Attribution that works on every visitor.
Not just the ones who accept your cookie banner.

A fully-loaded install of Refer App with every Pro feature enabled and realistic attribution data pre-seeded across eight channels, including AI Referrer traffic from ChatGPT, Perplexity, Claude, and Gemini. You can expand session journeys, review conversion data by channel, explore 90-day trends, content attribution, tags, and integrations. No installation required. Access is granted immediately on email submission and expires after 48 hours.

Almost certainly yes. Refer App works with any website, including WordPress (though the dedicated WordPress plugin is a better fit there), static HTML, Next.js, Shopify, Squarespace, custom CMS, or anything else. The only requirement is that your hosting account supports PHP, which is available by default on virtually all shared hosting plans. Your site itself does not need to be built in PHP.

PHP 7.4 or higher, and SQLite support (the sqlite3 extension, which is enabled by default on virtually all shared hosting and most VPS configurations). No MySQL, no Redis, no build tools, no composer. A standard shared hosting account is sufficient for most traffic volumes.

No. Both the free and Pro tiers record unlimited sessions. Performance scales with your hosting hardware, not with a plan limit. For very high-traffic sites, the SQLite database can be replaced with a more capable backend without changing the application layer.

None, by design. Each session is identified by a random token stored in a first-party cookie. The token contains no personal information and is not derived from any identifier. IP addresses are not stored. Device fingerprinting is not used. The session record contains: the entry channel, the referrer domain, UTM parameters if present, the pages visited during the session, and a timestamp. No name, email, or personally identifiable information is collected at any point.

In a SQLite database file on your own hosting account. Nothing is sent to any third-party server. No cloud service holds your data. The database file is a standard SQLite file you can read, back up, move, or delete yourself at any time.

The architecture is designed to minimise the personal data processed: no personal identifiers, no cross-site tracking, no third-party data transfer. Under both UK GDPR and EU GDPR, processing that does not involve personal data is outside the regulation's scope. However, compliance depends on your specific implementation and jurisdiction. We recommend reviewing your deployment with a qualified privacy professional. We are not lawyers and this is not legal advice.

Refer App sets one first-party cookie: rac_session. It contains a random session token, no personal data. Under UK PECR, cookies strictly necessary for a service the user has requested do not require consent. In standard deployments, no consent banner is required for the attribution cookie alone. Your existing consent setup (if any) is unaffected. This is our assessment; we recommend confirming with a privacy professional for your specific context.

The console is accessed via a URL you choose during setup. It is protected by a login screen. You can further restrict access by placing the console directory behind HTTP basic auth at the server level, restricting access by IP, or keeping it on a non-public path. The application does not expose any public API endpoints.

Yes. Refer App ships as readable PHP source code. There is no obfuscation, no binary, and no compiled component. Your technical team can review the full codebase before deployment.

Refer App is a standalone application, not a WordPress plugin, so it does not use auto-update infrastructure. New versions are released as zip downloads. Updating involves replacing the application files while retaining your database and configuration. The update process is documented step by step. Pro license holders receive email notification of new releases.

Free tier: documentation and community support. Pro: direct email support with a one business day response target. There is no phone support. Enterprise arrangements with higher SLA commitments are available by contacting us directly.

Annual ($199.99/year) includes all Pro features plus one year of updates and email support. At renewal, you keep your data and continue running the current version even if you do not renew, but you will not receive new features or priority support. Lifetime ($599.99, one payment) includes all future Pro updates and support with no renewal required.

One license covers one installation. For multiple sites, purchase a license per installation. An agency plan covering multiple installs at a reduced per-site rate is available. Contact support for current terms.

Refer App is built from the ground up for non-WordPress environments. It has no WordPress dependency: no admin panel, no plugin hooks, no WP functions. The attribution logic is the same nine-channel classification used in the WordPress plugin, but the implementation is a standalone PHP application with its own console, its own database, and its own beacon script.

GA4 is a full analytics platform: page views, events, user behaviour, conversion funnels, audience segments. Refer App is a focused attribution tool: it tells you which channel sent each visitor and what they did during that visit. The two are complementary, not competing. The specific advantage of Refer App is that it captures sessions GA4 misses entirely: ad-blocker users, consent-decliners, privacy browser users, and any session where the tracking pixel does not fire. It also identifies AI Referrer traffic as its own channel, which GA4 currently folds into Direct.

Your data stays on your server in a SQLite file. Nothing is held by a third party. You can export everything as CSV at any time, or simply leave the database in place. It is a standard file you can move, back up, or delete yourself.

The AIR Engine (Attribution Intelligence Report) is a Pro feature that analyses your unclassified Direct traffic using a deterministic rule-based system. It cross-references your Direct sessions against your historical channel data to identify the most probable explanation for each pattern: dark social sharing, cross-device revisits, branded search returns, AI halo traffic, and more. Each finding includes a confidence score, the raw data used to compute it, and a specific recommended action. It does not use any external AI or LLM. All analysis runs on your server against your own session database.