Proxying Third-Party JavaScript as First-Party JavaScript (and the Potential Effect on Analytics)

This article explores a simple Cloudflare Worker to proxy third-party JavaScript, effectively making it appear as first-party to the browser. Let's examine the ease of implementation and its implications for website analytics.

A basic Cloudflare Worker to proxy a URL is remarkably straightforward:

addEventListener("fetch", (event) => {
  event.respondWith(
     fetch("https://css-tricks.com")
  );
});

While lacking error handling, this demonstrates the core functionality. Consider websites providing JavaScript URLs for embedding, such as CodePen's Embedded Pens feature (https://cpwebassets.codepen.io/assets/embed/ei.js). This URL can be proxied equally easily using a Cloudflare Worker. The worker automatically handles the content-type header.

Cloudflare Workers provide their own URLs, but you can also easily integrate a worker into your website's routing. This allows you to serve the proxied JavaScript from your own domain.

By doing this, the JavaScript loads as if it were first-party, even though it's actually proxied. The advantage? First-party scripts are rarely blocked by ad blockers. This technique could potentially be misused to circumvent ad blockers, a practice with ethical implications. However, proxying can be beneficial for legitimate purposes, such as overcoming CORS issues.

The impact on analytics is particularly interesting. Using Plausible, a privacy-focused analytics service, as an example, we can see how proxying affects data accuracy. Plausible, while aiming for privacy, might still be blocked by users, leading to incomplete data. Proxying it as first-party mitigates this issue.

A comparison of data from using Plausible directly versus using the proxied version reveals a significant difference. The proxied version showed a substantial increase in unique visitors (around 13.8% compared to the non-proxied setup) and pageviews (19.4% more than the non-proxied setup). This strongly suggests a significant portion of users block third-party analytics scripts. The data suggests that 20-30% of users might be blocking third-party analytics on this specific website.

The percentage increase calculation used is: (final - initial) / final * 100.

The above is the detailed content of Proxying Third-Party JavaScript as First-Party JavaScript (and the Potential Effect on Analytics). For more information, please follow other related articles on the PHP Chinese website!