Airbnb beat its own revenue, bookings and profitability targets in Q2 and raised its full-year outlook, CEO Brian Chesky said, crediting its bet on becoming an AI-first travel platform. The company isn't just using AI for customer support — it's embedding it into product, search and operations to accelerate growth. If your brand still treats AI as an isolated support project, Airbnb shows that embedding it into your core business hits margin directly, not just customer experience.

Etsy reported that traffic from AI agent platforms remains under 1% of its total, CEO Kruti Patel Goyal said, even though that traffic brings higher-intent shoppers than average. The data confirms AI agents aren't yet converting at the rate of human traffic, but they're better at selecting who they bring. If you sell on marketplaces, get ready for that traffic to grow beyond the traditional checkout: converting agent traffic requires listings and catalogs built for them.

The RealReal accelerated GMV growth by 22% after rebuilding its operations around AI, using its Athena system to automate authentication and cataloging for secondhand luxury goods. A new pricing algorithm completes the shift, replacing manual processes that used to cap how fast it could scale. If your catalog still depends on manual review to authenticate, classify or price items, this is proof that automating those bottlenecks with AI translates directly into more sales.

ShipBob launched an AI suite that connects to merchants' live operations through its own agent, Bobby, or Anthropic's Claude, built on top of its in-house fulfillment tech stack. Giving AI agents direct access to real-time operational data speeds up inventory and shipping decisions that used to take days. If your logistics provider doesn't yet offer this kind of AI agent access, you'll fall behind brands already optimizing their supply chain with live data.

A US appeals court overturned a March order that had blocked Perplexity's AI shopping agents from accessing Amazon, after Amazon argued the agents violated a federal anti-hacking law. The ruling confirms marketplaces can't unilaterally block third-party AI agents shopping on your customers' behalf. If your Amazon listing isn't ready for these agents to read and recommend it, you risk losing sales to competitors who are.

Alibaba unveiled Qwen3.8-Max, a 2.4-trillion-parameter model it claims outperforms GPT-5.6 on autonomous computer-use tasks, though independent analysts question whether that performance translates into lower real-world cost. The race for the cheapest, most capable model for autonomous agents is heating up, and it directly affects what it costs to run your own AI agents. Before picking a model purely on its benchmark ranking, factor in what it will actually cost to run in production.

Shopify president Harley Finkelstein said on the Q2 earnings call that the platform and its ecosystem are "well-suited to this agentic era," after two years investing in its search index. Shopify isn't reacting to agentic AI — it's spent years building the infrastructure that makes it possible across its ecosystem. If you sell through Shopify, check how your store shows up in that search index: it's the gateway shopping agents will use to find you first.

Gorilla Commerce, which topped nine figures in sales largely through Amazon, has sold on Walmart's marketplace for two years and expanded into Walmart's physical stores last month. The brand shows that relying on a single marketplace is a risk even when it's working well: diversifying channels, including physical retail, protects your business from any one platform's algorithm or policy changes. If your brand lives on Amazon alone, now is the time to evaluate a second sales channel.

Ecommerce accelerator Pattern describes a "virtuous cycle": it uses brand data to deliver results, those results attract more brands, and those brands feed back even more data, co-founder and CEO Dave Wright said on its earnings call. The more brands that plug into a shared AI and data platform to sell on marketplaces, the better the system gets for everyone, not just the leading brand. If you manage catalog across multiple marketplaces, betting on tools that learn from aggregated industry data gives you an edge no single brand can build alone.