General Questions
What is the PeakMetrics MCP Server?
The PeakMetrics MCP Server is a secure connection that brings PeakMetrics narrative intelligence directly into AI assistants like ChatGPT, Claude, and Google Gemini. Instead of logging into PeakMetrics separately, you can ask questions and get insights within the AI tools you already use daily. Think of it as making PeakMetrics “available everywhere” rather than “accessed in one place.”Why did PeakMetrics build this?
We built MCP Server because we observed that teams were spending significant time switching between tools, exporting data, and reformatting insights. In fast-moving situations, especially crises, this friction slowed response times and led to decisions being made with outdated information. By embedding PeakMetrics directly into AI workflows, we eliminate that friction and ensure intelligence is always current and immediately actionable.What does MCP stand for?
MCP stands for Model Context Protocol. It’s an open standard that allows AI assistants to securely connect to external data sources. Major AI platforms like ChatGPT and Claude officially support it, making it a safe, standardized way to extend AI capabilities.Does this replace the PeakMetrics dashboard?
No. The dashboard remains the best tool for deep investigation, detailed analysis, and complex queries. MCP Server is designed for rapid insights, briefing generation, conversational analysis, and situations where you need intelligence without leaving your current workflow. Most customers use both: the dashboard for thorough investigation and MCP for quick access and routine tasks.Is this available to all PeakMetrics customers?
Yes, MCP Server is included with all PeakMetrics subscriptions at no additional cost. It’s a new way to access the platform you already have.With the MCP Server, what data can AI assistants access?
With the MCP Server enabled, AI assistants can bring in both data and analytical context from the PeakMetrics workspaces configured directly in your account. This includes outputs from Smart Categories, bot detection, filters, and other workspace-level analytics you’ve set up. Once a workspace is configured, the AI assistant can reference it whenever you ask a question, drawing on the latest available PeakMetrics outputs at the time of the request. Rather than analyzing raw data in isolation, the assistant uses the same structured signals and narrative context your team already relies on in PeakMetrics. This means you stay in control of what the AI sees through your workspace configuration, and you can always trace AI-generated insights back to the underlying analytics and mentions that informed them.Setup & Technical Questions
How long does setup take?
Setup takes approximately 5 minutes and is a one-time process. You’ll need to:- Enable MCP in your AI assistant’s settings
- Add the PeakMetrics server URL: https://mcp.peakmetrics.com
- When prompted, log into your PeakMetrics account
Which AI assistants work with PeakMetrics MCP Server?
You can use the PeakMetrics MCP Server in any AI assistant that supports custom MCP Servers (sometimes called Connectors or Apps). Currently that includes:- ChatGPT (Plus and Enterprise)
- Claude (Pro and Teams)
- Google Gemini (Advanced)
- Custom-built AI assistants
Do I need a special subscription to ChatGPT or Claude?
For ChatGPT, you need ChatGPT Plus or ChatGPT Enterprise. For Claude, you need Claude Pro or Claude Teams. The free versions of these AI assistants don’t currently support MCP connections.What is the MCP Server URL?
- Server Name: PeakMetrics (or whatever you prefer)
- Server URL: https://mcp.peakmetrics.com
Where do I get my credentials?
Your credentials are your unique PeakMetrics username and password. If you are locked out or need assistance accessing your account, contact support@peakmetrics.com.What if I lose connection?
You can disconnect and re-log in following the same instructions. Contact support@peakmetrics.com for questions.Can I use MCP Server on mobile?
Yes, if your AI assistant has a mobile app that supports MCP. ChatGPT and Claude both have mobile apps where MCP connections work once set up.Does MCP Server work offline?
No. MCP Server queries PeakMetrics in real-time, so you need an internet connection for it to work. The benefit is that you always get current data, not cached or outdated information.Usage & Capabilities Questions
What can I do with MCP Server that I can’t do in the dashboard?
MCP Server enables several unique capabilities:- Conversational queries: Ask questions in plain English without learning query syntax
- Combine data sources: Merge PeakMetrics intelligence with other data in the same analysis
- Workflow integration: Access intelligence without leaving your current task
- Automated routines: Set up recurring intelligence pulls and briefings
Can I ask questions in languages other than English?
Yes. You can ask questions in any language your AI assistant supports (which is dozens of languages for ChatGPT and Claude). PeakMetrics also analyzes content in many languages, and the AI can help translate and interpret that content for you.How fresh is the data?
Every query pulls live data from PeakMetrics. The data is as current as if you were looking at the dashboard at that exact moment. There’s no caching or staleness.Can I get historical data or is it only recent?
You can query any timeframe in your PeakMetrics data retention period (typically 12+ months). For example: “Show me narrative trends about our brand in Q3 2024” will retrieve that historical data.Can the AI create visualizations and charts?
Yes. AI assistants like ChatGPT and Claude can generate charts, graphs, and visualizations from PeakMetrics data. You can ask for specific formats: “Show me a line chart of narrative volume over the past 30 days” or “Create a bar chart comparing sentiment across top narratives.”What types of questions work best?
Effective questions are specific and actionable:- ✅ “What are the top 5 narratives about [our brand] in the past 48 hours?”
- ✅ “Generate a crisis brief on the emerging issue about [topic]”
- ✅ “Which accounts are amplifying the narrative about [topic], and do they show bot-like behavior?”
- ✅ “Compare sentiment trends for [our brand] vs. [our competitor] this month”
- ❌ “Tell me what’s happening”
- ❌ “Is everything okay?”
- ❌ “Give me a report”
Can I save queries for repeated use?
While MCP Server itself doesn’t save queries, you can save prompts in your AI assistant. For example, you might save a “Daily Threat Brief” prompt that you run every morning with one click. Most AI assistants have features for saving and organizing prompts.Can multiple team members use MCP Server?
Yes. Each team member gets their own authentication token and can set up their own MCP connection. Each person sees only the PeakMetrics data they’re authorized to access based on existing permissions.Security & Permissions Questions
Is MCP Server secure?
Yes. Security measures include:- OAuth-based authentication: Industry-standard security protocol
- Unique tokens per user: Every team member has their own access token
- Permission inheritance: Users only see data they’re authorized to access in PeakMetrics
- Encrypted connections: All data transmission is encrypted in transit
Can people access data they shouldn’t see?
No. MCP Server respects all existing PeakMetrics permissions. If a user can’t see certain workspaces, narratives, or data in the PeakMetrics dashboard, they won’t be able to access it through MCP Server either.Is data stored by ChatGPT or Claude when I use MCP?
AI assistants may temporarily cache responses for the duration of your conversation, but they don’t permanently store PeakMetrics data. Always check your AI assistant’s data handling policies for specifics. For enterprise customers, most AI platforms offer data residency and retention controls.Business & ROI Questions
How much does MCP Server cost?
MCP Server is included with all PeakMetrics subscriptions at no additional cost. However, you do need a paid subscription to the AI assistant (e.g., ChatGPT Plus, Claude Pro) to use MCP features.What’s the ROI of using MCP Server?
Teams report saving 10+ hours per week on routine intelligence tasks like:- Creating daily briefings (from 2 hours → 5 minutes)
- Generating crisis reports (from 30 minutes → 5 minutes)
- Analyzing narrative trends (from 1 hour → 5 minutes)
- Identifying key influencers (from 45 minutes → 5 minutes)
Will this reduce how many PeakMetrics licenses we need?
No. MCP Server makes PeakMetrics more valuable and increases usage, not decreases it. Teams find they use PeakMetrics more frequently because the friction is removed, which actually drives higher license utilization and often expansion as more teams want access.Can this help us onboard new team members faster?
Yes. New team members can start getting value from PeakMetrics without learning the dashboard first. They can ask questions in plain English and get answers, lowering the learning curve significantly. Most new users feel productive with MCP Server within their first hour. However, you will still need to set up your workspaces in the PeakMetrics dashboard before accessing the data in the AI assistant.Does this work for customer-facing use cases?
While MCP Server is designed primarily for internal team use, some customers are exploring customer-facing applications—like integrating narrative insights into client dashboards or reports. Contact your account team to discuss specific customer-facing use cases and compliance considerations.Comparison & Competitive Questions
How is this different from just asking ChatGPT about our brand?
Without MCP Server, ChatGPT only knows information from its training data (which has a cutoff date) and can’t access your PeakMetrics data. It might give you generic or outdated information. With MCP Server, ChatGPT pulls real-time, authoritative narrative intelligence directly from your actual PeakMetrics data—so insights are grounded in what’s happening now, not what happened months or years ago. (Good data = better decisions.) Unlike AI assistants that rely on sampled or inferred content from their training data, PeakMetrics provides at-scale analytics across the full universe of relevant narratives, sources, and mentions. This means you’re not seeing a partial snapshot or anecdotal signal—you’re getting a complete, data-backed picture of how a story is forming, spreading, and evolving across the information ecosystem. Just as important is trust and transparency. Because PeakMetrics delivers comprehensive analytics and full underlying mentions, you’re never forced to “take the AI’s word for it.” You can ask the AI assistant exactly where an insight comes from, review the source material yourself, and understand how conclusions were reached. The result is narrative intelligence that’s not only powerful—but explainable, auditable, and trustworthy.Other tools offer dashboards. What makes PeakMetrics MCP unique?
Most media monitoring and social listening tools are still “dashboard-first” products. You have to go to them to get insights. PeakMetrics MCP Server is “workflow-first”—intelligence comes to where you’re already working. This is a fundamental shift in how narrative intelligence is consumed.Can I use multiple MCP servers at once?
Yes. Most AI assistants allow multiple MCP connections. You could connect PeakMetrics alongside other data sources (e.g., your CRM, sales data, internal knowledge base) and ask questions that combine insights from all of them. This is where MCP becomes especially powerful.Is PeakMetrics the only company offering MCP?
No. MCP is an open protocol, and many companies are building MCP servers for their platforms. What makes PeakMetrics unique is our narrative intelligence, the depth of media analysis, bot detection, influence mapping, and threat identification that we provide. MCP is the delivery mechanism; the value is in the intelligence itself.Troubleshooting & Support Questions
What if my query returns an error?
Common reasons for errors:- Authentication issue: Your token may have expired or been regenerated. Try reconnecting.
- Permission issue: You may be asking about data you don’t have access to. Check your permissions in PeakMetrics.
- Syntax issue: The AI may have misunderstood your question. Try rephrasing more specifically.
- Timeout: Very complex queries may time out. Try breaking them into smaller questions.
The AI’s response seems incomplete. What should I do?
Ask the AI to elaborate or provide more detail. Sometimes AI assistants provide high-level summaries first and can dig deeper when prompted. You can say things like:- “Give me more detail on narrative #3”
- “Expand on the bot activity analysis”
- “Show me the top 20 instead of top 10”
Can I share MCP Server responses with colleagues who don’t have MCP set up?
Yes. You can copy and paste responses, export them, or share screenshots. However, to run queries themselves, colleagues will need to set up their own MCP connection (which takes 5 minutes).What if I want a feature that doesn’t exist yet?
We’re actively developing MCP Server based on customer feedback. Submit feature requests through your account team or directly to support@peakmetrics.com. We prioritize features based on customer demand and feasibility.Where can I get help if I’m stuck?
- Check out our MCP Server Documentation
- Support team: support@peakmetrics.com
- Your account team: Contact your CSM for personalized guidance
Does PeakMetrics offer training on using MCP Server effectively?
Yes. Your customer success team member will schedule training.Advanced Usage Questions
Can I automate queries to run on a schedule?
This depends on your AI assistant. Some enterprise AI platforms support scheduled execution of prompts. Alternatively, you can use API-based automation with tools like Zapier, n8n, or Make if your AI assistant offers API access.Can I build custom applications on top of MCP Server?
Yes, if you are building a custom AI agent that would benefit from up-to-date narrative intelligence and threat detection, the PeakMetrics MCP Server is a perfect fit for you. Please contact your account team for technical documentation and developer resources.What’s the query limit or rate limiting?
MCP Server is designed to handle typical analytical workloads without interruption. For very high-volume automated queries (e.g., hundreds per hour), there are fair-use limits. Contact your account team if you have an advanced use case that might exceed standard usage patterns.Can we white-label or embed MCP functionality in our own platform?
For enterprise customers with specific integration needs, we can discuss custom solutions. Contact your account team to explore options.Still Have Questions?
Contact us:- Support: support@peakmetrics.com
- Sales: sales@peakmetrics.com