Staples Inc. is out to more quickly engage business customers by getting sales and customer service reps on the same virtual page. The two departments are using machine learning and artificial intelligence to speed up the ordering process through its automated Easy System and make recommendations using algorithms developed for the website.
Staples Business Advantage, the office supplies retailer’s online B2B segment at StaplesAdvantage.com, is using technology to enable its human and machine resources to share information about customers’ preferences and ultimately predict their needs, says Zubair Murtaza, vice president of e-commerce product management and customer experience.
Machine learning refers to computer programs designed to absorb information and learn how to develop new insights, such as when the Staples Easy System learns about customers’ interests and anticipates their demands with personalized recommendations.
Staples has machine-learning tools in place now and is developing what it calls “new frontier” technology. “We have things that are very machine-learning-intensive, such as self-service on the website, triggers for savings and tools for reps powered by the same algorithms used on the website,” Murtaza says. “The new frontier features bot-driven artificial intelligence with IBM’s Watson.” Watson, a supercomputer named after IBM founder Thomas Watson, uses artificial intelligence and analytics technology to produce useful information tailored to specific applications. A bot is interactive software that uses artificial intelligence to simulate human conversation.
By tapping into IBM’s Watson Conversation service, Staples offers an ordering interface for customers across the Staples Easy System, whether via Staples’ Easy Button or other channels such as an app, text, or email. In addition to driving simpler interactions with customers, cognitive capabilities built into the system using a combination of Watson technology and Staples’ internal personalization application programming interfaces, or APIs, will help the Staples Easy System learn more about each businesses’ preferences over time, including their preferred products and quantities.
An API is a set of software instructions designed to get disparate software applications to integrate and share data.
As Staples sees it, the new frontier tools offered through its Easy system are evolving, but some are useful in the here-and-now. Here’s how the button-based technology works: When a customer speaks to the Easy Button, a bot receives the message and tries to interpret the request. If the message is too complex for the bot, the query goes to a human to handle, Murtaza says. However, the bot constantly listens to each request and learns, even if a human is answering the question. The bots will learn over time how to improve their answers and handle more requests, he adds.
Staples is using natural language processing—technology that deduces the searcher’s intent through speech to show relevant results—in its Easy Button testing. The company launched an Easy Button alpha test last summer and expanded it in the fall to a small group of Staples Business Advantage customers in the Austin, Texas, area. Beta testing will grow to hundreds of customers in spring 2017, Murtaza says.
Staples Advantage, No. 22 in the 2017 B2B E-Commerce 300, also wants machine learning to inform customer service and sales reps about those customer interactions and will provide online dashboards they can use to access customer data and recommend products based on the same algorithms used on the website.
Murtaza declined to provide details on Staples’ investment in machine learning, but notes it’s growing.
“We have always been investing in how we use machine learning to become more personalized and target the customer,” he says, noting funding for machine learning has become “a dominant part” of Staples’ investment in e-commerce. “Machine learning-driven engagement across our channels is a top priority,” Murtaza adds.
Sign up for a free subscription to B2BecNews, a twice-weekly newsletter that covers technology and business trends in the growing B2B e-commerce industry. B2BecNews is published by Vertical Web Media LLC, which also publishes the monthly business magazine Internet Retailer. Follow Bill Briggs on Twitter @BBriggsB2B.
Follow us on LinkedIn and be the first to know when B2BecNews publishes new content.