> ## Documentation Index > Fetch the complete documentation index at: https://www.conductor.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # Platform: Power Your AI Tools with Conductor Data (MCP) > You have the data: brand mentions, citation share, topic performance, keyword rankings. **Conductor customers:** [Review this lesson in our Community](https://app.conductor.com/u/community?ms_url=%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fstep%2F2cd14a8c-7c35-4499-9b07-8bdd6eaea93f) ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FcJhZtxOIut-updated_course_headers__9_.png) You have the data: brand mentions, citation share, topic performance, keyword rankings. Conductor's MCP puts it to work in two ways at once. It opens that data up to team members who don't live in the platform every day, and it gives power users a fundamentally different way to work with it: conversational, fast, and connected to whatever AI tool they're already in. This course walks you through everything you need to start using it β€” what MCP is, what data you can access, how to get connected, and how to use it to answer the questions your team asks every day. No engineering degree required. By the end, you'll know how to: * Understand what MCP is and why it matters for your workflow * Connect Conductor to the AI assistant your team already uses * Query your Conductor data in plain English β€” no dashboards, no filters * Turn data into deliverables without leaving your AI tool * Use pre-built Skills to run repeatable analyses in seconds ⏲️ This course takes approximately 30 minutes to complete. Each section is self-contained, so you can move at your own pace or come back to individual sections as a reference. Ready? Let's start with the basics. 🚨Note: Please make sure to click the "start course" button at the bottom of the course and the "complete course" button at the end to get credit for completing the course 🚨 **Conductor customers:** [Review this lesson in our Community](https://app.conductor.com/u/community?ms_url=%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fstep%2F6cf72914-2203-4302-842a-dc9e7fc91129) Let's start with a problem every AI user eventually runs into. You're in ChatGPT or Copilot, and you ask something specific about your brand, your customers, your performance data. The AI responds with something that sounds reasonable, confident, even insightful. But it's not working from your actual data. It's working from general knowledge about the world, which may be months old and has nothing to do with your specific situation. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2F3I4rV5siqR-1.png) This isn't a bug. It's how AI tools are built. They're trained on vast amounts of general knowledge, but they have no live window into your tools, your dashboards, or your data. Unless you build them one. That's what an MCP server is. MCP β€” Model Context Protocol β€” is an open standard, created by Anthropic and now used industry-wide, that defines how AI assistants connect to external data sources. Any company that wants to make their data available inside an AI conversation can build an MCP server. When your AI tool connects to one, it can query real data instead of guessing. The AI tool you're using β€” Claude, ChatGPT, Copilot, Perplexity β€” is the host: where the conversation happens. The MCP server is the adapter that lets that AI reach into an external data source and pull back real information. The protocol defines how the two talk to each other. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FU70mpjKB0W-tangled_wires.png) That's all you need to know about MCP as a concept. What matters is what a specific company chooses to build with it. Conductor built one for your search and AEO performance data. Let's talk about what that unlocks. Want to dive deeper? See[What is an MCP Server? The Universal AI Adapter](https://www.conductor.com/academy/mcp-server/) **Conductor customers:** [Review this lesson in our Community](https://app.conductor.com/u/community?ms_url=%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fstep%2Fa8e960ed-ea5f-49cf-a233-6a35e877df7a) Conductor's MCP server connects your AI assistant β€” whichever one you use β€” directly to your Conductor data. When you ask a question about your brand's visibility in AI search, your citation trends, or how competitors are showing up in AI responses, the answer comes straight from Conductor's data. Not a general estimate, not a scraped summary. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2Ftv0OTaOY2D-mcp_core_difference.png) #### Why this is a bigger deal than it sounds If you've ever asked an AI tool about your brand's search performance and gotten a confident-sounding answer, there's a good chance that answer was made up. Not out of malice, but structurally: AI language models don't have access to Conductor. They're guessing from general knowledge about your industry, which may be months old and has nothing to do with your specific account. Conductor's MCP fixes that foundation. Questions about your performance get routed to Conductor's API, which returns verified, structured, real-time data. This matters especially for AEO and AI search, where the landscape shifts week to week. You want answers based on last week's data, not what a model learned in a training run a year ago. But accurate data is only part of it. The bigger unlock is what you can do with that data without leaving your AI tool. Ask a question, follow up, go deeper, get a recommendation, ask the AI to turn your findings into a brief or a deck β€” all in the same thread. With Skills, your most-used analyses go from a multi-step conversation to a single command. That's the shift from data access to an end-to-end workflow. #### What's available through Conductor's MCP Through Conductor's MCP, your AI assistant has access to three categories of data: AI search data This covers brand mentions, citations, and sentiment across Conductor's supported AI search engines. * Brand data tells you who's showing up in AI-generated answers and how often. * Citation data tells you which URLs and domains AI engines are actually trusting as a source, not just mentioning in passing. * Sentiment data captures the quality of those mentions: positive, neutral, or negative, broken down by category (price, quality, ethics, experience, and more). Together, these give you a full picture of how your brand is perceived across AI search. Traditional search data This includes keyword rankings, search volume trends and seasonality, SERP feature breakdowns (Standard Links, People Also Ask, Local Pack, etc.), head-to-head competitive rankings, and single-keyword deep dives β€” everything you need to understand where you rank, how you're trending, and how you stack up against competitors. Account configuration data Your account metadata: the topics, prompts, brands, competitors, personas, intents, and search engines you're tracking. This is what helps the AI resolve your questions accurately β€” so "how's my UK brand doing?" maps to the right data rather than a guess. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FT4B6nL8lHl-2.png) Technically, the MCP connects your AI to a set of tools, each defining something the AI can ask for or do. Tools can surface data, trigger actions, or generate outputs depending on how they're built. In Conductor's MCP today, the available tools primarily surface your search and AEO data: brand mentions, rankings, account configuration, and more. New data types and endpoints are being added regularly. For the full breakdown of what's included in each category β€” including any platform-specific availability notes β€” see the Conductor Developer Experience:[What data is available in Conductor's MCP?](https://www.conductor.com/docs/mcp/what-data-is-available-in-conductors-mcp/) #### What the MCP unlocks for Conductor power users The MCP changes how you work with data, not just how fast you get to it. Instead of pulling reports and switching between tabs, you're having a conversation that goes from data to insight to strategy to action in a single thread. You can chain questions, cross-reference AI search performance with keyword rankings, and ask the kind of follow-up questions that usually require a whole new report. Then ask the AI to turn it all into a strategic analysis, quarterly content plan, or polished performance deck β€” without ever leaving your tool. The depth of analysis you can run in a 10-minute MCP session would take hours to replicate manually. #### What the MCP opens up for the rest of your team Conductor data has historically lived behind a login that only a handful of people regularly open. The MCP removes that bottleneck. Anyone in a connected AI tool can now pull real Conductor data, no platform login required. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FgxsbAnktC0-who_gets_access.png) Here's what that looks like across different roles: * Content teams can query which topics competitors are being cited for that they're not covering, get a prioritized list of what to write or optimize, and pull it all into a content brief. No platform training needed. * Brand and PR can track sentiment before a campaign goes live, catch early signals on reputation shifts by category, and identify which domains are driving negative mentions β€” in plain English, whenever they need it. * Leadership can get a 30-second verdict on AI search performance before a board meeting, or walk out of an AI tool with a full executive-ready deck built from live data. No report request needed. These are just a few examples. The range of what's possible depends on your role, your data, and the questions you're asking. For more, visit the[Conductor MCP use cases page](https://www.conductor.com/academy/conductor-mcp-server-use-cases/). **Conductor customers:** [Review this lesson in our Community](https://app.conductor.com/u/community?ms_url=%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fstep%2F1c3f7fe5-23b9-499c-9eba-066f25f78412) Before we get into setup, it's worth picturing what you're actually getting access to. Technically it's a new connection; functionally, it changes how your whole team can work with Conductor data. Here's a tour of what becomes possible once you're connected. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FZFNBhQGmQi-5_ways_to_put_conductor_to_work.png) #### AI search: brand mention analysis Brand mention analysis tells you how and when your brand is showing up in AI-generated answers, and how that compares to competitors across your tracked topics, personas, and search intents. Think of it as share of voice for the AI layer of search. Often, the most interesting finding isn't where you're dominating, but where you're showing up without leading. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FleHi19VgRj-ai_search__brand_mention_analysis_screenshot.png) #### AI search: citation analysis Brand mentions and citations are not the same thing. A mention means an AI engine referenced your brand; a citation means it trusted your content as a source. That gap might be where content strategy has the most room to move. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FiNkl29vKDl-mentions_vs_citations.png) #### AI search: sentiment analysis Beyond whether your brand is mentioned, sentiment analysis tells you how it's being talked about β€” positive, neutral, or negative β€” broken down by category: price, quality, ethics, customer experience, and more. For brand and PR teams, it's the earliest signal you'll get on a reputation shift. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FVA_-S7mhWi-ai_search__sentiment_analysis_screenshot.png) #### Traditional search: rankings, competition, and seasonal trends For teams with traditional search data connected, the same conversational approach applies to keyword performance: rankings, SERP features, competitive comparisons, seasonal demand. All in plain English, in one thread. The kind of analysis that used to mean pulling multiple reports now happens in a single conversation. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FbKARIFsTGF-traditional_search__rankings__competition__and_seasonal_trends_screenshots.png) #### Account configuration: pressure-testing your tracking strategy One underrated use case: asking the AI to review your Conductor configuration and flag what might be missing. Before a strategy review or QBR, one question is all it takes to pressure-test your coverage. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2F0pb7gUaGzA-account_configuration__reviewing_your_tracking_strategy_screenshot.png) Now that you know what's possible, let's get you connected. The next section walks through the setup process for each AI platform. **Conductor customers:** [Review this lesson in our Community](https://app.conductor.com/u/community?ms_url=%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fstep%2F3b3ffa56-f66c-4606-b81c-f5c941125222) Good news! Access to Conductor's MCP server is included in every customer's contract. No additional purchase required. If you're a Conductor customer, you're already in. MCP is also inherently LLM-agnostic, meaning Conductor's MCP server can connect to any AI tool that supports the protocol, not just the four listed below. The setup guides currently available cover the most widely used platforms, and more will be added as the feature evolves. Pick the assistant you use and follow its setup guide from the[Conductor Developer Experience](https://www.conductor.com/docs/): * [ChatGPT](https://www.conductor.com/docs/mcp/chatgpt) * [Claude](https://www.conductor.com/docs/mcp/claude/) * [Copilot](https://www.conductor.com/docs/mcp/copilot/) * [Perplexity](https://www.conductor.com/docs/mcp/perplexity/) Using a different AI tool? If your team works in a platform that isn't listed above, you can likely still connect β€” check out[Conductor's custom connectors guide](https://www.conductor.com/docs/mcp/custom-connectors) or reach out to your Conductor team and we'll help figure out what's possible. At a high level, each guide walks you through: 1. Verifying your plan and permissions 2. Adding a connector pointed at Conductor's MCP URL 3. Authenticating with your Conductor account Once you've enabled the connector, your assistant can call Conductor's[MCP tools](https://www.conductor.com/docs/mcp/what-is-an-mcp-tool/) on your behalf. **Conductor customers:** [Review this lesson in our Community](https://app.conductor.com/u/community?ms_url=%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fstep%2Fee57c516-556b-4fc3-bf7f-8881369ee5ee) You're connected. Now comes the part that actually matters: learning to have a productive conversation with the MCP. Before you ask about data, ask about yourself. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2F-VeX9LyU69-how_to_query_data.png) #### Step one: tell the MCP who you are Whatever your role, your Conductor experience, or your history with AI tools, the single best first move is this: `I'm a [your role] at [your company]. We use Conductor for our search and AI search performance data. How can the Conductor MCP help me? What kinds of questions should I be asking?` Let the AI answer that before you do anything else. This works because the MCP knows what data it has access to, and when you tell it your role and goals, it maps those two things together and gives you a personalized starting point. A content strategist, a PR manager, and a VP will each get different answers to the same prompt β€” and all three will be genuinely useful. You don't need to already know what to ask. The MCP figures that out with you. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FIzERCHwcY5-step_one__tell_the_mcp_who_you_are_screenshot.png) #### Step two: orient yourself on your data Once you have a sense of how the MCP can help you, take a moment to understand what Conductor data is available for your account. All of this surfaces through the MCP, no platform login needed: `What topics are we currently tracking in Conductor?` `What competitors are we tracking?` `What time ranges does my data cover?` `I'm working on a project about [X] β€” are we tracking anything related to that?` `Show me the tools and data I have available via Conductor.` This is especially useful if you're not a regular Conductor user and aren't sure what's been set up. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FciApgce39v-step_two__orient_yourself_on_your_data.png) #### The vocabulary that gets you better answers Bookmark this. The AI routes questions to the right data based on the words you use. Conductor-specific language gets precise answers; vague language gets vague ones. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FIGY5zf3uuw-conductor_vocabulary_cheat_sheet.png) Always name competitors specifically. "How do I compare to my competitors?" is hard to act on. "How does my brand mention share of voice compare to \[Competitor A] in the past 90 days?" gets you an answer. #### How to structure a good question Good queries follow a simple pattern: what you want to know + the data dimension + any relevant filter. ![dynamic content](https://storage.googleapis.com/storage-prod-public-xfr99i/spot%2Feb48ced0-dfb7-4d21-bc33-7f990b8ba150%2Fcourses%2Fcecdf0c2-0196-4510-9872-6810b415e537%2Fshared%2FAtzSJBUluX-vague_vs__specific_prompt_comparison.png) #### Iterate β€” it's a conversation, not a search One of the biggest advantages of the MCP is that you can follow the data wherever it leads. Start broad, then go deep on whatever's interesting. 1. `What topics are driving the most brand mentions for my site?` 2. `For that top topic, how does my share of voice compare to [Competitor]?` 3. `Which specific queries within that topic am I losing ground on` 4. `What does my sentiment look like on those queries?` Each question builds on the last.