Victor Chanet, a solo founder based in Warsaw, Poland, launched Datakool, a privacy-first analytics tool with a tracking script under 1KB. That's 75x smaller than Google Analytics' 75KB+ payload. The cookieless design eliminates consent banners and handles GDPR, CCPA, and PECR compliance out of the box. Chanet bootstrapped the project without venture funding.
Cookieless tracking works by processing visitor data server-side rather than storing identifiers on users' devices. No cookies means no consent banners cluttering your site. Plausible and Fathom take similar approaches, though each makes different trade-offs around accuracy and features.
Datakool's MCP (Model Context Protocol) integration is where it diverges from the pack. You can query your analytics data directly through Claude Code, Cursor, or any MCP-compatible AI assistant. Connect your API key and ask questions in natural language. It's one of the first analytics tools to treat AI assistants as a primary interface, not an afterthought. Many developers already use AI-powered tools like Pluck to streamline their workflow.
Pricing starts at $2/month for 10,000 pageviews on one site. The Pro tier runs $10/month for 100,000 pageviews across three sites, and the Founder plan costs $50/month for 10 million pageviews with unlimited sites. A launch promo code (DKLAUNCH) grants lifetime free access to the Solo plan. Some Hacker News commenters noted the lack of a permanent free tier for hobby projects, which could limit adoption among smaller users.