Helping product teams build, ship, and improve at scale, with AI built into every step.
Every company is trying to do the same thing: build a better product and move faster than their competitors. The way to do that is to learn faster, so more decisions are made from evidence instead of guesses.
But learning faster takes more than the ability to run an A/B test.
Your experimentation platform should help you run experiments quickly and rigorously, capture what you learn, and identify opportunities to apply those learnings to the next decision.
GrowthBook 5.1 brings the entire learning loop together, with AI across every step. It’s now available to all cloud and self-hosted users. Check out the full release notes for everything that shipped.
Scale testing without sacrificing rigor
GrowthBook 5.1 helps teams scale experimentation, both in volume and in what you can test, like measuring funnel completion or optimizing with contextual bandits. AI-powered workflows remove friction and built-in rigor ensures you can move faster without sacrificing confidence in your results.
Contextual banditspersonalize which variation a user sees at the individual user level, learning which variant performs best based on their context and automatically shifts traffic accordingly. They’re especially useful for time-sensitive campaigns, high-variant tests, or personalization where the best experience varies by user.
Funnel metrics let you measure funnel completion in your experiments and see where users are dropping off. They are especially useful for testing marketing funnels, onboarding flows, and other multi-step experiences.
AI lets you scale experimentation by turning the workflow into something fully programmable, giving agents the ability to publish flags, create experiments, and more. Our new Hosted MCP + OAuth server, and an expandingAI skills library of 30+ skills make it faster and easier for teams to build and run experiments.
See GrowthBook 5.1 in action
Join a live demo to see the MCP server, Slack app, and contextual bandits working together in a real experiment workflow.
Every experiment creates knowledge, even the losses. The challenge is making sure that knowledge doesn’t disappear into static documents, Slack threads, or the heads of the people who ran the experiment.
GrowthBook 5.1 helps capture what you learn and makes it easier to analyze more data with AI, uncover deeper insights, and share them across your organization.
Learnings gives you a living knowledge library to document everything you’ve discovered through experimentation, so your entire organization can access and build on it. GrowthBook’s agent can automatically scan your past tests to get you started or you can use our new API endpoints and agent skills.
The Slack App (Beta) brings your GrowthBook data into the conversations where your team is already working. Tag @growthbook to ask questions about your data or interact with the GrowthBook AI assistant directly from Slack. You can also send rich, customizable notifications and daily or weekly digests to keep your team informed about what’s happening across your projects and organization.
Product Analytics closes the loop
GrowthBook’s Product Analytics became generally available in 5.0, and 5.1 builds on this foundation making it easier to understand how users move through your product, where they encounter friction, and identify opportunities to test. This gives a clearer view of where to focus your experimentation efforts next.
User Journeys help to understand what happens before and after key moments in the user experience. Pick a starting or ending step so you can see the actions users take immediately before and after. Follow users deeper into your product or switch over to Funnel Explorer to do a deeper dive.
With a clearer view of how users navigate your product, you can spot friction, uncover opportunities, and make more informed decisions about what to test next.
AI-powered dashboards leverage AI skills to turn plain-language prompts into shareable product analytics or experiment dashboards. Describe the views you need through GrowthBook’s in-app assistant, MCP server, or wherever your agents live, and get the dashboard you need in moments, ready to share org-wide.
We rebuilt SQL Explorer to be easier to use for both humans and agents, so you can do richer explorations across your entire data warehouse without needing to predefine metrics. Use GrowthBook’s AI assistant or your preferred agent to query your data directly and ask questions that don’t map to existing metrics. It's another tool in the agent's kit for deeper, more flexible data exploration.
The full loop with GrowthBook 5.1
Together, these capabilities turn experimentation from a testing function into a continuous learning loop for the entire company.
Understand your product. Run more experiments. Learn from every result. Use what you learn to make better decisions. Then do it again.
.avif)
.avif)
