There has been a lot of talk over the years about hiring chatbots to serve as negotiation agents. Now it looks like AI-powered chatbots may be entering mainstream use. The question is, can they recognize a good deal when they see one?
Maybe not. But for the people on the front lines of negotiations, AI-powered chatbots can offer welcome help. “This is an age-old problem in public procurement: corporate buyers do not have time to fully negotiate with all suppliers,” Remko Van Hoek, professor of supply chain management at the University of Arkansas, wrote with co-authors in the Harvard Business Review. “Historically, this has left untapped value for both buyers and suppliers.”
Van Hoeck and his co-authors point to the prominent example of Walmart, which has implemented AI-based chatbot negotiation software to connect with suppliers. “So far, the chatbot has negotiated and concluded agreements with 68% of the suppliers it approached, with each party gaining something of value,” they note. AI helped Walmart negotiators deal with its more than 100,000 suppliers.
I recently met with Martin Rand, CEO and co-founder of Pactum, who helped lead the implementation of AI-based negotiation and procurement for Walmart Canada, where the effort first began in January 2021. “In procurement, we are now seeing, that AI is starting to help contract negotiations with suppliers,” says Rand. “Negotiations are ubiquitous and the basis for all value creation in mainstream businesses, making them an ideal place to apply AI technology.”
With thousands of suppliers to manage, it’s a challenge to conduct targeted negotiations for each one,” says Rand. “As a result, firms and their suppliers often settle for blanket agreements with little negotiation, leaving valuable trade terms on the table. With AI, intelligent chat platforms can interact with human suppliers, streamline tasks requiring little human input, generate savings and improve contract terms for both businesses and their suppliers.”
“We’re just now seeing companies wake up to the many possibilities,” says Ajay Agrawal, CEO and founder of SirionLabs, creator of a lifecycle contract management platform. “For example, companies have found that using AI to organize and analyze data in their contracts can flag risks, ensure regulatory compliance, help negotiate better terms and improve the quality of life of their legal teams by reducing time spent reviewing contracts. “
At Walmart, “instead of someone in the supply department contacting each individual supplier to negotiate terms and price, they use an autonomous negotiation platform to service thousands of contracts,” Rand continues. “The solution was first implemented in Canada and has now expanded to providers in the US, Chile and South Africa. So far, the chatbot has generated an average savings of three percent.
Agrawal identifies the killer application for AI in contracting as helping businesses manage risk. “All commercial transactions are governed by contracts,” he says. “So businesses have literally thousands of contracts with millions of pages of dense text, jargon and legalese. Manually reviewing this data for risk identification and regulatory compliance is a nightmare.”
The biggest issue holding back the adoption of AI in the negotiation process is the fear of job loss. “The adoption of AI has been hindered by the belief that the technology will take away the jobs of human employees,” says Rand. “Instead of replacing human employees, we eliminate the mundane aspects of a procurement manager’s job at low cost, freeing up time for strategic negotiations.”
AI must look beyond simply completing transactions and provide insights that offer greater value in contract negotiations – and scale appropriately to the size of the transactions involved. “In contract negotiations, a common mistake is for AI to blindly optimize what is being asked for, instead of accurately defining the value of the terms,” says Rand. “While a term value of plus or minus 20% may not matter for a single trade, AI can execute thousands of trades simultaneously. If the value turns out to be 20% lower than it should be, then the technology can reach agreements that harm the enterprise. For example, tell the AI to accept at least a 1% improvement, while thinking of a supplier with costs of $10,000, that would mean a profit of $100 would be the minimum. But what if AI was negotiating with a $1 billion supplier? An improvement of 0.9% would mean a loss of $9 million in profit. Autonomous negotiations bring scale to what people do. That’s how they create enormous value, but it’s also how they can do harm if they’re given inaccurate information.”
AI can do even more in the negotiation process. “Applying AI to contracts provides answers to all sorts of critical questions, such as why some business relationships are not performing well and where to sell more customers,” says Agrawal. “This makes contractual terms, obligations, service levels and service data details easily accessible to everyone in the enterprise. From a cost perspective, using AI for data analysis leads to significant savings in contract negotiation and exponentially greater value when used to track contract performance.”
Follow me at Twitter.