In my first article from the 2019 SAS Global Forum, I tried to give a sense of the company’s overall vision, and in particular its vision for its marketing analytics and engagement services. In this second piece, I take a deeper dive into use cases, with a SAS executive and a SAS customer.
Michele Eggers is senior director of the SAS customer intelligence product line, which includes its flagship Customer Intelligence 360 offering. As so often, the story for many customers starts with silos.
Bringing the silos together
“One of the biggest trends I’ve seen from customers is the modernization of their marketing platforms, their processes, and how they’re building customer journeys. What I’m also seeing is the fact that they have silos; they’ve been building email programs in one channel, and they have website personalization, social media. They need to bring these together, not just through the technology, but in how they’re engaging the customer in a more omnichannel way.
“A good example is Office Depot Europe. We’ve been working with them for some years, and frankly they weren’t as modern as they needed to be. I’ve really seen, though, in the last year, them really amp up how they deal with marketing — in-store, online, lots of direct marketing activities. They’ve brought the silos together and become much more personal in their marketing.”
The inevitable CDP question
SAS was doing analytics decades before we had the CDP term to content with. CDP or not?
“I look at SAS as a CDP — and more. We want to be our customers’ central hub of all their customer information, although we’ve built a marketing platform that is open, and can link in [to other solutions].” Customers usually don’t want to deploy SAS as just a CDP. “They want us not only to manage and gather and enrich the data, but they also want us to be able to take it to execution: to deliver the emails, to personalize the website, to deliver the content in a mobile app or push notification.”
Taking it to execution
SAS is clearly recognized by observers of the space for its execution capabilities, but its name still doesn’t trip off the tongue in that context as readily as Salesforce, Oracle, or Adobe.
“You’re right. I like to think we’re the best kept marketing secret. Much of it just comes down to the sheer size of marketing investment others put into it; and that’s okay. We build our customer base very organically, and through our accolades from the analyst community. When there are RFPs going on, people pull up the Gartner reports and say, wow, I didn’t know SAS did that. One of the top U.S. retailers had a large RFP process to replace their legacy marketing automation system. They assumed that Salesforce was going to get it, but we had bid for it. We ended up winning the business — and they were surprised. I was talking with the Chief Customer Officer, and she said she didn’t even realize we were in it. I’m proud of that, as a product leader, that we are showing what we can deliver for customers technology-wise.”
The future of customer intelligence
“When we launched Customer Intelligence 360, we were very focused on digital, and insights which could go beyond typical web analytics; and we allowed those digital marketing engagement capabilities that really filled a gap. What I’m excited by now is tying all this together with the planning capabilities — marketing resource management. That gives us a great foundation.
“When I think about the roadmap, much of the special sauce of what we do is about making smart marketing decisions. How do we embed AI into the applications? We’re doing it very pragmatically, providing useful analytical ‘helpers’ for the marketers. An example, embedding AI into testing. Marketers aren’t using analytics enough in their testing, so we’ve embedded AI and optimization capabilities to perform self-learning tests (which go faster); and to optimize multi-variant tests to minimize the volume you have to do to get the results you need. Another example is embedding things like self-learning, automated segment discovery.”
Leaving the journey to AI
The next great opportunity arises, according to Eggers, because CMOs are still struggling to understand what’s working, and what’s not. “Attribution helps me know what’s working and what’s not. We do that today, using analytics embedded in the system to provide those insights. That leads to customer journey analysis, which can change marketers’ strategies. The next evolution — and I don’t believe the market is ready for it — is really leveraging reinforcement learning to create more self-derived customer journeys. Take the prescriptive development of customer journeys out, and let AI drive those decisions.”
Let AI gently guide the customer down the path? “Yes, and it’s going to take time, because that’s a lot of control you’re taking out. I think it’s a multi-step, multi-year process.”
Expanding beyond existing SAS customers
I had wondered whether businesses already using SAS analytics for a range of functions were most likely to be interested in the marketing and customer experience tools. “We definitely have the large customer base that we have because we’ve gone to organizations which see the value of analytics; because one of our differentiators is that we’ve embedded analytics there. But what I’ve seen with Customer Intelligence 360 is that we’ve actually been able to expand our ability to go into accounts which maybe don’t know SAS. And it’s mainly because the analytics are democratized into the application.”
|
Personalizing the campus life Yes, they’re students, but in a practical sense they’re customers too. That’s why the University of Oklahoma uses SAS solutions through what might be called the student life-cycle, from acquisition, through engagement, and hopefully to completion. What I heard about from Shawn Hall, senior data scientist in the Office of Business Analytics at the University of Oklahoma, was the agile and creative use of predictive analytics to improve student experience, as well as support the university in its role as a business. “We function like a business in many ways, and I’ve seen that highlighted with the recommendation system we’ve built. It’s very much akin to what Amazon does.” Tell me about the recommendation system. “We use what we learn about students from the data system about their activities around campus. It started as a simple association analysis using the SAS EM package.” That’s SAS Enterprise Miner, a descriptive and prescriptive modeling tool. “It’s grown into text analytics to pull information from their entrance essays, for example, to make more personalized recommendations for summer courses, student organizations, campus events; to create that sense of community for students.” The recommendations are currently distributed by email, but consideration is being given surfacing personal recommendations on the college website when students log in; and where students could also provide feedback on the relevance of the recommendations. How are the students responding to the recommendation initiative? “We did host a number of focus groups when we started. We wanted to hear from them what they thought would be useful. We knew it was an avenue we could use to increase engagement, but we didn’t always have the best sense of what would work for the students. Summer course and student organization recommendations were things they didn’t feel they were getting enough information about across campus. And the recommendations are specifically personalized for them.” One-on-one, personalized, relevant engagement. Sounds familiar. What about use cases for predictive? “There have been a lot of projects around identification and early intervention programs, for example for financially at-risk students. More recently, we’ve been predicting bar passage rates for students in the law school, as well as identifying stages at which students might benefit from parallel planning, where students’ first major doesn’t work for them.” Pattern recognition can reveal problems before they occur. The office constrains itself when it comes to the types of data analysed. Social media activity is publicly available, and the ID cards students use to swipe into locations around campus — “Loads of useful information there,” said Hall — but there’s an understandable reluctance to be “creepy.” Analytics are also being used to determine where recruitment efforts are best directed, based on what the college is able to predict about what turns a prospective student into an enrolled student. Essentially, it’s about predicting conversion and deploying resources accordingly. “We’ve looked at how much it takes in scholarships to turn a prospective into an enrolled student. Where to even look in the first place, and then how much is it going to take to make you our student.” What about predicting performance of enrolled students? “Yes, we’re making headway in predicting retention as well as graduation rates, and looking at what supplemental efforts could be put in place to boost those numbers, and increase the likelihood of a student being successful in a given program.” Hall first came across SAS in grad school. It’s been used as an educational tool by the University “forever.” It’s more recent that it’s been used in the Office of Business Analytics for the kinds of purposes described here. “For me personally, I’ve been familiar with the tool from the get-go. SAS EM is a very intuitive set-up, and it’s in line with my training as a statistician and audience scientist.”
|
SAS covered DMN’s expenses to attend Global Forum.
