Today, your customer comes to your website, browses and, hopefully, places an order. Tomorrow, your customer’s AI purchasing agent will communicate with your AI sales agent—and the agents of your competitors—decide which supplier has the best offer and place the order.
That’s the vision laid out today by Trent Gillespie, CEO of consulting firm Stellis AI, in a keynote address at Distribution Strategy Group’s Applied AI for Distributors conference in Chicago.
To win those sales, distributors first must build those AI-powered sales agents. And they must enable those agents to access product catalogs and pricing to make offers to buyers’ AI purchasing agents—securely—or they won’t have a chance to get those orders, Gillespie said.
“The old internet was built for humans,” he said. “If you want to be part of the future of AI commerce you’re going to need to be discovered there.”
Nor is this decades away, he said, as some companies are already experimenting with this kind of AI-to-AI communication. “We’re at the edge of this becoming a regular thing,” he said. “We’ll see examples of this this year and in two to three years it will be commonplace. If you’re not ready for this, you’re not ready for the future of commerce.”
Gillespie acknowledged that most companies are nowhere near implementing artificial intelligence at this level of sophistication. In fact, he cited a Distribution Strategy Group survey that found only 13% of distributors see AI as vital to their long-term success.
The important thing, he emphasized is to get started in a disciplined way. The way to do that is for every team within an organization to set a monthly goal of solving a problem with AI and learning from those small projects.
“Every month, discuss what new ways can we use AI this month? Try it, and reflect on it at the end of the month: What did we learn? Did it work? Adjust course. It is a race, and organizations moving with that discipline are outpacing the rest,” he said.
He gave examples of distributors already getting results with AI. For example, one company he’s worked with applied AI to invoice matching and achieved an 82% reduction in invoice processing cost and a 72% increase in speed. “Those are savings that go right to the bottom line,” Gillespie said.
Cost and Security
Getting started with AI need not be expensive, he added. A company can give employees access to a general-purpose generative AI system like ChatGPT for $30 per month, and license software for creating AI-powered agents for sales, customer service and other purposes for $499 monthly.
But protecting company data is essential, and every organization will need to create guardrails about what AI systems employees can use and what data those systems can access. For starters, he said, it’s important to pay for enterprise licenses for AI technology for systems like ChatGPT rather than using free versions that capture data that’s entered, potentially exposing proprietary information.
Once AI agents start interacting with other companies’ AI agents, he said, it will be important to establish rules about what information can be shared, and with which other organizations. Selling organizations may not want to reveal all their inventory or pricing information and buyers may be reluctant to reveal every product they’re purchasing.
On both sides of agent-to-agent interactions, Gillespie said, “there will be internal work required to define those protections.”
Eventually, he said, distributors may have agents dedicated themselves to each customer that know what those customers buy, when they buy it and at what price. That could enable a selling agent to suggest replenishment orders and new products. “It might even start to sell stuff you don’t have because it can find other products and services to sell to your customers without you knowing it,” he said.
Getting to that level of AI functionality, Gillespie said, “won’t come from one project. It will come from 100 projects, maybe 1,000, and it will come from the people closest to the work, which is you.”
How will AI help you serve your future customer?
To get there, companies need to redesign their processes around AI and incentivize their employees to use AI regularly, with appropriate safeguards. He encouraged attendees to get out the word about AI experiences and learnings, through company meetings and newsletters. “Small wins every single month will get you there,” he said.
Drawing on his years of experience as a technology leader at Amazon.com Inc., he recalled how Amazon always tries to envision what its customers will want in five years and work backwards to determine what the company needs to build now to meet those customer expectations. Customers are already starting to expect AI-driven service and that will be increasingly common, he said.
“Position your company to win with AI and design your company for AI-enabled customers,” Gillespie said.
The AI race is just starting, he said, so don’t worry that you’re too far behind to catch up, he said. “The bold aren’t waiting for AI to happen to them,” Gillespie said. “They are using it to reinvent what’s possible.”
Don Davis, former editor-in-chief of Internet Retailer magazine and Vertical Web Media, is a freelance writer based in Chicago. His experience in retail and distribution goes back to his childhood when he worked in the toy wholesale business founded by his father and two uncles and in their discount department stores located throughout the New York metropolitan area.