> For the complete documentation index, see [llms.txt](https://docs.growthfactor.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.growthfactor.ai/getting-started/mcp.md). # MCP The GrowthFactor MCP lets you use GrowthFactor's site-evaluation tools directly inside the AI assistant you already use — Claude (Desktop, Web, Code), Cursor, or any MCP-compatible client. Instead of switching tabs into the GrowthFactor app, you ask Claude things like "score this address" or "rank my pipeline" and it pulls real data from your GrowthFactor workspace. *** ## What you can do with the MCP Every figure the assistant cites comes from a real tool call against your account. The MCP uses the same tools that power the GrowthFactor app: | You can ask about… | What it does | | ---------------------- | ------------------------------------------------------------------------------------ | | **Scoring a location** | GrowthFactor Score (0–100) with the lens breakdown and explanations | | **Demographics** | Population, income, age, education for a trade zone (drive-time, walk-time, or ring) | | **Foot traffic** | Pedestrian patterns at the exact point | | **Analogs** | Similar stores in your own portfolio to benchmark against | | **Cannibalization** | Trade-zone overlap with your existing sites | | **Your portfolio** | List, view, create, and update sites | | **Your pipeline** | List, view, create, and update deals | | **Reports** | Generate and retrieve site evaluation reports | | **ML predictions** | Run any sales-prediction model your workspace has configured | ## MCP vs. the Growthfactor Agent The MCP and the in-app [Growthfactor Agent](/analyze-site-intelligence-and-selection/growthfactor-agent.md) run on the same analytics tools, so they share the core data. The difference is where they run and what each is best at. **In most cases, build your strategy in the Agent.** It's where you draw and combine trade zones into custom territories, find your whitespace, drop and compare pins live on the map, and annotate it all for your team — and it remembers your work across sessions. **Reach for the MCP** when you want GrowthFactor data inside the AI tool you already use, most often to turn a finished strategy into a committee-ready deck. | Feature | MCP | Agent | | --------------------------------------- | ------------------------------------------------------------- | ----------------------------------------------------------------------------------- | | **Where it runs** | Your AI client (Claude, Cursor, ChatGPT) | Inside the GrowthFactor app | | Demographics, traffic, scoring, analogs | ✓ | ✓ | | Sites & deals (list, create, update) | ✓ | ✓ | | Compute a trade zone | ✓ | ✓ | | Sales-forecast reports | ✓ | ✓ | | **Build & combine trade zones** | — | Union, intersect, and subtract zones into custom territories | | **Whitespace / gap analysis** | — | Subtract existing coverage to find uncovered demand | | **Cannibalization** | Single point vs. your portfolio | Pin-based — compare a pin against every other pin, area, and site, drawn on the map | | **Live market-planning visuals** | Returns data to your client | Drops pins, draws zones, and renders the full map and tables as you plan | | **Annotate & share the map** | — | Mark up, comment on, and rename study elements for your team | | **Memory** | Stateless — each call starts fresh | Remembers you and your team across conversations | | **Committee-ready decks** | Pipe GrowthFactor data into your AI client to assemble a deck | — | | **Omnichannel access** | — | On the roadmap | {% hint style="info" %} **Public beta** — the Agent's Market Planning capabilities (building trade zones, whitespace, on-map cannibalization, and annotation) are in public beta and improving quickly. {% endhint %} **Suggested workflow:** build your strategy in the Agent — draw your territories, run cannibalization and whitespace analysis, and shape your market plan visually — then use the MCP to turn it into a committee-ready deck in your AI client. ## Supported clients The MCP works with any MCP-compatible client. The most common are: * **Claude Code** * **Cursor** * **ChatGPT** ## Connecting GrowthFactor to your MCP client * [ChatGPT Setup documentation](https://help.openai.com/en/articles/11487775-apps-in-chatgpt) * [Claude Setup documentation](https://support.claude.com/en/articles/11176164-use-connectors-to-extend-claude-s-capabilities) ### Authentication The MCP uses your existing GrowthFactor login. When you add the server, your client opens a browser tab and prompts you to sign in. Approve the connection and you're done. Re-auth happens automatically when your session expires. You can revoke the connection at any time from your AI client's connector settings. ### Configuration The server URL is the same everywhere: `https://mcp.growthfactor.ai/mcp`. The steps differ by client. #### Claude Desktop and Claude.ai 1. Open **Settings** {% columns %} {% column %} 2\) Select **Connectors**
{% endcolumn %} {% column %} 3\. Click **Add custom connector** at the bottom of the page ![](/files/tUuTfkYPbOT7n18RZtEz) {% endcolumn %} {% endcolumns %} 4. Click **Add custom connector** at the bottom of the page 5. Name it **GrowthFactor** and paste `https://mcp.growthfactor.ai/mcp` as the URL 6. Save. Claude opens a browser tab to GrowthFactor — sign in with your usual login and approve the connection. #### Claude Code 1. Add the server by running the following in your command prompt or terminal: ```bash claude mcp add --transport http growthfactor https://mcp.growthfactor.ai/mcp ``` 2. Start the authentication flow by running `/mcp` inside Claude Code and selecting **GrowthFactor.** 3. Claude Code opens a browser tab to GrowthFactor — sign in with your usual login and approve the connection. You'll see `Authentication successful. Connected to growthfactor.` in the terminal once it's done. *** ## Example prompts Open a chat in your AI client and ask the assistant directly. A few starting points: * "Score 123 Main St, Austin with a 10-minute drive-time trade zone." * "Pull demographics and analogs for the corner of Newbury and Dartmouth in Boston." * "Which 5 deals in my pipeline have the best foot traffic? Rank them." * "Where should I open next in Nashville?" — the assistant will pick candidate corridors, score each, and compare them in a table. ## Workspace context If you belong to more than one workspace, the assistant will ask which one to use. Say "use the Acme workspace" once and it'll stick for the rest of the conversation. The MCP respects the **active workspace's** Site Score lenses, default trade zone, tracked brands, and brand-specific model defaults — the same context the in-app [Growthfactor Agent](/analyze-site-intelligence-and-selection/growthfactor-agent.md) uses. ## Privacy and permissions * Access uses your existing GrowthFactor login and respects your workspace permissions. If you can't see a site in the app, the assistant can't see it either. * **Write actions** (create deal, update site, apply tags) require you to specify the exact target and value. The assistant won't experiment with your data. * You can revoke the connection at any time from your AI client's connector settings. ## Tips for good results * **Be specific about trade zones** ("10-minute walk", "2-mile ring") or let the default apply — the assistant will tell you which zone it used. * **Tell it your intent** ("I'm scouting for a second location near my best store") — it tailors the analysis and helps us improve the product. * **For broad asks** ("where in Boston?"), the assistant proposes candidate areas first and waits for your go-ahead before running a full scan. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.growthfactor.ai/getting-started/mcp.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.