Why Login Walls Are Invisible to AI Shopping Agents

Some retailers hide prices behind login walls. "Sign in to see pricing" or "Create an account for exclusive access." The business logic makes sense: capture email addresses, build customer profiles, offer personalised pricing. But for AI shopping agents, a login wall is a stop sign.

The Agent's Perspective

A user asks their AI shopping agent: "Find me the best price on a Dyson V15 across these five retailers." The agent visits each site, looks for the product, and tries to read the price. On sites with open pricing, it takes seconds. On a site that requires login, the agent sees a product page with a "Sign in for price" message where the price should be. The agent has no account. It has no way to create one (CAPTCHA, email verification, terms acceptance). It simply cannot complete the task.

The agent doesn't report "I couldn't check this site." It just returns results from the other four. The retailer with the login wall has removed itself from consideration without knowing it.

It's Not Just Prices

Login walls block more than prices. We catalogued what gets hidden behind authentication across 80 B2B and B2C sites:

• Pricing (most common): 34% of B2B sites hide all pricing
• Stock availability: 18% require login to see if a product is in stock
• Product specifications: 12% gate detailed specs behind authentication
• Bulk/wholesale options: 41% of B2B sites hide volume pricing entirely
• Customer reviews: 8% require login to read reviews (not just write them)

Every piece of hidden information is information an AI agent can't use when helping a user make a purchasing decision. If your competitor shows prices openly and you don't, agents will recommend your competitor.

The Credential Problem

"Can't the user give their login credentials to the agent?" Technically yes, but practically this creates real problems.

First, most users don't want to hand login credentials to an AI agent. Security concerns are legitimate. If the agent stores credentials, that's a target for data breaches. If it doesn't store them, the user has to provide credentials every time, which defeats the purpose of having an agent handle the task.

Second, even with credentials, automated login is fragile. The way agents handle authentication and sessions is still evolving, and sites use CAPTCHA, two-factor authentication, device fingerprinting, and rate limiting. An agent logging in from an unfamiliar IP address, browser fingerprint, or location will trigger security measures. The user's account might get locked for suspicious activity.

Third, sharing credentials may violate the site's terms of service. Many retail sites explicitly prohibit automated access, even by the account holder.

What B2B Sites Get Wrong

B2B companies have the strongest attachment to login walls. The argument: "Our pricing is complex and customer-specific, so we can't show it publicly." That's fair. But the solution isn't to hide everything.

Show base pricing or "starting from" prices publicly. Let agents see that a product exists, what its specifications are, and roughly what it costs. The customer-specific discount can stay behind the login. An agent that sees "Industrial Widget X, from £45 per unit" can tell the user this product exists and roughly what it costs. An agent that sees "Sign in for pricing" skips the product entirely.

Agents as a Sales Channel

Think of AI agents the way you thought about search engines fifteen years ago. You wouldn't hide your product pages from Google and expect to get traffic. AI shopping agents are becoming a meaningful source of purchase intent. When someone asks an agent to compare products and find the best option, the sites that provide accessible information get recommended.

Our data from running shopping agent benchmarks shows that when one retailer in a comparison set requires login and others don't, the login-gated retailer appears in agent recommendations 91% less often. Not because agents are biased against login walls, but because they literally cannot access the information needed to make a recommendation.

Practical Changes

Show prices publicly. If you're B2C, there's almost no reason to hide prices. The "sign in for exclusive pricing" approach captures fewer emails than you think and loses more sales than you realise.

Make product data accessible. Even if you keep some features behind login, ensure product names, images, descriptions, specifications, and base pricing are publicly visible. This is the minimum an agent needs to include your product in a comparison.

Offer a guest checkout path. Agents that can add items to a cart and reach a checkout page without creating an account can complete purchases on behalf of users. Forced account creation adds another step where the agent (and many human users) will abandon complex checkout flows.

Provide a product data feed. Structured product feeds (Google Merchant Centre format, for instance) give agents direct access to your catalogue data without visiting your site at all. If the agent already knows your prices and stock levels, the login wall on your site becomes irrelevant.