Agentic commerce: Google and Microsoft launch in-chat purchases inside AI assistants

Traditional e-commerce has relied on a purchasing process that remained virtually unchanged for twenty-five years: a shopper searches for an item in a browser, visits a website, clicks through category pages, adds products to a cart, and completes a checkout form.

That conventional funnel has entered a period of structural disruption. Google and Microsoft have almost simultaneously rolled out official protocols that allow AI assistants to close complete commercial transactions directly within chat interfaces.

What happened

Google introduced its open Universal Commerce Protocol (UCP), an architecture engineered to link AI agents across Google Search and Gemini directly with commerce backends like Shopify, Etsy, and Walmart. Under this standard, a consumer can ask an assistant to search, evaluate, and purchase an item within a single conversational thread, without being redirected to the merchant’s storefront.

Almost simultaneously, Microsoft launched Copilot Checkout, integrating payment infrastructures such as Stripe and PayPal directly into the Copilot chat pane. In initial merchant pilot programs, the company reported substantial increases in instant conversion rates by removing the traditional friction of abandoned shopping carts and repetitive customer registration forms.

Together, these initiatives mark the formal transition from conversational search to agentic commerce: a model in which the buyer delegates product discovery and checkout execution to autonomous software operating within pre-authorized budget and delivery parameters.

Why it matters for business

For brand leaders, wholesalers, and online retailers, the rise of agentic commerce introduces three fundamental strategic shifts:

1. The gradual decline of the digital storefront: Direct visits to product landing pages will decline for routine replenishments and standardized goods. Purchase decisions will depend less on web page visual design and far more on the precision with which machine agents can parse technical specifications, verified stock, and live pricing.
2. The elimination of checkout drop-off: The steepest revenue loss in digital commerce occurs on checkout pages, where up to 70% of shoppers abandon carts due to lengthy forms or unexpected steps. Processing purchases instantly with saved ecosystem credentials drastically shortens the sales cycle.
3. Competing for algorithmic endorsement: Merchants will no longer compete solely for organic search keywords, but for becoming the primary recommendation that language models verify as dependable, accurately reviewed, and continuously in stock.

How to apply to day-to-day work

Businesses that sell products online do not need to rebuild their software stack, but they must adapt their data feeds for autonomous machine agents:

• Structure your product catalog in Google Merchant Center: Verify that your feeds contain standard product identifiers (GTIN, MPN, SKU), unambiguous attribute descriptions, dimensions, and explicit return policies in clean, machine-readable formatting.
• Enable accelerated checkout gateways: If your storefront runs on systems like Shopify, activate native integrations with conversational channels and one-tap payment vaults (such as Shop Pay or Stripe Link).
• Maintain real-time inventory synchronization: Autonomous purchasing agents penalize merchants who fail to fulfill confirmed orders. Real-time ERP or inventory management integration is now critical to avoid disqualification from automated purchasing networks.