The primary financial risk of agentic commerce to e-commerce companies is not the transaction fee but the potential loss of high-margin retail media advertising revenue. Since many retailers derive most or all of their profit from on-site ads, agents threaten their core business model.

Related Insights

OpenAI's 4% fee for in-app purchases creates a risk for merchants. If consumers start using ChatGPT as their primary purchasing interface, it could intercept sales that originated from a brand's own marketing. A customer might see a product elsewhere, then buy it via ChatGPT, imposing a new tax on an otherwise organic conversion.

AI agents shop based on optimized specs, not human heuristics like brand trust. This shift to "agentic commerce" could neutralize the power of major brands like Walmart and Amazon, and eliminate the interpersonal relationships that sustain local, small businesses.

As AI agents shift e-commerce from high-margin cost-per-click models to lower-margin commissions, search platforms will likely retaliate. They will make free, direct, and unpaid traffic more difficult to acquire, forcing a higher volume of transactions into their paid ecosystem to compensate for the lower per-transaction revenue.

The "DoorDash Problem" posits that AI agents could reduce service platforms like Uber and Airbnb to mere commodity providers. By abstracting away the user interface, agents eliminate crucial revenue streams like ads, loyalty programs, and upsells. This shifts the customer relationship to the AI, eroding the core business model of the App Store economy's biggest winners.

Amazon's potential commerce partnership with OpenAI is fraught with risk. Allowing ChatGPT to become the starting point for product searches threatens Amazon's highly profitable on-site advertising revenue, even if Amazon gains referral traffic. It's a classic battle to avoid being aggregated by another platform.

AI agents acting as gatekeepers will make direct product discovery harder for consumers. This forces new and existing brands to rely more heavily on large-reach social and video platforms for initial awareness, ultimately increasing the value and ad spend on those platforms.

Unlike service platforms like Uber that rely on real-world networks, Amazon's high-margin ad business is existentially threatened by AI agents that bypass sponsored listings. This vulnerability explains its uniquely aggressive legal stance against Perplexity, as it stands to lose a massive, growing revenue stream if users stop interacting directly with its site.

The 4% fee on ChatGPT's checkout isn't comparable to ad spend because it doesn't grant merchants a long-term customer relationship. With restrictions on remarketing, it's a simple transaction cost that erodes margins, not an investment in acquiring a customer with future lifetime value (LTV).

Just as newspapers ceded their audience to Google for traffic, retailers are being tempted to let AI chatbots handle customer interactions. This trade sacrifices brand identity and direct customer relationships for short-term volume—a historically catastrophic move that leads to commoditization by an aggregator.

While a commerce partnership with OpenAI seems logical, Amazon is hesitant. They recognize that if consumers start product searches on ChatGPT, it could disintermediate Amazon's on-site search, cannibalizing their high-margin advertising revenue and ceding aggregator power.