# Hello Retail MCP server for Claude, ChatGPT and Codex

> The Hello Retail MCP server connects Claude, ChatGPT or Codex to every Hello Retail product, so ecommerce teams build and run their own personalization by asking in plain words.

- **Page**: [https://helloretail.com/en/mcp/](https://helloretail.com/en/mcp/)
- **Server address**: https://core.helloretail.com/mcp
- **Setup guide**: [https://developer.helloretail.com/mcp/setup/](https://developer.helloretail.com/mcp/setup/)
- **Every tool, documented**: [https://developer.helloretail.com/mcp/](https://developer.helloretail.com/mcp/)
- **Works with**: Claude (Code, Desktop and Projects), ChatGPT and OpenAI Codex
- **Price**: Free for Hello Retail customers
- **Tools**: 165 across 10 areas of Hello Retail. 94 read a store's data, 71 change its setup.

## What you can build by asking

You bring the idea. The AI tool you already use turns it into search rules, pages, designs, newsletters and Product Agents, through real Hello Retail tools. Each request is one the Hello Retail MCP documents.

### Put the right product first in search

Name the product and the search. The rule is added, and shoppers see it straight away.

- **Product area**: Search
- **Example request**: "When people search for wool, show them the Camel wool coat first"
- **Hello Retail tools used**: `productData_search`, `search_listProductEngines`, `search_addQueryRule`
- **Result**: Done. Anyone who searches for “wool” now sees the Camel wool coat first. (Live in search)

### Build a campaign page by asking

Ask for the page, say what belongs on it and what should come first. It arrives as a draft for you to check.

- **Product area**: Pages
- **Example request**: "Build a page for our summer campaign"
- **Hello Retail tools used**: `pages_copyDesign`, `pages_createConfig`, `pages_updateConfigProductFilters`, `pages_updateConfigProductBoosts`
- **Result**: Draft, ready for you to publish

### Restyle every recommendation box at once

Describe the look in plain words. One design is changed and set on every box, so the store looks like one store.

- **Product area**: Recommendations
- **Example request**: "Give all our boxes the new look"
- **Hello Retail tools used**: `recoms_getDesign`, `recoms_copyDesign`, `recoms_updateDesign`, `recoms_updateSelectedDesign`
- **Result**: Ready for you to check

### Fill your newsletter with products for each reader

Newsletter Content puts product tiles in the email you already send, picked for each reader. Setting it up is one request.

- **Product area**: Newsletter Content
- **Example request**: "Set up a newsletter campaign for our Mailchimp or Klaviyo newsletter"
- **Hello Retail tools used**: `newsletterContent_listDesigns`, `newsletterContent_createCampaign`, `newsletterContent_previewCampaign`
- **Result**: Your campaign is set up. Here is a preview of the product tiles a reader gets. (Preview ready)

### Turn on a Product Agent, on your terms

Say which agent, which products and which channel. Product Agents then email each customer about the product that matters to them.

- **Product area**: Product Agents
- **Example request**: "Start emailing customers about price drops, but only for our own brand"
- **Hello Retail tools used**: `productAgents_getChannels`, `productAgents_getSettings`, `productAgents_updateAgent`
- **Result**: Agent on

## What you can ask about your store

Your AI tool reads your search, recommendation and sales data in Hello Retail, tells you what it found and offers to act on it. The answers below are illustrative, for a demo store.

### Find the searches that come up empty, and fix them

Your AI tool reads what shoppers searched for, matches it to products you sell and closes the gap.

- **Product area**: Search
- **Example request**: "Which searches find nothing, and can you fix them?"
- **Hello Retail tools used**: `search_getTopSearchesWithoutResults`, `productData_search`, `search_addSynonyms`
- **Result**: Four of them match products you sell, like “trainers” for sneakers. They are added as synonyms and work once your catalog has been re-indexed. (Fixed)

### Spot the recommendation box that slipped

You get the box, the reason and a better strategy to try, ready as a draft.

- **Product area**: Recommendations
- **Example request**: "Which recommendation box is underperforming, and what should we change?"
- **Hello Retail tools used**: `recoms_getAnalyticsGrouped`, `recoms_getAnalyticsDailyForKey`, `recoms_listBestPracticeAlgorithms`
- **Result**: The cart box. Its clicks have dropped. I can switch it to a best-practice strategy as a draft for you to review.

### Know what to promote this weekend

Product Intelligence ranks your products by the day they sell best, and by how they sell at a discount.

- **Product area**: Product Intelligence
- **Example request**: "What should we promote this weekend?"
- **Hello Retail tools used**: `productIntelligence_getTopProducts`
- **Result**: These sell best on Saturdays, and two of them also sell well at a discount. Want them first on your sale page and in the next newsletter?

## Built on Product Intelligence

An AI coding tool can build a store in a weekend. Now give it over a decade of global buying patterns data and see your business flourish.

Through the same connection, your AI tool asks Product Intelligence how each of your products sells: by day of the week, time of day, season and discount, and to which customers. Product Intelligence learns from over a decade of buying patterns across 250 million products.

## What it covers

- **Search**: 52 tools (24 read, 28 change). Example: "Which searches find nothing, and can you fix them?"
- **Recommendations**: 23 tools (13 read, 10 change). Example: "Which recommendation box is underperforming, and what should we change?"
- **Pages**: 25 tools (12 read, 13 change). Example: "Our category page converts badly - why?"
- **Newsletter Content**: 24 tools (16 read, 8 change). Example: "Design product tiles for our newsletter"
- **Product Agents**: 5 tools (3 read, 2 change). Example: "Start emailing customers about price drops, but only for our own brand"
- **Retail Media**: 7 tools (4 read, 3 change). Example: "Which sponsored campaigns are running, and are they working?"
- **Product Intelligence**: 1 tool (1 read, 0 change). Example: "What should we promote this weekend?"
- **Product data and feeds**: 18 tools (13 read, 5 change). Example: "Why does this product show the wrong price?"
- **Orders**: 1 tool (1 read, 0 change).
- **Operations**: 9 tools (7 read, 2 change). Example: "Who changed this, and when?"

## How to connect

1. Add the server address https://core.helloretail.com/mcp to Claude, ChatGPT or Codex.
2. Sign in with your Hello Retail account. No API keys needed.
3. Ask in your own words.

A tool that changes something asks for confirmation first, and changes are recorded in the audit log. New to Hello Retail? [Book a demo](https://helloretail.com/en/demo/) and our team sets you up first.

## Frequently asked questions

### What is the Hello Retail MCP server?

The Hello Retail MCP server connects an AI tool such as Claude, ChatGPT or Codex to your Hello Retail account. MCP stands for Model Context Protocol, an open standard that lets an AI tool use another system's functions as tools. Once connected, you ask in plain words, and the AI tool reads your store data and changes your Search, Recommendations, Pages, Newsletter Content, Product Agents and Retail Media setup for you.

### What is the best MCP server for ecommerce?

It depends on what your store runs on, because an MCP server works on one system. For search, recommendations and personalization on Hello Retail, it is the Hello Retail MCP server. It has 165 tools and 71 of them make changes, so your AI tool can find the searches that return nothing, check a product feed and change a search rule or a recommendation design, asking you to confirm first. It is free for Hello Retail customers.

### Which AI tools does it work with?

Claude (Code, Desktop and Projects), ChatGPT and OpenAI Codex. The [setup guide](https://developer.helloretail.com/mcp/setup/) has the steps for each.

### What can I do with it?

The server has 165 tools across 10 areas of Hello Retail. 94 of them read (search analytics, recommendation performance, product data, feed runs, the audit log) and 71 make changes (search rules and synonyms, page and recommendation designs, newsletter campaigns, Product Agents settings, Retail Media campaigns). The full list is in the [developer documentation](https://developer.helloretail.com/mcp/).

### Is it safe to let an AI tool change my store?

You sign in with your Hello Retail account through OAuth, so no API keys are pasted around. Tools that make a change ask for confirmation before they run, and changes are recorded in the audit log. New pages, recommendation boxes and Retail Media campaigns are created as drafts that you publish in My Hello Retail.

### What does it cost?

The Hello Retail MCP is free for Hello Retail customers. You need your own subscription to the AI tool you connect.

### Do I need a developer?

No. Connecting takes the server address and a sign-in, and after that you ask in your own words. The same server works from coding tools such as Claude Code and Codex, for developers and agencies who build custom features on Hello Retail. If you want help, [our team builds with you](https://helloretail.com/en/demo/).

## Further reading

- [What is MCP? A store owner’s guide](https://helloretail.com/en/learn/what-is-mcp-ecommerce/)
- [How to connect your AI tool to your store](https://helloretail.com/en/learn/connect-ai-tool-to-store/)
- [What your AI tool can change in your store](https://helloretail.com/en/learn/what-your-ai-tool-can-change/)
- [What you can build, product by product](https://helloretail.com/en/learn/what-to-build-with-your-ai-tool/)
- [Build your own ecommerce personalization](https://helloretail.com/en/learn/build-your-own-ecommerce-personalization/)

---

## About Hello Retail

Hello Retail is an ecommerce personalization platform based in Copenhagen, Denmark. Online retailers use it to build and run their own search, recommendations, merchandising, email and retail media on its Product Intelligence engine.

- **Website**: [helloretail.com](https://helloretail.com)
- **Demo**: [Let's talk](https://helloretail.com/en/demo/)
- **AI information**: [helloretail.com/en/ai-info](https://helloretail.com/en/ai-info)
