
Agentic commerce is online shopping in which an AI agent, such as ChatGPT, Gemini, Microsoft Copilot or a personal assistant, finds products, compares them and sometimes completes the purchase on the shopper's behalf. For a store, it means your product data and checkout have to work for software as well as for people. In late 2026 most of the effect is on discovery, since checkout inside AI tools is still early and uneven.
If you sell on Shopify, most of the plumbing is already switched on, and your job is clean product data and a few settings, which typically costs $6,000 to $21,000 of agency time in our planning estimate. If you run an open-source, headless or custom store, expect $20,000 to $80,000 or more, because you have to connect your own catalog and checkout to the new protocols. Adobe forecasts that AI-referred traffic to US retail sites will rise 130% this holiday season, and so far that traffic converts better than other traffic. The returns depend on how much of your traffic comes from AI, which our model below lets you test.
We build online stores on Shopify and on custom code, so we have a point of view. We also tell merchants not to pay for protocol work they do not need, and this guide says when that applies.
Date | What happened |
|---|---|
January 2026 | Google, with Shopify, Etsy, Wayfair, Target and Walmart, introduced the Universal Commerce Protocol (UCP), an open standard for agent shopping from discovery through checkout. Microsoft launched Copilot Checkout on January 8. |
March 2026 | Shopify turned on Agentic Storefronts by default, sending merchant products to AI channels. OpenAI pulled back its Instant Checkout, which had charged merchants a 4% fee, after few merchants went live. |
September 28, 2026 | Adobe forecast $275.1 billion in US online holiday sales (up 6.7%) and a 130% rise in AI-referred traffic to US retail sites in November and December. |
October 2026 | Google added a UCP checkout hub and a Universal Cart for US merchants in AI Mode and Gemini. |
November 10 to 17, 2026 | OpenAI announced Businesses in ChatGPT and, on November 13, ChatGPT Wallet, which lets agents buy through retailer checkouts. Google announced Gemini checkout with Google Pay on November 17. Shopify's Fall '26 Edition connects stores to nine AI agents. |
Modern Retail called holiday 2026 the first season when shopping agents really complete purchases, in its November 19 roundup, noting that Walmart lets shoppers cap an agent's spending at $50, $100 or $200 and that Amazon blocked Meta's Muse agent from its site.
Adobe Digital Insights reported that AI-referred traffic to US retail sites rose 71% year over year in September and converted 90% better than other traffic, a reversal from August 2025, when it converted 9% worse. Adobe's holiday forecast puts AI-referred traffic growth at 130% for November and December. Kearney reports that AI-referred traffic converted about 42% better than other channels in early 2026. Bain estimates that roughly one in four online shoppers plans to start holiday shopping on an AI platform.
Most of these figures measure shoppers clicking from AI tools to retail sites, not purchases made by agents on their own. They are fast-growing from a small base, and they describe retail overall, not your store. Sources also differ on details, and several come from press coverage of forecasts. Use them to decide where to look, then measure your own traffic.
Piece | What it does | Who it matters to |
|---|---|---|
Product feed (Google Merchant Center, Shopify Catalog) | Tells AI tools what you sell, at what price and in what stock | Every store |
Universal Commerce Protocol (UCP) | Open standard from Google and Shopify covering discovery, cart, checkout, identity and orders | Stores that want checkout inside Gemini and AI Mode |
Agentic Commerce Protocol (ACP) | Checkout standard from OpenAI and Stripe, used for ChatGPT | Stores that want ChatGPT purchases |
Shopify Agentic Storefronts | Manages AI channels from the Shopify admin | Shopify merchants |
Wallets and tokens (ChatGPT Wallet, Google Pay) | Hold the saved card and the shopper's consent, and pass a token in place of the card number | Stores accepting agent-initiated orders |
Platform | Where it stands | What you still do |
|---|---|---|
Shopify | Agentic Storefronts on by default for eligible stores. Google's checkout reportedly enrolls Shopify stores by default, with an opt-out. A November 2026 probe by TechnologyChecker found about 73% of Shopify storefronts published a UCP manifest. | Clean product data, check eligibility and channel settings, review terms. |
BigCommerce and Wix | The same probe found 85.7% of BigCommerce and 65.0% of Wix storefronts with a manifest. Third-party figure, so verify. | Check your platform's agent settings and feed. |
WooCommerce, Adobe Commerce, OpenCart | Usually need a clean feed and a connector or app, and extra work for protocol checkout. | Feed, structured data, connector. |
Headless or fully custom | You build the connections yourself. | Feed, schema, protocol endpoints, order and risk handling. |
Our ecommerce website development cost guide compares these platforms on cost, and our custom software vs SaaS guide covers the build-or-buy math behind that choice.
There are no published benchmarks for this work yet, so these are our own estimates: hours of work multiplied by a blended US agency rate of $100 to $150 an hour. They are not a quote.
Store type | Typical effort | Planning range | What it covers |
|---|---|---|---|
Shopify-hosted | 60 to 140 hours | $6,000 to $21,000 | Catalog audit and cleanup, feed and policy checks, channel and eligibility settings, tracking. |
Open-source or headless | 200 to 540 hours | $20,000 to $81,000 | The above, plus structured data and a connector or protocol integration. |
Fully custom platform | 300 to 700 hours | $30,000 to $105,000 | The above, plus protocol endpoints and controls for agent-initiated orders. |
Catalog size and quality. Thousands of products with missing attributes cost far more to fix than a few hundred clean ones.
Integrations. Inventory, pricing and promotions have to be accurate in real time, because an agent will quote what you tell it.
Checkout. Custom checkouts need more work than platform checkouts.
Risk controls. Agent-initiated orders need fraud checks, spending limits and clear records of consent.
Extra gross profit per year equals AI-referred visits times conversion rate times average order value times gross margin, times 12. Every input is an assumption you should replace with your own analytics.
Take a store with 100,000 visits a month, a 2.5% baseline conversion rate, a $90 average order value and a 40% gross margin. If AI-referred traffic converts 90% better, as Adobe measured in September, its conversion rate is 4.75%. A $15,000 readiness project then pays back like this:
AI share of visits | Monthly AI visits | Yearly revenue from them | Yearly gross profit | Payback on $15,000 |
|---|---|---|---|---|
0.5% | 500 | $25,650 | $10,260 | 17.5 months |
2% | 2,000 | $102,600 | $41,040 | 4.4 months |
5% | 5,000 | $256,500 | $102,600 | 1.8 months |
The answer swings from under two months to nearly a year and a half depending on one number: how much of your traffic comes from AI tools. So look at your analytics first. Two cautions: some of those sales would have happened anyway, so if only half is truly incremental, double the payback. And Adobe's conversion figure covers retail overall, so your own result may differ. If AI-referred visits are well under 1% of your traffic today, do the cheap fixes now and wait on the expensive ones.
Check your feed. Confirm titles, prices, stock, shipping and returns are accurate in Google Merchant Center and, on Shopify, in Shopify Catalog.
Check your channel settings. On Shopify, look at Agentic Storefronts and the eligibility terms before the season starts, and decide whether you want each channel on.
Fix the worst products first. Start with your top sellers: add attributes, clear descriptions, sizes and compatibility.
Test checkout on mobile. Agents and shoppers both abandon slow or broken checkouts.
Tag AI traffic. Set up analytics so you can see sessions from ChatGPT, Gemini, Copilot and others.
Review fraud rules. Decide how you will treat orders placed by an agent for a shopper.
An agent quotes your price and stock to a shopper. Wrong data means unhappy customers and possible disputes, so keep inventory and promotions in sync.
Wallets and tokens carry the shopper's consent and a stand-in for the card number, but you still need rules for unusual orders, chargebacks and returns. Large retailers are adding spending caps for agents, which tells you where the concerns sit.
Terms are changing quickly. OpenAI's Instant Checkout, which charged a 4% merchant fee, was retired in March 2026, and reports of ChatGPT Wallet's launch say it started without a commission. Read the current merchant terms, and avoid designing your business around any one channel.
Returns, shipping and warranty terms should be clear and consistent, because agents will read and relay them. Our guide to AI agent development for US businesses covers the guardrails that matter when software acts for people.
Audit your catalog, clean your top products, check your feed and set up AI-traffic tracking.
Enable the channels that fit your platform and customers, test checkout end to end, and review fraud and returns rules.
Compare AI-referred conversion and revenue with the model above. If the share is meaningful, invest in deeper integration. If not, hold at the cheap fixes and revisit next quarter.
Which AI channels does my store feed today, and can I turn each one off?
How do you keep inventory and prices accurate in real time?
Which protocols do you support, and what is the roadmap?
How do agent-initiated orders appear in my fraud and reporting tools?
Who owns the integration code and the data feeds?
What does this cost to set up and to maintain each year?
Will you tell me if a cheaper option covers what I need?
Our software outsourcing guide explains how to compare vendors across locations.
If you are on Shopify and mostly need data cleanup and settings, you may not need an agency at all, and we will say so. We are a good fit for open-source, headless and custom stores that need real integration work, or for merchants who want a second opinion on what to spend. We build on Shopify, Magento, WooCommerce and OpenCart, and on custom code.
We do not publish a rate card. A quote starts with a one-to-two-week discovery that settles the catalog, the integrations, the channels and the risks. It produces a roadmap and a scoped estimate tied to roles and hours. We work on a fixed-cost, dedicated-team or staff augmentation basis, depending on how settled your scope is. Our engineers are in-house, and you own the IP.
Start with our eCommerce development company in the USA page and our eCommerce solutions. If you are in Florida, California or New York, see eCommerce development in Miami, in Los Angeles or in New York.
None of these is an agentic checkout, and we will not dress them up as one. They are stores and commerce systems where catalog, checkout and operations had to work together.
We built an OpenCart store for a precious metals retailer serving customers in Canada, covering product discovery, cart, checkout, customer accounts, payments and order history. Read the case study.
We built a QR-linked apparel platform where every garment carries a tag that opens a profile the buyer controls after purchase. Read the case study.
We built one connected system that runs three retail repair locations and a nationwide mail-in repair operation. Read the case study.
Our Google rating is 4.9 from 126 reviews, and we have delivered 180+ projects across 15+ industries.
Agentic commerce is online shopping in which an AI agent, such as ChatGPT, Gemini, Microsoft Copilot or a personal assistant, finds products, compares them and sometimes completes the purchase for the shopper. For a store, it means product data and checkout must work for software as well as people. In late 2026 the biggest effect is on discovery; checkout inside AI tools is still early.
Start with accurate product data in Google Merchant Center and, on Shopify, Shopify Catalog. Shopify's Agentic Storefronts sends eligible merchants' products to AI channels from the admin. On other platforms you need a clean feed, structured data and a connector or app. Check each channel's merchant terms before enabling it.
Not necessarily. UCP (from Google and Shopify) covers discovery through checkout, and ACP (from OpenAI and Stripe) covers checkout for ChatGPT. Shopify and some other platforms handle them for you. Custom and open-source stores may need a connector or their own endpoints. Start with clean product data and add protocol work only if your AI-referred traffic justifies it.
In our planning estimates, a Shopify-hosted store needs about 60 to 140 hours, or $6,000 to $21,000. An open-source or headless store needs about 200 to 540 hours, or $20,000 to $81,000. A fully custom platform needs about 300 to 700 hours, or $30,000 to $105,000. These are estimates at a $100 to $150 blended rate, not published benchmarks.
Often yes, but measure first. Adobe reported in September 2026 that AI-referred traffic to US retail sites converted 90% better than other traffic and forecast a 130% rise this holiday season. In our model, a $15,000 project pays back in under 2 months if AI is 5% of your visits and about 17 months if it is 0.5%. Check your own analytics.
It can be, with controls. Wallets and tokens carry the shopper's consent and replace the card number, but you still need rules for unusual orders, chargebacks and returns, accurate inventory and prices, and clear machine-readable policies. Large retailers are adding agent spending caps. Review each platform's current merchant terms before you enable agent checkout.
Largely. Shopify turned on Agentic Storefronts by default for eligible stores in March 2026, and its Fall '26 Edition connects stores to nine AI agents. You still need clean product data, correct channel settings and a review of eligibility and terms. Google's checkout reportedly enrolls Shopify stores by default with an opt-out, so confirm what is on.
If you want to know what to fix before the next shopping peak, book a technical consultation. Bring your platform, your catalog size and your analytics for AI-referred traffic. We will tell you what we would check in discovery, which fixes are cheap and which are not, and whether we think a bigger investment pays back, whether or not you end up working with us.
Subscribe to the Akoode newsletter for carefully curated insights on AI, digital intelligence, and real-world innovation. Just perspectives that help you think, plan, and build better.