Accurate metrics for your AI assistant.
Geckoboard MCP gives you real-time business metrics from tools like HubSpot, Aircall, and Google Analytics. Use it to analyze changes, automate workflows, and build your own tools on data you can trust.
// the problem
AI assistants struggle with metrics
Today’s assistants can read, reason, and act across your whole business. But ask one to calculate a metric and it stumbles. The MCPs from tools like HubSpot and Zendesk were built to look up records, not compute them. So the assistant tries to work the metric out itself, and that’s where things go wrong.
Lookups work, metrics don’t
Ask how many deals you created this month and it pulls the answer straight from the database. Counts and totals are easy. But win rate, conversion rate, median response time? Those are calculations the query tools can’t return.
So it fails silently
And you won’t know. Left to crunch raw records on its own, it works off the first few hundred out of thousands, or counts the wrong stage as won. It’s slow, because getting the answer right means paginating through every record. What comes back looks confident and complete. It’s wrong.
And the fix isn’t vibe-codable
Computing metrics correctly needs a real layer underneath: every metric calculated in advance, kept fresh as data changes, and checked for gaps across the whole dataset. That isn’t a prompt or a weekend script. It’s a serious engineering project.
// the solution
A metrics engine, not another raw connector
For over 15 years, we’ve helped teams turn messy business data into accurate metrics they can trust on a dashboard. Behind those dashboards is our metrics engine — it connects your CRM, help desk, call centre, and analytics, and turns the raw data into clean, consistent metrics. Geckoboard MCP opens that engine up to your AI assistant.
// what it’s for
Put your metrics to work
Ask a question, automate a routine, or build something entirely your own — all on the same trusted metrics.
// get started
How it works
Getting trustworthy metrics in front of your AI assistant takes minutes.
Connect your tools
Create a Geckoboard account and connect it to your integrations. We’ll then switch on MCP access for you (there’s currently a waitlist).
Enable Metrics MCP
Connect the Geckoboard MCP to Claude, ChatGPT or any MCP-compatible assistant in a few clicks.
Get started
Ask about your metrics, pull reports, create your own tools.
Anyone who uses metrics as part of their role will benefit from direct access to their data via their AI client — business leaders, analysts, and customer-facing teams who want quick answers in the flow of their work, alongside the dashboards they already rely on.
It comes into its own for the people who build and maintain the reporting layer: Ops, RevOps, data and analytics teams, internal tooling teams, and implementation consultancies. We also see strong fits in Customer Success and Support Ops, where real-time metrics drive day-to-day workflows.
HubSpot, Pipedrive, Zendesk, Intercom, Aircall, and Google Analytics, with more being added all the time. If you need one that isn’t there yet, tell us what you’d like to see.
Create a Geckoboard account, connect your integrations, and either fill in the form on this page or ask for access in the app. There’s currently a waitlist for MCP. Once you’re switched on, you connect Geckoboard to Claude, ChatGPT or any MCP-compatible assistant in a few clicks.
No. Pointed straight at a tool, an AI assistant only sees raw records and has to guess how each metric is defined. Geckoboard gives it pre-computed metrics with consistent definitions, so the answers match the dashboards your team already uses.
A native connector is great for pulling up individual records and ad-hoc lookups. Geckoboard is a metrics engine: it returns ratios like win rate and conversion rate, durations like response time and time to close (including medians), breakdowns by rep with real names, and this-period-vs-last comparisons. Raw connectors can’t reliably calculate those, so the two work well together.
Dashboards and Metrics MCP use the same metrics in different ways. A dashboard gives your team a shared view, with secure sharing and permissions handled for you, and it works straight out of the box. Building your own tool can take a long time and bring security risks of its own. Metrics MCP is better for asking a one-off question, automation, or building your own tool. In the future we expect most teams will want both.
Anything that supports MCP. That includes assistants like Claude and ChatGPT, build tools like Lovable and Replit, and automation platforms like Zapier and n8n. As more tools adopt the standard, they’ll work with Metrics MCP too.
No. Metrics MCP provides read-only access to your metrics. It can answer questions about your data, but it can’t edit your data, dashboards or account settings. We are working on MCP control for your dashboards though.
The MCP is available to try free for 14 days. It’s included on all Geckoboard plans during this pre-release phase. However, data limits may apply.