Know what brings you paying customers

Yes. Every plan starts with 7 days free, and you can cancel any time before the trial ends.

A pageview, a goal (a click or a signup you chose to track) or a payment. Crawler hits are counted separately and don’t use your quota. Pick the size on the pricing slider, from 25K to 10M events a month; Starter and Growth are the same ladder; Growth adds sites and people.

Nothing is lost. The dashboard keeps counting, and you’ll get a note suggesting the next plan. Two months over in a row and we’ll ask you to move up.

No. The script is 6.6 KB gzipped, has no dependencies, and loads after your page.

Stripe, Dodo Payments, Polar and Creem work out of the box (one webhook each). Any other provider can send payments with a simple API call.

Yes. GeckoStats is an MCP server: add it to Claude, Claude Code, ChatGPT, Cursor, Grok or any MCP client with one command or a connector, and ask about your own traffic, revenue, search, funnels and customers. It reads the same numbers the dashboard shows, and only the site whose key you gave it.

Everything the dashboard draws: overview, sources, pages, campaigns, goals, funnels, customers, search queries, releases and the live view, through 26 tools. The same queries, the same numbers. Agents read; they cannot change settings or delete anything.

Only what the agent asked for in that conversation: the rows that answer its query, the same way a chart answers a click. Nothing is sent on its own, and GeckoStats never passes data to any AI company itself. The key is per site and can be revoked from settings at any time.

The checkout carries the visitor id, and the payment webhook brings it back. When that’s missing, the return page, the customer’s earlier payments, the email you identify()’d and finally checkout timing are tried in that order, and the dashboard tells you which one matched. Credit goes to the last visit that came from a real source, so a tweet still gets it when someone returns directly two weeks later.

Yes, through your own Search Console: every query and page you rank for, with the revenue each earned, opportunities ranked by money rather than by volume, what Google’s index says about the pages that earn, and the queries losing clicks at an unchanged position. There is no crawler; it reads what Google already knows about you.

Yes. Your site reports what GPTBot, ClaudeBot, PerplexityBot and the rest read (a crawler runs no JavaScript, so a tracker script cannot see them), and the dashboard puts that against the visitors those same services sent back, with the revenue they brought.

You can serve the script from your own domain with a one-line rewrite, so blockers don’t see it. Payments are always recorded either way, because they come from the webhook.

By default, no. The tracker sets no cookies at all: a visitor is a salted hash of IP, user agent and site that rotates every day, and the IP itself is never written anywhere. Nobody can be followed across days or across sites, so there is nothing to ask consent for. If you turn on cookie mode (a per-site switch, for when identity across days matters more) that changes, and then it is worth asking your own lawyer.

No. Revenue per search query is marked ≈ because Google never ties a query to a person; a page is only priced once it has traffic; Core Web Vitals wait for 30 samples; and demo data says it is demo data everywhere it appears.

Yes, up or down, at any time. Moving up takes effect right away; moving down takes effect at the end of the period you’ve paid for.