San Francisco-based Demandbase, a business-to-
Spiderbook, which Demandbase acquired in May, offers machine learning and data science to analyze web pages and target accounts. Demandbase capitalized on those capabilities to create DemandGraph, an AI-powered business graph designed to help marketing and sales teams collect business data like vendor, customer and partner relationships.
Demandbase CEO Chris Golec told CMSWire the DemandGraph AI capabilities take Demandbase's digital footprint of information on companies and "exposes it to a much bigger set of data."
"If you think about what Salesforce is doing with Einstein, that's all about applying AI to (customer relationship management) CRM data and the people you're doing business with," Golec said. "Think about applying AI to not only that but all web traffic and the information that exists about those companies across the web."
AI Put to Work
Demandbase officials said DemandGraph will read 400 terabytes of unstructured text every month along with 12 billion website traffic hits of B2B buyers. That information gets structured into a connected account-based network of data assets for AI to translate.
According to company officials, DemandGraph also includes:
- Unstructured business knowledge (SEC filings, annual reports)
- Demandbase’s proprietary identification technology that maps billions of network IP addresses to businesses
- Corporate hierarchies extending beyond subsidiaries and remote offices to include vendor, customer and partner relationships
- Web activity by businesses including ad impressions and web traffic from more than 3 billion B2B interactions every month
With traditional personalization methods, companies can target certain segments and show them specific content.
Learning Opportunities
"You're better off learning what content is actually effective," said Alan Fletcher, chief product officer for Demandbase. Fletcher was Spiderbook's CEO and co-founder. "You want to get an understanding of who's engaging with that certain content. You can let machine learning understand what content will be received the best. And if that visitor comes back tomorrow, they may want a different piece of content."
Personalization is the best way to market and sell, Fletcher said, but it has been impossible to do it at scale.
"But that's what we do by adding an AI layer," he said.
Better Connections than Social
Golec claimed that while social networks such as Facebook and LinkedIn map connections between people based on likes and interests, technology like DemandGraph is much better at doing the same for business. Its repository of information about B2B relationships leads to better targeting, he said.
"What Spiderbook has brought is not just how companies are related inside their four walls but how they do business with their customers, partners and vendors," Golec told CMSWire. "That's what a business graph is. It extends beyond its four walls and is a much better predictor of B2B commerce than a typical social connection."