An IBM lead-sharing tool highlights how AI is making inroads in systems that facilitate partner engagement.
Launched in 2018, Smarter Cognitive Opportunity Recommendation Engine (SCORE) uses IBM cognitive computing to guide and expedite the lead-sharing process. So far, IBM’s software organizations have adopted SCORE. The system is expanding to IBM’s hardware business units, the company said. SCORE follows on the heels of IBM Business Partner Connect, a tool that uses Watson to help partners team up for customer opportunities.
“What [we are] trying to do … is make cognitive the epicenter of all our transformation initiatives, and we are starting here with SCORE,” said Mike Fino, COO of IBM Partner Ecosystem.
IBM had been focused on automating its lead-sharing engine for about two years, Fino said. “Up until we started with the SCORE effort, writing static programs or static algorithms didn’t lend themselves well to the lead-[sharing] environment, because it is such a fluid space,” he said. In late 2017, executives considered tapping into IBM cognitive technology, and, in doing so, improve the speed and quality of lead sharing.
Fino said IBM’s AI technology gives SCORE a powerful edge over traditional lead-sharing engines. “That is really the secret of what SCORE gives us, which is the ability to learn from every positive and negative decision or recommendation that we make, factor that [knowledge] into the next one, and have that be the living engine that we use to engage with our partners,” he said.
How SCORE makes recommendations
SCORE’s top priority was to establish a system that would “pass the right opportunities — in other words, the opportunities our business partners are most likely to win — to the right business partner … in a timely fashion,” said Alan Zwiren, cognitive and AI transformation manager, IBM partner ecosystem.
To determine the right IBM business partners, SCORE assesses partner firms against an array of attributes. Attributes can include the partner’s win rate, experience, geographical proximity to the prospect and competencies. Much of the information that SCORE uses is based on partners’ historical performance data but is extending to other data sources as well, Zwiren said. When SCORE generates a recommendation, it also gives users an explanation for how it came to its conclusions.
“We tried to provide explanations to show end users which key features of specific business partners are influencing the recommendation,” noted Oznur Alkan, research scientist at IBM Research.
A feedback loop, which collects feedback from IBM sales, business partners and other sources, lets SCORE continuously learn and improve its recommendations.
Zirwin added that SCORE evaluates the attributes used to rank partners on a monthly basis. “We look at many different dimensions … and each of these dimensions are weighed once a month automatically by SCORE to say, ‘Which is the most predictive [dimension] of a won deal?’ ” he said.
‘A North Star’ for IBM business partners
SCORE has proven valuable in two fundamental respects, Fino said. The first is for its assistance in getting IBM’s leads into the right partner hands. The second benefit is that it has helped incentivize certain behaviors among IBM business partners. Those behaviors can help partners improve their ranking within the engine.
“[We] are starting to see our partners focusing on the same things that we believe are important to not only help them improve their status, but … we are seeing more and more partners use [SCORE] as a bit of a North Star to drive their enablement efforts,” Fino said.
More partner tools that use IBM cognitive computing are on the horizon. In addition to SCORE and IBM Business Partner Connect, Fino said his team is developing a tool that uses AI to help partners with Cross-selling by recommending adjacent offerings.
“I think this is really just the beginning,” he said.