Analytics for open-source projects: adoption signals, not surveillance

Open-source maintainers are rightly wary of analytics: their users are developers who block trackers and dislike being profiled. But without any data, it's hard to know whether the new docs helped, which install method people use, or what to tell a potential sponsor.

VisitTrack's cookieless mode stores nothing in visitors' browsers and IP addresses are never stored. File downloads and outbound clicks (to GitHub, package registries, GitHub Sponsors, Open Collective) are tracked automatically, docs engagement is measured per page, and a public read-only dashboard lets you publish the numbers. One thing to know up front: VisitTrack is a hosted product, not open-source software.

The open-source metrics that matter on your website

1. 1.Install-command copies

   The closest thing a website has to an install count. Track the copy button per package manager (npm, pip, brew) to see what your users actually use.

2. 2.Clicks to the repository and registry

   Outbound clicks to github.com, npmjs.com, pypi.org and others, per destination. They complement GitHub's own traffic insights, which only cover the repository and a 14-day window.

3. 3.Downloads

   Binary and archive downloads are recorded automatically as file_download with the file extension.

4. 4.Docs engagement and search queries

   Active reading time, scroll depth and Search Console queries per docs page — where to improve docs and what people search for.

5. 5.Sponsor and donation clicks

   Outbound clicks to GitHub Sponsors or Open Collective. A useful number when writing a sponsorship ask.

Questions maintainers ask

How many people try to install it?

Count install_copied events per package manager, and file_download events for binaries.

Where do new users discover the project?

The sources breakdown: Hacker News, Reddit, newsletters, search and AI assistants each appear separately.

Which docs pages need work?

Pages with traffic but low scroll depth, plus 404s after restructures (recorded automatically).

Can we publish our stats?

Yes. Turn on the public dashboard for a read-only link anyone can open, or embed a widget on the project site.

How to set up analytics for an open-source project

1. 1

   Install the cookieless tag

   On the project site and docs (same site id if they share a domain). Static-site generators and docs frameworks just need the tag in the head — see Astro or the documentation sites guide.

   <script defer data-site="SITE_ID" data-cookieless src="https://visitrack.app/tracker.js"></script>

2. 2

   Count install-command copies

   Send an event from your copy button with the package manager as a prop, or give each button its own data-vt-goal so they show up as separate goals.

   <button data-vt-goal="install_copied_npm">Copy</button>
   <button data-vt-goal="install_copied_brew">Copy</button>

3. 3

   Let downloads and outbound clicks track themselves

   Release archives and installers are recorded as file_download; clicks to GitHub, registries and sponsor platforms as outbound_click. Nothing to add.

4. 4

   Publish the numbers (optional)

   Turn on the public dashboard in Settings. It shows the whole dashboard read-only — for a docs-only project that's usually just traffic, sources and events. Or embed the visitors-trend card in the README site.

5. 5

   Say what you collect

   Add a short note to the site's privacy page: cookieless, no IP storage, aggregate data only. It matters to developer audiences. See cookieless mode.

Example: a CLI tool's project site

A popular open-source CLI with a docs site, ~18,000 monthly visitors.

Illustrative numbers for a typical site of this kind, not real customer data.

Two surprises: a sizable Windows audience that never shows up in issues, and AI assistants recommending the tool. The first argues for better Windows install docs; the second for keeping the docs crawlable and accurate.

VisitTrack vs open-source analytics and GitHub insights

Third-party details as of October 2026.

If running open-source infrastructure is a principle for your project, choose Umami, Plausible CE or GoatCounter. If you'd rather not maintain a server, VisitTrack is a hosted, cookie-free option. See VisitTrack vs Umami and vs GoatCounter.

What VisitTrack costs for an open-source project site

Project sites are usually small; docs add pages per visit.

Every feature is on every plan; the price only follows monthly events (pageviews, custom events and payments, counted across all your sites). Bot and AI-crawler hits and session replays are never billed. Yearly billing is ten months' price for twelve, every plan starts with a 14-day free trial with no card, and there's a one-time lifetime deal from $59.