You run A/B tests. You collect data. But getting answers from that data still means clicking through dashboards, filtering date ranges, and cross-referencing results across tests.
What if you could just ask?
Today, we're launching Optibase's MCP Server — a direct integration that connects your Optibase workspace to AI assistants like Claude and Cursor. Ask questions about your A/B tests, heatmaps, traffic, and conversions in natural language. Get answers from your actual data in seconds.
No dashboards. No exports. Just ask.
The Model Context Protocol (MCP) is an open standard that lets AI assistants connect directly to your tools and data. Think of it as a universal API for AI — instead of your assistant guessing or working from screenshots, it pulls live data from your actual Optibase workspace.
For CRO teams, this changes the workflow entirely. Instead of switching between tabs, building reports, and interpreting charts, you stay in one place and ask questions like you would ask a colleague who has all your data memorized.
Once connected, your AI assistant has access to your full Optibase workspace. Here's what that unlocks:
Skip the dashboard. Ask directly:
The assistant pulls live results — conversion rates, visitor counts, statistical significance — and gives you a clear answer.
The assistant can fetch your actual test pages and analyze what each variant looks like. Then it can suggest improvements based on the performance data:
This is where AI-powered A/B testing gets practical — not generating random ideas, but making recommendations grounded in your real test data.
No more digging through analytics tabs:
This is the real unlock. Combine multiple data points and ask for high-level guidance:
Instead of a dashboard that shows you numbers, you get an analyst that interprets them.
https://my.optibase.io/api/mcpOptibase, Type: HTTP, URL: https://my.optibase.io/api/mcp
Any client that supports the MCP protocol works. Same URL, same OAuth login flow.
A/B testing tools have always been good at collecting data. The bottleneck has been acting on it. Teams run tests, results sit in dashboards, and decisions get delayed because someone needs to pull the numbers, interpret statistical significance, and write up a recommendation.
MCP removes that bottleneck. Your test data becomes conversational — accessible to anyone on the team who can type a question. That means faster decisions, shorter test cycles, and more experiments shipped.
This is what AI-powered CRO actually looks like: not AI running your tests for you, but AI making your test data instantly useful.
The MCP integration is available now on all Optibase plans. Connect your workspace in under two minutes and start asking questions about your A/B tests, heatmaps, and conversions.
MCP is an open standard that lets AI assistants like Claude and Cursor connect directly to external tools and data sources. It means your AI assistant can pull real data from Optibase instead of working from context you paste in.
Claude Desktop and Cursor are the primary supported clients. Any MCP-compatible client can connect using the same server URL and OAuth authentication flow.
No — the MCP integration is available on all plans, including Free.
Yes. Authentication is handled via OAuth. The AI assistant can only access the workspace you explicitly authorize during setup.
No. The MCP integration is read-only. It can query your data and provide analysis, but it cannot create, edit, or delete tests.
Optibase is the experimentation platform built for Webflow (and WordPress). Run A/B tests, split tests, and multivariate experiments with heatmaps and session recordings included — starting free. Try Optibase →
