What Data Liwan Collects

Liwan is designed to be privacy-friendly and respects the privacy of your visitors. It collects only the data that is necessary for generating useful analytics while ensuring that visitors remain anonymous.

Liwan is self-hosted. Analytics data is processed and stored on your infrastructure, not on Liwan-managed cloud servers.

Collection happens per request and is scoped to the entity (website/app) that received the tracking event. Liwan intentionally avoids storing directly identifying values such as IP addresses.

Where is the data stored/processed?

Liwan runs on your server or hosting environment. By default, collected analytics data stays within your deployment and data storage.

What data does Liwan collect?

Liwan collects the following data:

• URLs: The URL of the page that was visited (excluding query parameters and fragments).
• Referrers: The URL of the page that referred the visitor to the current page. Only the domain is stored.
• Browser: The browser used by the visitor (e.g., Chrome, Firefox). The version is not stored.
• Operating System: The operating system used by the visitor (e.g., Windows, macOS). The version is not stored.
• Device Type: The type of device used by the visitor (e.g., desktop, mobile).
• Timestamp: The time of the visit/event, used for time-series analytics and daily aggregations.
• Country (optional): The country of the visitor based on their IP address.
• City (optional): The city of the visitor based on their IP address.
• Visitor ID: An anonymized ID that is unique to each visitor. This is used to track unique visitors across multiple visits.

What data does Liwan NOT collect?

• IP Addresses: Liwan does not store IP addresses. Instead, it uses a hash of the IP address to generate the visitor ID. This ensures that visitors remain anonymous.
• Full Referrer Paths: Liwan stores only the referrer domain, not the full referrer URL path/query.
• Query Parameters and URL Fragments: Liwan strips these from tracked page URLs.
• Usernames: Liwan does not collect usernames or any other personally identifiable information.
• Cookies: Liwan does not use cookies to track visitors.
• Cross-Site Tracking: Visitors have a unique visitor ID that is specific to each website. This means that visitors are not tracked across different websites.
• Persistent Identifiers: Visitors are not tracked across multiple days. Each day is treated as a separate session.

How are Visitors Identified?

Liwan uses a combination of the visitor’s IP address and user agent to generate a unique visitor ID. This ID is pseudonymized using a one-way hash function and a unique salt. The salt is rotated daily to ensure that the IP and user agent combination cannot feasibly be reversed, making users anonymous. The IP is discarded right after generating the visitor ID and never stored on disk.

# Pseudocode for generating the visitor ID
blake3(ip, user_agent, daily_salt, entity_id)[0:16]