What's New

About the Datadrew MCP connection — security, data access, and revocation

What is Datadrew MCP? Datadrew MCP is a read-only endpoint at https://mcp.datadrew.io/mcp that lets external AI assistants — Claude, ChatGPT, and any other Model Context Protocol client — query your Datadrew data. It uses the open MCP standard and the same OAuth 2.1 sign-in flow you are used to from other apps. This article explains what your AI assistant can and cannot do, how sign-in works, and how to revoke access. Available on every plan Datadrew MCP is included on every Datadrew plan, including Free. You do not need a paid subscription to connect Claude, ChatGPT, or any other MCP-compatible AI assistant. Read-only by design Every tool exposed via Datadrew MCP is read-only. Your AI assistant can: - Query orders, products, customers, and inventory from Shopify. - Pull campaign performance from Meta Ads, Google Ads, Amazon Ads. - Read traffic and conversion data from GA4 and Google Search Console. - Read email metrics from Klaviyo and Brevo. - Pull subscription data from Recharge, Skio, Stripe. - Read warehouse, review, and shipping data from Unicommerce, Judge.me, AfterShip. It cannot: - Create, edit, or delete anything in your store. - Send emails, SMS, or ads on your behalf. - Change Datadrew settings or billing. - Access integrations you have not connected in Datadrew. - Access data from any other Datadrew account. How sign-in works 1. The AI client registers itself with Datadrew via OAuth 2.1 Dynamic Client Registration (RFC 7591). 2. Your browser opens to app.datadrew.io to sign in. This is the same sign-in you use for Datadrew itself — email/password or Google. 3. If your account has multiple shops, you pick which one the AI assistant will query. 4. Datadrew issues an access token (1h) and a refresh token (30d) to the AI client. 5. Every tool call uses the access token. When it expires, the client silently rotates it using the refresh token. You never share your Datadrew password or API keys with the AI client — it only ever sees short-lived tokens. Token lifecycle at a glance | | | | | --- | --- | --- | | Token | Lifetime | What it does | | Access token | 1 hour | Signs every tool call. Auto-refreshed by the AI client. | | Refresh token | 30 days | Used to get new access tokens. Rotated each use. | | Session idle | 30 days | 30 days without use → sign in again. | Multi-shop accounts One Datadrew MCP connector is scoped to one shop at a time. If your account has multiple shops, switch inside the chat: - "List my Datadrew shops" - "Switch Datadrew to [shop name]" How to revoke access From Datadrew: 1. Click your profile avatar in the top-right and open Connected AI assistants (also reachable from Settings → Datadrew MCP). 2. Click Revoke next to the AI client. What happens on revoke. The refresh token is invalidated immediately, so the AI client can no longer mint new access tokens. An access token that was already issued keeps working until its 1-hour expiry — this is how stateless OAuth 2.1 tokens work. For most uses the next tool call will fail within a few minutes as the client tries to refresh. If you need to cut off in-flight access instantly, also remove the Datadrew connector inside the AI client itself. From your AI client: remove the Datadrew connector from its settings. This deletes the local tokens so the client stops making requests. What Datadrew logs Datadrew records which tools your AI assistant calls, the shop the call is scoped to, and the timestamp — the same way we log other activity in your account. We do not log response content beyond what is required to serve the request. Frequently asked questions Does Datadrew MCP cost extra or consume AI credits? No. It's included on every plan at no additional cost. AI credits are only consumed when you use Drew AI (Datadrew's in-app AI analyst). External MCP connections do not draw from your credit balance. Which AI clients are supported? Any client that supports custom remote MCP servers, including Claude Desktop, Claude.ai, ChatGPT (Pro/Business/Enterprise), Claude Code, Cursor, Windsurf, and most self-built MCP clients. Will the AI assistant see my payment details? Stripe tools return invoice, charge, and payout summaries — the same data visible in the Stripe dashboard. They do not return raw card numbers. Can I limit which tools the AI assistant can use? Today the connector exposes read access to all connected integrations. To exclude a specific platform, disconnect it under Datadrew Integrations. Is my data leaving Datadrew? Your AI assistant receives only the data it requests to answer your question — same as your browser receiving dashboard data. What your AI client does with that data (including whether it's used for training) is governed by that client's privacy policy, not Datadrew's. Need help? Reach out at support@datadrew.io or use the in-app chat.

Connecting Datadrew to Claude, ChatGPT, and other AI assistants

Connect any AI assistant to Datadrew Datadrew exposes an open Model Context Protocol (MCP) endpoint at https://mcp.datadrew.io/mcp. Any MCP-compatible AI assistant — Claude Desktop, Claude.ai, ChatGPT, Claude Code, Cursor, Windsurf and others — can use this URL to read your connected data in a safe, read-only way. Once connected, you can ask your AI assistant things like "What was my Meta Ads ROAS last week across all accounts?" or "Pull the top 20 Shopify products by revenue for Q1" and it will fetch live data directly from the integrations you have connected in Datadrew. Which plans include Datadrew MCP? Every plan, including Free. You do not need a paid subscription to connect an AI assistant to Datadrew. Open Settings → Datadrew MCP inside the app to get started. Requirements - A Datadrew account with at least one Shopify store connected. - The integrations you want the AI assistant to access (Meta Ads, Google Ads, Klaviyo, Shopify, GA4, Stripe, etc.) already connected in Datadrew. - An MCP-compatible AI client. This guide covers Claude Desktop, Claude.ai, ChatGPT, Claude Code, and generic clients. What your AI assistant can see Your AI assistant gets read-only access to the integrations you have connected. It can run queries and retrieve data, but it cannot modify your store, send emails, or change settings. Supported data sources: - Shopify (orders, products, customers, inventory) - Meta Ads, Google Ads, Amazon Ads - Google Analytics 4, Google Search Console - Klaviyo, Brevo - Stripe, Recharge, Skio - Amazon Seller Central, Unicommerce - AfterShip, Judge.me Your AI assistant only queries integrations you have already connected in Datadrew. It cannot reach any platform you have not connected, and it cannot access data from any other Datadrew account. Connect Claude Desktop 1. Open Claude Desktop. 2. Go to Settings → Connectors. 3. Click Add custom connector. 4. Enter: - Name: Datadrew - URL: https://mcp.datadrew.io/mcp 5. Click Add. A browser window opens for sign-in with your Datadrew account (same email/password or Google). 6. If you have more than one Datadrew store, pick the one you want Claude to analyse. Try it: "Using Datadrew, show me my Shopify revenue for the last 30 days." Connect Claude.ai 1. Open claude.ai and sign in. 2. Go to Settings → Connectors. 3. Click Add custom connector. 4. Paste https://mcp.datadrew.io/mcp as the connector URL. 5. Click Connect. A popup takes you to Datadrew to sign in and pick your store. 6. Return to Claude.ai. The Datadrew connector is available in any chat — click the connector menu at the bottom of the message box to enable it. Connect ChatGPT ChatGPT supports custom MCP connectors on Pro, Business, and Enterprise plans. 1. Open chatgpt.com and sign in. 2. Go to Settings → Connectors. 3. Click Add custom connector. 4. Paste https://mcp.datadrew.io/mcp as the connector URL. 5. Sign in with your Datadrew account and pick the store you want ChatGPT to analyse. 6. Enable the Datadrew connector in your chat's tool menu. Connect Claude Code Run the following in your terminal: claude mcp add --transport http datadrew https://mcp.datadrew.io/mcp The first time you use a Datadrew tool, Claude Code opens a browser for sign-in. Connect any other MCP-compatible client Datadrew speaks standard MCP. Any client that supports custom remote MCP servers can connect using the same URL. Connection details: - URL: https://mcp.datadrew.io/mcp - Protocol: Model Context Protocol (MCP) - Auth: OAuth 2.1 with dynamic client registration (RFC 7591) and PKCE - Transport: streamable-HTTP (JSON-RPC over HTTPS POST) Your client will automatically register itself and send you through a browser sign-in to Datadrew. Examples of clients known to work with generic MCP endpoints: Cursor, Windsurf, Zed, Continue, and most self-built MCP clients using the official SDKs. Using Datadrew if you have multiple stores If your account has more than one Shopify store, sign-in prompts you to pick the active store for the assistant. You can switch later without re-adding the connector — ask the AI assistant: - "List my Datadrew shops" - "Switch to [shop name]" Revoking access - From Datadrew: click your profile avatar in the top-right and open Connected AI assistants, then click Revoke next to the client. You can also reach this screen from Settings → Datadrew MCP by clicking the Connected AI assistants card. - From your AI client: remove the Datadrew connector from its connector settings. For security details and how tokens work, see About the Datadrew MCP connection. Troubleshooting Connector added but tool calls fail. Make sure the integration you are asking about (e.g. Meta Ads) is connected in Datadrew under Integrations. The AI assistant can only reach integrations you have connected. Signed out after a while. Access tokens expire hourly and are auto-refreshed by your AI client using a 30-day refresh token. If you go 30 days without using the connector, you'll be asked to sign in again. Need help? Reach out at support@datadrew.io or use the in-app chat.

Drew AI launch — your AI growth analyst

Drew AI Launch — Your AI Growth Analyst We are excited to introduce Drew AI, your dedicated AI-powered growth analyst built directly into Datadrew. Drew AI lets you ask natural language questions about your e-commerce data and get instant, data-backed answers. What Is Drew AI? Drew AI is a conversational analytics assistant that understands your Shopify store data, ad performance, customer behavior, and more. Instead of navigating dashboards and building reports manually, you can simply ask Drew AI a question in plain English. Examples of what you can ask: - "What was my ROAS on Facebook Ads last month?" - "Which products have the highest repurchase rate?" - "Show me my LTV by acquisition cohort for Q4." - "How did my Google Ads campaigns perform compared to last quarter?" - "What are my top customer segments by revenue?" - "Compare my email revenue from Klaviyo vs. paid ad revenue." How It Works Drew AI is powered by advanced language models and has access to: - Your data warehouse: Drew AI can write and execute SQL queries against your store's analytics data in our data warehouse — the same data that powers your dashboards. This includes orders, customers, products, LTV cohorts, RFM segments, and ad metrics. - Live ad platform data: Drew AI can pull real-time data from your connected Facebook Ads, Google Ads, GA4, Klaviyo, and Google Search Console accounts for up-to-the-minute metrics. - Domain expertise: Drew AI is equipped with specialized knowledge about e-commerce analytics, retention analysis, ad optimization, and customer segmentation — so it provides context-aware answers, not just raw numbers. AI Models Drew AI offers multiple modes to match your needs: | | | | | --- | --- | --- | | Mode | Best For | Credit Usage | | Light | Quick, straightforward questions | Lower | | Pro | Standard analytical questions | Medium | | Deep Analysis | Complex, multi-step investigations | Higher | Visual Outputs Drew AI does not just give you text answers. It can generate: - Charts and graphs — bar charts, line charts, and more to visualize trends. - Data tables — structured tables with the exact numbers behind the analysis. - Summaries — clear, actionable insights with recommendations. Getting Started 1. Open Datadrew from your Shopify admin. 2. Click Drew AI in the sidebar. 3. Type your question in the chat input. 4. Drew AI will analyze your data and respond within seconds. Drew AI is available on all plans. Start asking questions today. Need help? Contact us at support@datadrew.io. Related articles - Release notes and product updates - What is Drew AI and how does it work? - Asking your first question to Drew AI - Understanding AI model tiers Need help? If you have questions or run into issues, reach out to us at support@datadrew.io or use the in-app chat. We're happy to help.

Release notes and product updates

Release Notes and Product Updates Stay up to date with the latest improvements, new features, and fixes in Datadrew. We update this page regularly as we ship new functionality. March 2026 March 1, 2026 — GMV-Based Pricing Launch - New pricing model: Essentials and Pro plan pricing now scales based on your store's annual Gross Merchandise Volume (GMV), ensuring fair pricing that grows with your business. - Updated plan structure: Three clear tiers — Free, Essentials, and Pro — with distinct feature sets for each stage of growth. - AI credit system: Drew AI now uses a credit-based system (Free: 200, Essentials: 1,000, Pro: 4,000 credits/month) for transparent usage tracking. February 2026 February 20, 2026 — Drew AI Enhancements - Live ad data: Drew AI can now pull real-time data from your Facebook Ads and Google Ads accounts using MCP tools, in addition to querying your data warehouse. - Model selection: Choose between Light (fast, lower credit cost) and Deep Analysis (thorough, higher credit cost) modes depending on your question complexity. - Chart generation: Drew AI can now generate visual charts and tables directly in the conversation. February 10, 2026 — Hourly Performance Tracking - New dashboard: Hourly performance tracking is now available for Essentials and Pro plans, letting you see intra-day trends in sales, traffic, and ad performance. - Hourly syncs: Essentials and Pro plans now sync data hourly instead of daily for more up-to-date dashboards. January 2026 January 15, 2026 — Klaviyo Segment Sync - RFM to Klaviyo: You can now sync your Datadrew RFM segments directly to Klaviyo for targeted email campaigns. Tag your best customers, at-risk customers, or any custom segment and push it to Klaviyo automatically. January 5, 2026 — Google Search Console Integration - New integration: Connect Google Search Console to see your organic search performance — queries, impressions, clicks, average position, and CTR — alongside your other analytics data. This page is updated as new features ship. For questions about any update, contact us at support@datadrew.io. Related articles - Drew AI launch — your AI growth analyst - Requesting a new feature - What is Drew AI and how does it work? - Feature comparison across plans Need help? If you have questions or run into issues, reach out to us at support@datadrew.io or use the in-app chat. We're happy to help.

New Feature Alert - Industry Benchmarks

As part of our ongoing commitment to empower our valued Shopify merchants with cutting-edge insights, we are thrilled to introduce our latest feature: Industry Benchmarking. You can know compare your key performance metrics with that of the competitors in the same industry and selling to similar customers in same/similar currencies. Some of the KPIs include - 1. Overall growth rates(Revenue, New customer, Order) last month compared with competitors 2. Check whether your customer lifetime value(LTV) grow as your top 25%, Average or Bottom 25% of your competitors. 3. How the cumulative repeat rate from customers increases over their 1st year. Check Industry insights Uncover opportunity areas where you perform better than competitors. Leverage this competitive edge to drive growth and retention. Also, Identify the problem areas which need founder’s attention and intervention. Set realistic and impactful goals based on achievable benchmarks to grow to be among the top players in your industry.