The next customer to buy from an online store may not be a person at all. Shopping assistants built into ChatGPT and other AI tools can now research products, compare options, and in a growing number of cases complete the purchase on a shopper’s behalf. For the brands that sell online, that changes a basic assumption behind every storefront: the visitor reading the page is no longer always human.
Netalico, an ecommerce agency that has specialized in Shopify since 2013, has spent the past two years rebuilding its playbook around that shift. The firm now operates as an agentic commerce agency alongside its core development work, preparing merchant storefronts for a world where AI agents read catalogs, evaluate merchants, and move through checkout with instructions from the humans they serve.
From NASA Systems To Storefront Machinery
The company’s direction owes a lot to the background of founder Mark Lewis, an engineer whose early career included enterprise systems work at NASA before he turned to ecommerce. Lewis still operates as a fractional ecommerce CTO for several larger brands, and that engineering-first orientation shows up in how his agency approached the AI question: rather than publish opinions about where commerce is heading, the team built instruments to measure it.
The first was an internal probe that tests which brands the major AI assistants actually name and cite when shoppers ask for recommendations. The second became public: a free scorecard that fetches any store and grades, from A to F, how well AI agents can read its products, trust its data, and complete a purchase. Merchants use it the way they once used site-speed graders, and it frequently delivers uncomfortable news. Most stores, Netalico has found, fail at least one of the tests that matter: product data that machines can parse, crawler access for AI systems, content that renders without JavaScript tricks, and brand facts that stay consistent everywhere the models look.
“Merchants spent twenty years optimizing for a shopper with a mouse and ten seconds of patience,” Lewis says. “The new visitor reads every word, checks your data against the rest of the web, and never gets tired. Stores that are legible to machines are going to compound an advantage that others will struggle to catch.”
Two Protocols, One Preparation
Behind the shift sit competing technical standards. OpenAI has published a protocol that lets merchants sell directly inside ChatGPT conversations, while a broader effort backed by Google and Shopify aims to make any catalog readable and purchasable by any agent. Which standard wins matters less than most merchants think, in Netalico’s view, because the preparation is nearly identical either way: structured product data, open and deliberate crawler policy, machine-readable brand information, and checkout flows that automated buyers can traverse.
That preparation overlaps almost completely with a discipline the agency was already practicing: generative engine optimization, the work of earning brand citations inside AI-generated answers. The same structured data and consistent facts that help a store get recommended by an assistant also help an agent buy from it, which is why Netalico folds both into one program as a Shopify AI SEO agency rather than selling them as separate futures.
An Unusual Shape For An Agency
Netalico itself runs against several industry norms. The company, a Shopify Partner since 2016 that today holds Shopify Plus Premier Partner status, the most selective tier of Shopify’s partner program, keeps its entire team of Shopify developers in-house with nothing offshored. Development is AI-accelerated, with senior engineers directing the tools and controlling quality, an approach the firm credits for compressing build and migration timelines across hundreds of Shopify projects.
Its client roster leans toward mid-market and enterprise DTC brands in the $2M-$50M GMV range, including Big Green Egg, the ceramic grill maker whose Shopify Plus migration and ongoing support Netalico handles, along with names like Vinyl Me Please and HydraPak. The same team that builds a store stays on it after launch, on retainers that run $2,700-$10,000 per month with the average client near $4,500, while project builds span $25,000-$250,000 and up. Hubs in New York and Los Angeles anchor a team working with merchants across the US and Canada.
That continuity, the firm argues, matters more in the AI era than it did before. Agent readiness is not a one-time project but a property a store either maintains or loses, the same way site speed decays without attention. A team that carries a storefront from build through every subsequent season is positioned to keep the machine-facing layer current as the protocols themselves evolve.
What Comes Next
Lewis is careful not to overclaim on timing. Human shoppers still complete the overwhelming majority of purchases, and the protocols are young. But he draws a line between uncertainty about the calendar and uncertainty about the direction.
“Nobody can tell you the quarter when agent-driven orders become a meaningful share of revenue,” he says. “What you can know today is whether an agent that shows up to your store can do its job. Fixing that costs a fraction of what merchants spent on replatforming waves of the past, and everything it requires, cleaner data, faster pages, consistent facts, makes the store better for humans in the meantime.”
For a company that started more than a decade before AI assistants could hold a conversation, the bet is consistent with its history: commerce keeps changing shape, and the merchants who win are the ones whose infrastructure is ready slightly before the customers arrive. Netalico’s business is making sure that this time, when the customer arrives as software, the store is ready to sell to it.
