Hoodie Model Context Protocol (MCP): The Whole Cannabis Market, Inside Your AI Tools
Connect Claude, ChatGPT, and Microsoft Copilot, the AI tools your team already uses, straight to Hoodie's live data. Ask in plain language, and get back an answer plus the next move, not just a report.
ChatGPT Microsoft CopilotHoodie MCP is the connection that lets AI tools plug straight into Hoodie's live cannabis data. MCP stands for Model Context Protocol, an open standard AI assistants use to reach outside data sources. Instead of exporting a spreadsheet and pasting it into a chat, your AI asks Hoodie directly, working from the same market, storefront, retail, and wholesale data your team already trusts, with every request authenticated as you.
Wherever your AI already lives
No new dashboard to learn. Hoodie MCP plugs into the assistants your team opens every morning, so the whole market is one question away from wherever the work is happening.
Claude
Ask for brand reads, distribution shifts, and account briefs in the Claude app, and have Claude build the dashboard or summary from live Hoodie data, with Hoodie skills that package common workflows.
Claude Code
Query Hoodie data from the command line while you build internal tools, reports, and automated workflows on top of it.
ChatGPT
Bring Hoodie's market and account data into the ChatGPT conversations your team already has every day.
Microsoft Copilot
Put market reads next to the docs, decks, and spreadsheets your team already works in across Microsoft 365.
Using another AI tool?
Hoodie MCP is built on the open Model Context Protocol standard. If your team works somewhere else, ask your Hoodie account manager.
One plugin to set up
The Hoodie Analytics plugin sets up the connection for you in each tool. See the setup guides.
One request. A finished analysis.
Hoodie skills turn a single plain-language request into a complete, interactive analysis built on live Hoodie data, right inside the AI tool you already use.
Give me a month over month comparison of my distribution in NY?
- Checking subscriber brands and markets…
- Wyld & Good Tide found in New York.
- Generating volume decomposition from the
/volume-decompositionskill…
/volume-decomposition skill from Hoodie's New York market intelligence, June and July 2026.Four ways in, through one connection
Your AI reaches whatever your Hoodie subscription already covers. Plain questions in, real numbers and a next step out.
Market Intelligence
Market share, distribution, pricing movement, and store-level wins and losses by state and county, so you can see where the market is moving and who is moving it.
Show me state sales growth of pre-rolls in Ohio and compare that to my same store sales as well as my competitors.
Ohio pre-roll sales are up 18% over the last 12 months. In stores that carried you in both periods, your pre-roll sales grew 11%, trailing the category by 7 points. Competitor A is outpacing the category at 24%, driven by new distribution in Franklin and Cuyahoga counties, while Competitor C is down 3%.
Storefront Data
What is on the shelf right now across 12,000+ dispensaries: items, prices, stocking levels, promotions and assortment changes, down to the individual store and SKU.
Which competitors are running promotions near me, and which of my products could be affected?
Recommended: a 35% all-week price reduction on your 28g SKUs only.
Retail Performance
Point-of-sale performance for the stores you operate, or the stores that share their data with you through Hoodie Connect, on rolling windows that stay current.
Take this inventory sheet and tell me which of these products are going to perform best based on historical sales or, if new, your predicted performance.
Blue Dream 3.5g flower and the watermelon gummies are your safest bets, backed by 90 days of steady sales. The new mango live resin cart is predicted to land third, based on how similar carts sell in your stores and nearby. Two items, the 2g rosin and the 14g shake, project under 10 units a week, so consider cutting them from the order.
Wholesale Accounts
Your own wholesale orders, customers, and inventory records from your connected ERP, so account conversations start with what actually shipped.
Show me my weighted sales in New York by breaking down doors that I won or lost this month and which sales person was tied to those accounts.
You added 32 new doors this month and lost only 5, for 27 net new doors. Ritchie Cunningham was largely responsible for 24 of the new doors and $43,523 in sales to those accounts, while losing only 2.
| Sales rep | Won | Lost | Net weighted |
|---|---|---|---|
| Ritchie Cunningham | 24 | 2 | +$39.6K |
| J. Rivera | 5 | 2 | +$5.1K |
| A. Patel | 3 | 1 | +$3.6K |
Example answers describe what comes back; actual results depend on your subscription, markets, and connected data.
What opens up when your AI has the whole market
A connection like this doesn't answer one question and stop. Once your AI can query Hoodie directly, it can build with that data, watch it, and hand it to whoever on your team needs it next.
Live dashboards, on request
Ask your AI to build a dashboard, and it pulls straight from Hoodie instead of last week's export. In Claude, packaged Hoodie skills like the ACV decomposition dashboard build a full won, lost, and whitespace read for a brand in one request.
Monitoring that doesn't wait to be asked
Have your AI check every Monday for new stores that picked up a competitor's line in your territory, or any other watch worth keeping running.
One brief instead of five reports
Before a call, ask for an account brief: what a store carries, how it's selling, when it last ordered. One question, not five spreadsheets stitched together.
Whitespace, mapped
Ask which counties carry your closest competitor but not you, and get the list of stores worth a call, not just a map to stare at.
One connection, as many doors in as your team needs
The same Hoodie connection works across Claude, ChatGPT, and Microsoft Copilot, so sales, retail, and leadership can each ask from the tool they already use.
AI is only as smart as the data underneath it
Hoodie integrates Retail POS, Wholesale, and storefront data from 12,000+ dispensaries, providing both the best coverage and the most statistically accurate market reads in the business. So when your team asks, it knows.
Built on contributed POS data
Sales reads come from 4,000+ integrated POS systems, not panels or surveys, totaling more than $45 billion in sales since 2021.
Cross-validated from all sides
Market estimates are checked against billions of dollars in wholesale transactions and against state-reported sales data.
Trusted where accuracy is public
Public cannabis companies use Hoodie data in their own investor and market reporting, as do the analysts who cover them.
Connected in minutes
Turn it on
Current customers ask their Hoodie account manager to enable Hoodie MCP. New to Hoodie? Start with a demo.
Install the Hoodie plugin
Add the Hoodie Analytics plugin in Claude, Claude Code, ChatGPT, or Microsoft Copilot. It sets up the connection for you. Setup guides
Ask your first question
Start with a market you know well and check the answer against your gut. That's usually the moment it clicks.
Built for the people making the calls
- Brand sales leaders and reps: account briefs, whitespace, and reorder follow-ups without a spreadsheet.
- Retail buyers and operators: store performance and nearby pricing in one conversation.
- Distributors: which accounts and items are slowing down across the book.
- Leadership and finance: market share and trend reads on request, for the board deck or the Monday meeting.
- Data and engineering teams: a standard connection to build internal tools and automated reports on, including from Claude Code.
Priced to fit what you use
Pricing varies depending on the applications you use and the number of markets you'd like access to. There is no one-size-fits-all plan, so the best next step is a conversation.
Talk to our teamHoodie MCP FAQ
What is Hoodie MCP?
Hoodie MCP connects the AI tools your team already uses directly to Hoodie's live cannabis data over the Model Context Protocol, an open standard. Instead of exporting a spreadsheet and pasting it into a chat, your AI asks Hoodie directly and gets back an answer grounded in real sales, storefront, and wholesale data.
Which AI tools work with Hoodie MCP?
Hoodie MCP works with the Claude app, Claude Code, ChatGPT, and Microsoft Copilot. Setup guides for each are in the Hoodie docs. Hoodie MCP is built on the open Model Context Protocol standard, so if your team uses another AI tool, ask your Hoodie account manager about it.
What data can my AI reach?
Whatever your Hoodie subscription already covers. Market Intelligence questions need Market Intelligence access; storefront data needs Sales Enablement or Dispensary Analytics; retail performance needs Hoodie Connect; and wholesale questions need a wholesale (ERP) data subscription. If a module isn't on your account, your AI will tell you it can't reach that data rather than guess.
Can my AI see data my account doesn't have access to?
No. Every request is authenticated as you, and results are limited to the data your subscription already covers. Connecting an AI tool does not give it access to anything you cannot already see in the Hoodie app.
What are Hoodie skills?
Skills are packaged instructions bundled with the Hoodie plugin that help Claude carry out common Hoodie workflows for you. The first, the ACV decomposition dashboard, builds a two-month read of stores won, lost, and retained for a brand in a state, with a dollar bridge, whitespace, and a county map. More are on the way.
Does this replace the Hoodie platform?
No. Hoodie MCP is another way into the same data. Your dashboards, alerts, and applications stay exactly where they are; MCP lets you ask the same data questions from inside your AI tools, and have those tools build with the answers.
How accurate are the answers?
Your AI works from the same data behind every Hoodie report: contributed POS data from 4,000+ integrated POS systems, cross-validated against wholesale transactions and state-reported sales. Because it queries Hoodie directly instead of guessing, answers trace back to real numbers you can check in the platform.
How long does setup take?
Minutes. Once Hoodie MCP is enabled for your account, you install the Hoodie Analytics plugin in your AI tool, which sets up the connection for you, then ask your first question.
How much does Hoodie MCP cost?
Pricing varies depending on the applications you use and the number of markets you'd like access to. There is no one-size-fits-all plan, so the best next step is a conversation with your Hoodie contact.
Where insights meet relationships, now one question away.
When you're ready, the same data you already trust is waiting inside the AI tools you already use.
