Introduction
Datahop is agent-first analytics — product analytics, people, web analytics, feature flags, experiments and feedback in one place.
Datahop answers the everyday product questions: how many people did X, where do they drop off, do they come back, did the change work? It runs at app.datahop.dev. You add a small SDK to your app, and then you or your AI agent use the data.
What makes it different:
- Agents can do everything you can. Connect Claude, ChatGPT or a coding agent over MCP or the CLI. It can read your data, build charts and dashboards, create cohorts and flags, and run experiments. Every change it makes is logged under its own name.
- Results explain themselves. Every chart comes with a short plain-text summary of what it shows. You see it under the chart, and agents read the same text.
- Only what you need. Switch off products you don't use in project settings, and they disappear from the app and from your agent's tools.
What's inside
Product analytics
Trends, funnels, retention, dashboards and a live event feed.
People
Profiles merged across devices, cohorts and groups such as companies.
Web analytics
Visitors, sources, pages, devices and countries.
Feature flags
Roll out to a percentage, a cohort, a plan or a company.
Experiments
Describe an A/B test in a sentence and get a clear verdict.
Feedback
Collect feedback from your own UI, tied to who sent it.