s also ahead of the pack (along with Adobe) when it comes to customer analytics, thanks to the SAS Customer Intelligence 360 offering (Forrester), making it a prime-time CX player.
And when you look at Gartner’s current, April 2019, assessment of the the leaders among multi-channel marketing hubs, you find SAS (best ability to execute) sitting alongside Oracle, Adobe, Salesforce, SAP, and Marketo. Be honest, when someone asks you to name the first three marketing hubs you can think of, do you name SAS?
All of which is a long way of saying that I had some questions for SAS CMO, Randy Guard. But first…
Announcements from SASGF
At Global Forum, SAS released a range of announcements traversing the broad horizon of their business, from compliance projects with Citi, to demand and supply chain planning with Nestlé. Here are the ones you need to know about.
Note: The ubiquity of SAS systems of classroom tools was reflected by the heavy presence of academic institutions, alongside tech and data vendors, on the expo floor.
I wanted to ask Randy Guard about the accessibility of analytics, both about the concept of AI working under the hood to support any kind of user, and about what seemed to me to be a low code to (almost) no code approach. He told me that embedded AI was key to the current SAS offering. What, I asked, differentiated it from Salesforce Einstein or Adobe Sensei?
“The combination of two things, I believe, is a differentiator. The open platform: we have capabilities across a lifecycle of data, discovery, and deployment. You can bring in the data, do an analysis of it —
including visuals —
and where do you want to put it? That, combined with transparency and management of the analytics environment, that’s different, that’s where we have uniqueness.”
And the opportunities for business users to be hands-on with analytics tools?
“It’s very deliberate,” Guard said. “It’s very much an intentional strategy. We want to be very visible and vocal about it because there’s a gamut of users who can benefit from analytics.”
He cited the main stage example of building a recommendation engine. “You’ve got to know a little bit about what you’re doing, what you’re trying to solve, and to build the model took maybe a minute and a half or two minutes to run through several cycles of data prep, and then we’d built a pipeline. You saw probably five or six models of different types, and then we created the ensemble model, and published the API so you could call it up from wherever. If you want to go in and change the model; if you’re not skilled at that level, you aren’t going to know how many tweaks to make, how many layers of the neural network you want, but you can get it very good, very functional through automation.” It’s automated, Guard reiterated, but it’s not a black box. “Your data science team can go in and adjust it.”
Finally, Guard was emphatic about SAS’s commitment to offerings which serve marketers. “A lot of what we hear is that clients have always used SAS in marketing, and a lot of times they’ll say it’s on the ‘science side’ versus the direct engagement side. To answer your question directly, we want to be on the science side and on the engagement side. That’s a big part of where we’ve pushed, and consciously created Customer Intelligence 360. It leverages the best on the analytics side, but also puts business context on the application side for engagement.”
It’s marketing-centric, Guard insisted. “It’s not like we’re going to create a customer service piece, or we’re going to do a salesforce automation piece; so we’ve been very focused on the marketing side, and intend to be. If you look at the three big application areas (with data and analytics, visualization, and business context): customer intelligence, fraud management, and risk management are those three from SAS, the three big ones.”
Isn’t it the case, though, that SAS analytics is used by a lot of brands to prep data which is then pushed to an engagement of MA engine from a competitor?
“We do see that. We love that we’re used for that. But we want to be part of the customer touch-point too. We know that sometimes there’s legacy systems in there, okay. That’s why we pushed CI 360 to be broad: it’s got engagement, discovery, planning, so we can feed the breadth of marketing, and we’ll continue to put more application logic in there.
“Customer intelligence, for us, is definitely something we want to be known for, even expanding into the engagement model, meaning engaging and mapping across the customer journey.”
SAS covered DMN’s expenses to attend SASGF