Buyer and seller value from optional automated bargaining in electronic commerce
This paper develops a checkout-relative value model to demonstrate that while optional automated bargaining can generate combined value for both buyers and sellers, its net economic benefit is highly conditional on specific product contexts, policy configurations, and interaction costs, often favoring the seller at the expense of the buyer or failing to benefit both parties simultaneously.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the world of online shopping, two distinct ways to buy things have long existed. The first is the posted price, where a seller lists an item for a fixed amount, and the buyer either pays it or walks away. The second is bargaining, where a buyer and seller haggle over the price until they agree on a number. For decades, economists have studied how these two systems interact, asking whether letting customers negotiate actually helps the market or just creates confusion and wasted time. The central question is not just whether a deal gets struck, but whether the act of negotiating creates more value for both sides than simply buying at the fixed price would have. If a customer spends five minutes haggling to save a few dollars, but in doing so loses the chance to buy the item at the original price, or if the seller lowers the price for someone who would have paid full price anyway, the system might actually be losing value.
Researchers Bankatesh Choudhary and Meenakshi Sharma from the Birla Institute of Technology in India set out to answer this question by building a detailed computer simulation of an online store. They did not ask real people to bargain; instead, they created a virtual marketplace where automated computer agents acted as sellers, trying to negotiate with simulated buyers. The goal was to see if adding an optional bargaining feature to a standard checkout process would benefit both the buyer and the seller, or if it would simply shift money around or cause losses. They tested this by comparing a standard "buy now" checkout against a scenario where a buyer could choose to negotiate first. They measured not just the final price, but the entire journey: how many new sales were created, how many existing sales were lost because the price was lowered, how many deals fell apart completely, and how much time and effort both sides spent during the process.
The researchers found that the answer depends entirely on the specific situation and the behavior of the people involved. In a standard, average scenario where buyers have a reasonable chance of buying at the full price, an automated bargaining system can indeed create extra value for both parties. In their main test, the automated seller agent managed to generate a small but positive gain for itself and a small gain for the buyer for every original shopping opportunity. However, this positive result was not universal. The study revealed that a policy that makes the seller the most money is not necessarily the one that makes the most money for both sides combined. In fact, the most sophisticated automated negotiator the researchers tested, which used a complex learning method to make decisions, actually produced a slightly lower total value for the pair than a simpler, more predictable rule-based approach.
The simulation also showed how easily things can go wrong. When the researchers tested a scenario where buyers were "stubborn" and refused to move from their initial low offers, the automated bargaining system caused the seller to lose money. The system spent time negotiating, failed to reach a deal, and in the process, the seller lost the chance to sell the item at the original price to a customer who was willing to pay it. Similarly, when the researchers simulated a group of buyers who simply did not value the product enough to pay the full price, the bargaining process still cost them time and effort. Even though the bargaining sometimes enabled a sale that wouldn't have happened otherwise, the time spent negotiating was not worth the small discount they received, leaving the buyers worse off overall.
A key discovery was that the benefits of bargaining are hidden by the average. While the overall numbers looked positive in the best-case scenario, the researchers looked closely at individual products and found that only about 42 percent of the specific product situations actually benefited both the buyer and the seller simultaneously. For the other products, one side or the other lost out. This means that a blanket decision to turn on bargaining for an entire store could hurt many customers and the business itself, even if the average numbers look good. The study also calculated the maximum amount of time or effort a buyer or seller could afford to spend on a negotiation before the deal stopped being worth it. They found that if the cost of bargaining is too high, or if the chance of returning to the original price after a failed negotiation is too low, the system breaks down.
The researchers were careful to state that these results are based on a computer model, not on real-world sales data. They used public records of past prices and past negotiation conversations to build their simulation, but the specific behaviors of the buyers and the costs of their time were assumptions built into the model. They did not claim to have discovered a new law of economics or a perfect way to run an online store. Instead, they provided a clear framework for how to think about the problem. They showed that to know if bargaining is useful, a business must look at the specific details of their customers and their products, rather than just looking at the final price of a deal. The study concludes that while automated bargaining has potential, it is a delicate tool that requires careful setup. If a store wants to offer it, they must ensure that the system does not waste the time of customers who are ready to buy, and that it does not force sellers to give away profits to customers who would have paid the full price anyway. The value of the system lies not in the technology itself, but in how well it is matched to the specific people and products it serves.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.