Vince

Vince is a self‑hosted web analytics engine that positions itself as a lightweight, privacy‑first alternative to Google Analytics and Plausible. From a technical standpoint it ingests JavaScript events from a tiny <script> tag, stores them in a highly efficient columnar format based on Compressed Roaring Bitmaps, and exposes a REST API for dashboard queries. The entire stack is delivered as a single statically‑compiled binary, eliminating runtime dependencies and allowing developers to deploy it in any environment—from a single server VM to a Kubernetes cluster—using only one binary and optional configuration flags.

Technical Stack & Architecture

Vince is written in Go (Golang), leveraging its excellent performance and built‑in concurrency for event ingestion and background processing. The core data store is a custom key/value engine built on Pebble, a fork of LevelDB optimized for fast writes. On top of Pebble, Vince compresses event dimensions (e.g., page path, referrer, device) into Roaring Bitmaps that enable sub‑millisecond set operations for filtering and aggregation. The application exposes a minimal HTTP API (JSON over HTTPS) that is fully documented on the site’s /tags/api/ page. TLS is handled automatically via Let’s Encrypt, so developers can focus on metrics rather than certificate management.

Core Capabilities

• Event Ingestion: Accepts standard web analytics events (page view, click, download, outbound link) plus custom events via POST requests.
• Goal & Funnel Tracking: Supports conversion goals and basic funnel steps, though it does not implement the full funnel UI of Plausible.
• Advanced Querying: The API allows filtering by date range, dimension values (city, source, device), and event type. Internally these queries are resolved using bitmap intersections for blazing speed.
• Public & Shared Dashboards: Dashboards can be made public or shared via unique, password‑protected links. The rendering layer is a lightweight React app bundled into the binary; no separate frontend deployment is required.

Deployment & Infrastructure

Vince ships as a single statically‑compiled binary; no external services are required. It can run natively on Linux, macOS, or Windows. For containerized environments the official Docker image (ghcr.io/vinceanalytics/vince) is available, and a Helm chart lets you deploy it into Kubernetes with minimal configuration. Because the data store is embedded, scaling horizontally requires running multiple instances behind a load balancer and sharding data by domain or time slice—a pattern developers can implement using the provided --shard flag. The application’s memory footprint stays under 100 MiB for moderate traffic, making it cost‑effective on shared hosting or low‑tier cloud VMs.

Integration & Extensibility

Vince offers a RESTful API that can be consumed by custom dashboards, CI/CD pipelines, or third‑party analytics tools. Webhooks are supported for event forwarding; developers can hook into the POST /api/v1/events endpoint to trigger downstream processes. The lightweight script tag (<script src="https://your.vince.instance/v1/track.js"></script>) is under 1 KB and can be embedded in any HTML page without affecting load times. While it does not expose a plugin system, its API surface is intentionally minimal so that external services can extend functionality without modifying the core binary.

Developer Experience

Configuration is handled via command‑line flags or environment variables—no YAML files are required. The documentation on the official site is concise and API‑first, with example payloads and query parameters. Community support is growing; the project’s GitHub repository hosts an active issue tracker, and a Slack channel (invite link on the website) provides real‑time assistance. Because all code is open source under an MIT license, developers can audit the privacy guarantees and contribute performance improvements directly.

Use Cases

• Privacy‑Conscious Web Projects: Sites that must avoid cookies and GDPR/CCPA compliance can use Vince to collect anonymized metrics without user consent.
• Small to Medium Enterprises: Organizations that need unlimited sites and events on a single server can deploy Vince without paying for tiered analytics services.
• CI/CD Analytics: Automated pipelines can push custom events to Vince and query them via the API for build success rates, deployment frequencies, or user engagement metrics.
• Self‑Hosted SaaS: Multi‑tenant SaaS platforms can deploy a single Vince instance and isolate metrics per tenant by using unique domain identifiers.

Advantages

Vince’s main selling points for developers are performance, simplicity, and privacy. The bitmap‑based storage delivers sub‑millisecond aggregation even for billions of events, while the single binary eliminates dependency hell. Its automatic TLS and zero‑configuration operation mean developers can spin up a production instance in minutes. Licensing is free under MIT, so there are no hidden costs or feature restrictions—unlike Plausible’s enterprise tier. For teams that value control, speed, and compliance, Vince offers a compelling alternative to cloud‑hosted analytics platforms.