Deep Steep
What regulated commerce requires
For regulated merchants, infrastructure ownership has moved from a technical preference to a business advantage.
This summer, vape and electronic nicotine delivery system merchants received notices telling them to remove their products from a major commerce platform.
Their storefronts, customer data, order history, and integrations had all been built on infrastructure controlled by someone else. The final decision was not theirs, and neither was the timeline.
The story is often framed as a platform policy change. The platform updates its rules. Merchants scramble. Analysts debate the regulatory environment. Eventually, the news cycle moves on.
But the larger lesson is about control.
Cannabis merchants know the pattern. Firearms accessories sellers have faced it. Supplement brands operate under changing enforcement expectations, while age-gated alcohol merchants watch processor and platform requirements shift around them.
Regulated commerce places more weight on every infrastructure decision.
A merchant builds on a platform. The platform decides the category creates too much complexity or risk. The merchant then discovers that access to the storefront was conditional all along.
The merchants best prepared for these changes are the ones that addressed the infrastructure question early.
They control their hosting. They control their data. They control their compliance configuration. A policy change from a platform they do not depend on remains an industry development rather than an operational emergency.
This is where heritage matters.
Webscale has supported regulated merchants for years. Famous Smoke Shop runs production traffic on Webscale infrastructure.
Age verification, geographic product restrictions, payment requirements, and compliance-heavy catalogs create real production challenges. Webscale's infrastructure has developed around those challenges through years of supporting merchants that operate inside them every day.
Webscale has maintained 100% uptime over 10 years. That number carries additional weight in regulated commerce.
For these merchants, an infrastructure failure can create consequences beyond lost revenue, particularly when age verification, product restrictions, or required transaction controls are involved.
Webscale brought that operational experience into the design of Agentic Commerce OS.
The AI layer runs alongside infrastructure already built for compliance-heavy catalogs, live first-party data, and merchant-controlled rules. The AI Shopping Assistant reads the merchant's own catalog and current product context. The merchant decides what it can say, which products it can surface, and when a shopper should be routed elsewhere.
Access to a capable foundation model is no longer the primary differentiator.
The advantage comes from the commerce context around the model, the data available to it, the controls governing its behavior, and the infrastructure keeping the experience available.
For regulated merchants evaluating their next move, the real question is who controls the infrastructure beneath the storefront, the data, and the AI, and whether that foundation was designed for the requirements of their category.
The Pour
What shipped, what changed, and what commerce leaders should know.
Product
Product page context went live in the AI Shopping Assistant this month. The assistant can now read the full PDP a shopper is browsing and answer questions about that specific item without the shopper restating what they're looking at.
Fitment banner launched for auto parts and appliance catalogs. Per-store search settings let merchants tune relevance scoring by storefront. The Playground preview gives merchants a sandbox to test assistant behavior before pushing live.
Market
Essentials Plans launched at $499 per month, creating a mid-market entry point for merchants that want Commerce Cloud infrastructure without an enterprise-sized commitment.
From the Blog
Prepare your store for AI-mediated commerce.
Adrian Luna · July 2026
Non-browser rendering and structured product data are the priorities. Crawler-friendly architecture ties them together. The protocols, Google UCP and OpenAI ACP, are shipping now.
Read the post →By the Numbers
254%
More revenue per visit from AI-driven referrals during the 2025 holiday season, year over year. The merchants capturing that lift are making their catalogs readable by AI crawlers and keeping their infrastructure fast enough to serve them.
Source: Adobe Digital Insights, January 2026.
$0.62 vs $7.40
That's the cost of an AI-handled customer resolution against a human agent. The gap isn't closing. As AI assistants get better at answering sizing and compatibility questions, the merchants who deploy them first absorb the savings longest. Order status lookups are next.
Source: McKinsey, 2026.
Merchant's-Eye View
What product page context changes
If you're running the AI Shopping Assistant, the product page context update matters more than it looks.
Before: the assistant could answer general catalog questions. "Do you carry brake pads for a 2019 F-150?" worked. But "Is this one compatible with my truck?" while the shopper was already on a PDP required the shopper to restate the product name. That's friction. And friction on a PDP is where carts die.
Now: the assistant reads the page the shopper is on. It knows the SKU and the product details. "Is this one compatible?" just works.
What to test first
Run a week of assistant conversations on your highest-traffic PDPs and look at the question patterns. If shoppers are asking clarifying questions the PDP should already answer, fit or materials, that's a signal your product data needs tightening before the assistant can do its best work. The AI is only as good as the catalog it reads.
What to tune
If you're running per-store search settings, also new this month, pair that with product page context. Different storefronts have different catalog strengths. A storefront heavy on technical specs benefits from stricter relevance scoring. A storefront heavy on lifestyle products benefits from looser matching. The assistant adapts to both, but only if the search settings match the catalog shape.
The Brew Ha Ha
Forward this to the team that keeps the site standing.
The Chai gives you the strategy. The Brew Ha Ha gives Devs the stack trace.
Subscribe
Webscale Networks, 35111F Newark Blvd, Unit #2025, Newark, CA 94560, United States
