It’s about to become far more important for merchants to offer consumers appropriate credit as agentic commerce emerges.
In this new shopping model, autonomous AI agents will act on behalf of customers to discover products, compare options and complete purchases with minimal human intervention. Bain & Company forecasts that the U.S. agentic commerce market could be worth $300 to $500 billion by 2030, representing 15% to 25% of total ecommerce.
This is problematic for merchants relying on legacy Buy Now, Pay Later (BNPL) credit. BNPL has become mainstream at checkout, with billions of dollars flowing through installment platforms. The method suits higher-ticket items, such as fashion, electronics, travel and furniture, where installment payments reduce sticker shock and boost conversion, especially on products with strong margins to balance provider fees.
Previously, retailers faced a simple calculation. If BNPL significantly improves conversion and the process remains simple, it’s beneficial. If margins are tight or the workflow too complex, traditional credit might be preferable.
When AI Starts Making Purchase Decisions
Agentic commerce complicates this equation. Now, AI agents include credit checks in recommendation and purchase decisions. When an AI tries to buy and credit is denied, it moves on. When this happens often enough, the agent deprioritizes both the lender and the merchant.
Worse, merchants may not even know why they’re losing sales. The agent shifts the consumer’s business elsewhere because it favors payment methods with high authorization rates.
Why Payment Performance Shapes what Gets Recommended
AI systems are designed to optimize outcomes, not just surface options. When an agent recommends a product, it implicitly assumes responsibility for whether the purchase succeeds. Failed transactions introduce friction into that loop.
Over time, systems learn from those failures. Products and merchants associated with incomplete or declined transactions are perceived by the agent as less reliable. In contrast, merchants with consistent authorization outcomes and smooth payment flows become safer recommendations.
This creates a feedback loop. Payment performance doesn’t just affect checkout; it influences which products are surfaced, how often they appear and whether they are recommended at all.
Unfortunately, traditional BNPL’s industry approval rate is only 35%-40%, making it less reliable in agentic commerce environments, where every decline can work against the merchant.
For retailers, this shifts how payment strategy should be evaluated. The question is no longer just whether a payment option increases conversion at checkout. It is whether that option can support consistent, predictable outcomes in an AI-mediated journey.
That requires a broader view of payment performance, including approval rates, friction in the flow and the reliability of transaction completion across channels. Methods that perform well in a traditional checkout may introduce too much variability when decisions are made programmatically.
As AI begins to influence more purchase journeys, retailers will need to assess payment options not just as conversion tools, but as part of how their products are discovered and selected.
Be Predictable
Retailers can avoid BNPL pitfalls in an AI-driven shopping environment by ensuring their payment methods are reliably predictable. Payments that consistently work will earn AI agents’ repeat business and appeal to today’s consumers.
Combining BNPL’s convenience with traditional credit’s predictability is appealing. Credit products that split purchases into interest-free payments and draw directly from an existing credit card are ideal because no new credit applications or BNPL accounts are required. With card-linked payments, approval rates typically top 85%. Using existing card limits and history sidesteps new credit risk and avoids hard/soft pulls, unlike traditional BNPL. Merchants receive full payment upfront from the card issuer, often the same day.
By converting a total price into monthly payments, installments change the affordability signal an AI agent uses when evaluating options. That increases the likelihood that higher-value products appear in the recommendation set.
For example, say a consumer asks AI to buy a new grill for the backyard. The agent locates a $400 option that meets all requirements: size, number of burners, construction quality and positive reviews. Unfortunately, the price gives the consumer sticker stock.
With card-linked installments, the agent can offer a solution: “If you want to pay over time with your existing credit card, you could pay $100 a month for four months. No new account, no application, no additional fees.” Everyone wins: the shopper gets the grill they want, and the merchant makes a sale that might have been lost.
What Retailers Should do Now
Retailers should act now, even as agentic commerce continues to evolve. The goal is not to predict exactly how AI-driven transactions will unfold, but to ensure the fundamentals are in place.
Start by auditing payment performance. Look closely at approval rates, decline patterns and where friction enters the checkout flow. These signals will become more important as AI-driven traffic increases.
Next, evaluate whether payment options support consistent outcomes across environments. AI does not distinguish between online, in-store or assisted channels; it prioritizes reliability.
Finally, treat payment flexibility as part of the product experience, not just a checkout feature. When affordability can be incorporated earlier in the journey, it expands what customers can consider and what AI systems can recommend.
As AI agents compare, negotiate and place orders on behalf of consumers, they will increasingly favor retailers with payment infrastructures that are predictable, reliable, and easy to execute. Those that align their payment strategy to that reality will be better positioned to compete as commerce becomes more automated.
Nandan Sheth is CEO of Splitit, a global fintech enabling interest-free installment payments using existing credit cards. A seasoned payments and fintech executive, Sheth brings more than 20 years of experience scaling companies and advancing digital commerce. Prior to Splitit, he led global digital commerce initiatives at Fiserv and held senior leadership roles at American Express. A proven entrepreneur, he co-founded Harbor Payments and Acculynk, both acquired by industry leaders.
