Enterprise procurement teams have been able to use e-auctions to source and secure supplier contracts for more than 25 years. With these online negotiations, multiple vendors are able to compete directly against each other in real time to win a contract to supply an enterprise with specific products or services. The technology has had a transformative effect, leveling the playing field for vendors bidding on contracts, removing human biases from the procurement equation, and making the process of identifying and coming to terms with suppliers far more efficient.
Now, a new technique, autonomous negotiating—where a machine determines the factors that are negotiable in a supplier contract and carries out negotiations over email or a chat interface—is giving procurement teams even greater capabilities. Amazon is among the companies that have been automating vendor communications and negotiations for years. Both approaches offer advantages over traditional, manual negotiations and the awarding of contracts without any negotiations at all, which can lead to poorer results overall, but which technique, e-auctions or autonomous negotiations, is likely to be the dominant form of corporate buying in the future?
To gain insight on the differences and advantages of the two, I recently spoke with two experts on digital sourcing and negotiations, Jacob Gorm Larsen, Head of Digital Procurement at Maersk Group, and Rik Vera, Partner at nexxworks and a renowned speaker and advisor on customer centricity and digital disruption.
Where the Similarities Lie
Larsen is one of the world’s best-known experts in the field of e-sourcing and he wrote the definitive book on the subject of commercial e-auctions, A Practical Guide to E-auctions for Procurement: How to Maximize Impact with E-Sourcing and E-Negotiation. Surprisingly, Larsen argues that e-auctions and autonomous negotiations are almost the same.
“Auctions, when conducted properly, have two stages: the value discovery stage, where joint value creation is explored, and the competitive bargaining stage, which is the auction process,” says Larsen. “A negotiation is similar, as it involves the same stages, only as part of a single process.”
According to Larsen, the objectives are the same with both approaches, too. Enterprise procurement specialists are looking to get a good deal for both parties, to make the process as efficient as possible and to maintain a strong relationship with their counterparts. The allocation problem they are trying to solve is always the same: discovering who they should give their business to under the most efficient terms possible. But he cautions that credibility and commitment are the most important aspects of the e-auction process, as vendors need to trust that they will indeed be awarded the business if they win.
“An e-auction has to be part of a holistic sourcing process, where you explore joint value creation and understand what factors both sides might flex on besides just the price,” Larsen adds. “What I’m preaching is the concept of ‘negotiauctions’—where you combine the competitiveness of auctions with the interaction and joint value creation of negotiations.”
Another similarity between the two techniques is competitiveness. One might argue that auctions are inherently more competitive than negotiations, as each bidder’s assessment of value throughout the auction process is influenced by their competitors’ assessment of value. That shared information also makes auctions highly efficient and typically enables bidders to avoid the winner’s curse, where the auctioned value of a contract exceeds its intrinsic value. But market-driven negotiation is not unique to e-auctions. Today, autonomous negotiations can generate a similar level of competitiveness, as they rely on a machine that’s able to negotiate with many counterparts simultaneously and take market feedback into account in real time during the negotiation.
What About the Differences?
Using e-auctions in complex spend areas requires human involvement in the value discovery process to ensure a total value approach, which means these auctions are often used for larger, more strategic deals where enterprises can justify investing significant resources in the process. Procurement teams facing time and budget constraints simply don’t have the resources to complete a full sourcing process and e-auction for every tail-spend vendor contract.
Rik Vera, an outspoken proponent of automation who advises Fortune 500 companies on AI and intelligent automation and who regularly speaks and writes about the impact of digitization on business and society, notes that there are fundamental differences between autonomous negotiations and humans’ actions during negotiations.
“People have a tendency to simplify meaning,” Vera says. “Give people two parameters to negotiate over and they can cope well. Give them three and they will struggle. Give them four? The truth is that humans cannot optimize well across four or more parameters. So, if they have seven items to negotiate over, they don’t try to optimize across all seven, but to reduce those to three, which feels like a more manageable set. Machines, however, will embrace complexity and even look for additional items that they can factor in.”
Vera notes that he’s “a big believer” in generic algorithms and that a formula used in an autonomous negotiation will continually adapt based on new data. The technology’s ability to learn throughout the process is a major difference from e-auctions, where the original boundaries of the negotiation are inflexible. “If a machine handles the whole process, it will learn from each negotiation,” Vera says. “If a process relies on humans carrying out the whole phase of value discovery, the AI misses out on learning.”
E-Auctions Require a Competitive Market
E-auctions require multiparty competition, while negotiations require just two parties and, so, have wider applicability. Ostensibly, any business deal can be negotiated, but not every contract can be easily auctioned. For example, holding an e-auction for a real estate deal in a unique location or for maintenance services in an area with no competition isn’t practical—auctions require multiple bidders to be efficient.
Autonomous negotiations, however, can be useful in a huge array of situations. They can be triggered based on external data—such as changes in commodity indices or consumer demand—and, most importantly, the software that enables them can continuously learn and improve, as both value creation and value distribution are part of the same process.
The Takeaway
AI has made massive competitive negotiations possible in the commercial world, as evidenced over the past several years by Amazon and other major players in the retail and logistics industries. Both e-auctions and autonomous negotiations provide a credible alternative to traditional, manual negotiations and typically enable enterprises to generate far better results than awarding vendor contracts without negotiating at all. Given that autonomous negotiations are applicable to a wide range of circumstances and offer many of the same benefits as e-auctions, in addition to machine-learning capability, it’s a safe bet that they will become a critical solution in the future of procurement. Whichever of these two digital sourcing methods an enterprise chooses, it’s critical that procurement professionals focus on joint value creation rather than just on price.
