SAS partners played a bigger role than ever in the event this year, as SAS continues to move beyond its legacy direct sales days and invest more in its channel.
TORONTO – Curiosity Meets Capability was the theme of the main keynote at the annual SAS Toronto roadshow for executives. The event this year was held Wednesday at the Globe and Mail Centre across the street from SAS’s offices east of downtown Toronto, and there didn’t appear to be a vacant seat in the house. Many of the execs in attendance were focused on AI, rather than being general line of business people, and the focus of the event was to explain to them how to optimize the chances of their projects being successful.
“The goal is to demystify AI and walk through real world examples, to show why it’s not that hard or that new,” said Shadi Shahin, SAS’s VP of Product Strategy, in the main keynote. “If you approach it in the right mindset, it’s not science fiction. We want to equip customers with tools to go back to their organizations so that AI is more than a science lab and becomes real workflows.”
Most organizations face a fundamental challenge around AI, Shahin said. They understand the value of data, and want to be data-driven, but don’t know how to get there.
“Most are still doing basic data preparation work, like BI [Business Intelligence] and dashboards,” Shahin noted. “Most get those analytics components, but most are stuck there. They want to participate at scale, but haven’t figured out how to do it. They want to, but are stuck with legacy practices. Analytics is a bridge, not an outcome, so in getting to true analytics, the issue is how do you make it achievable for you.”
Most organizations don’t do so effectively, he emphasized.
“50% of data models never see the light of day,” he said. “90% of those that do take more than three months to get out of the door because of internal friction. That means it’s probably too late. Models in regulated industries like credit risk typically last about 9 months. That means that if it takes three months to deploy, you are falling behind.”
Shahin also noted that only about 5% of enterprise data today is being mined – a fairly optimistic data point as some other sources report it as even lower, between 2 and 3%.
“It’s hard to manage and cleanse, and there are a lot of tools in the chain of modelling and converting to output,” he stated.
Overcoming these hurdles is not a technology problem, Shahin stressed.
“It’s not a technology barrier – it’s a process barrier,” he said. “If you think about analytics as a lifecycle in the same way we think of DevOps, there is a process gap.”
Still, he cited multiple examples to the audience of SAS customers, and the very different ways those SAS reference customers have been able to use AI successfully. Volvo’s Truck division uses SAS’ analytics platform to do remote diagnostics and improve uptime. Swisscom uses text analytics in conversational AI to better understand customer needs and solve issues. Seacoast Bank, a small regional bank in Florida competing with larger banks, uses SAS analytics to generate a customer lifetime value model to help them identify and retain top customers. Healthy Nevada Project, a large community-based population health study, uses SAS analytics to systematically assess demographic, genetic and environmental data to improve health.
Dennis Nilsson, AVP and Chief Data Scientist, at TD Insurancem was part of a panel on how customers can unlocking the barriers to operationalizing analytics.
“In traditional analytics, accuracy is the key,” he said. “With AI, it’s opening the Black Box and making the model more transparent. You need to start small and think big, and work with data infrastructure, in a cross-functional engagement, not just give it to the data scientists.”
Raymond Outar, Director of Omnia Artificial Intelligence at SAS partner Deloitte Canada, who leads Deloitte’s SAS Practice in Canada, emphasized how Deloitte had brought divergent groups together to get AI synergies. Omnia is actually an integrated a team of data scientists and broader thought leadership at Deloitte.
“Building out the right use cases and bringing thought leadership and data scientists together has been a challenge,” he said. “But people are starting to explore the art of the possible. Some of the biggest impacts are identifying business problems and then working back, instead of starting with the data and trying to figure out a problem to solve with it.”
Achille Ettorre, an AI expert at Queen’s University’s Smith School of Business, also stressed the importance of getting the process right.
“A lot of people think AI is just a commodity where you hire some consultants, and it’s done,” he said. “It’s not easy to do a true AI solution. The culture is key, and going in for small wins rather than trying to solve all the problems at once. There needs to be fully alignment of people, processes and technology, and understand what technology partner you need to solve the problem. All need to be aligned for the Last Mile in operationalizing analytics to be successful.”
Harper Reed, a technologist and entrepreneur who was CTO of Obama for America 2012, emphasized the importance to prioritize ethics throughout the process. He noted that few engineers took ethics courses, which is something that needs to change. He also emphasized the importance of securing proper permissions for every piece of data that was used and making sure that attribution of the data to those who use and present it is clear. Transparency of the whole process is critical.
“The best results come when we pair AI with people,” Reed added.
SAS historically sold direct for years, and the channel business is new and relatively small but is growing. The channel, because of the nature of their trusted relationship with customers, can have an advantage in leading customers through the intricacies of the process to successful AI deployments.
“SAS isn’t new to analytics, but it is relatively new to partnering,” said Doug McLaren, Director, Alliances & Channels at SAS Canada. “It’s a critical part of how SAS views the future of analytics. Many times, they are better positioned to take customers on those journeys. Many partners are developing use case and customer stories which are Powered by SAS, but it’s their IP, and their advisory services that stand out. The way our partners sell is not a commodity product.”
McLaren explained SAS’s phased approach to managing the partner business.
“This year in Canada, the channel-only segment is customers with $650 million in revenue and below,” he said. “The midmarket – $650 million – $3 billion is channel-first.” Over $3 billion is enterprise, where SAS leads.
“We felt we weren’t ready to go completely channel in the midmarket, but we are getting ready to announce our 2020 strategy for Canada, which will have a higher channel percentage,” McLaren added.
“This is our premier event for the year in Toronto and partners were in every aspect of it,” he stressed. They led sessions, and were in panels. Of the 6 demo pods, three of them were led by partners. While it is a phased approach, partners are critical to the future of this organization, and are intimately part of this.”
McLaren said that the SAS channel consists of three types of companies.
“We have recruited a small number of what we call Connected Partners, who have a lot of clients, a strong sales capacity and sales engine, and who would like to grow into analytics, but don’t have deep domain expertise or deep SAS capabilities today. Some of them, like CT Global and Zencos, have deepened their expertise and hired dedicated SAS salespeople.”
McLaren said that SAS has ramped up some of these partners by pairing them with the second partner group, Service Partners.
“Service Partners have a long history with SAS, typically on the delivery side,” he stated. “We have connected many of our Connected Partners with our Service Partners.
The third partner group, which has larger SAS capabilities, are termed Multi-Dimensional partners. The large integrators are key here, although McLaren pointed out that they aren’t just an enterprise play.
“Deloitte and KPMG have service offerings on risk, powered by SaaS, where they do sell into the midmarket,” he said.
Going forward, McLaren emphasized that the channel role will expand further, to help customers on their AI journey.
“There is a lot of competition and clients have a lot of choices,” he said. “But we have a strong market presence, strong customer references, and a focus with our partners where we are investing significantly for the long term.”