How It Works
A deep dive into the proxy architecture that powers SEO page delivery.
Overview
Nexting uses a reverse proxy pattern to serve SEO pages under your domain. This means when someone visits a specific path on your website, your hosting platform fetches the content from our servers and returns it to the visitor—all without changing the URL in their browser.
This is the same technique used by large-scale CDNs and enterprise applications. It's secure, performant, and completely transparent to end users.
Request Flow
Let's trace what happens when a user visits a proxied page:
GET https://yourdomain.com/blog/seo-articleVercel/Netlify finds a matching rewrite rule in your configuration
GET https://api.nexting.ai/proxy/YOUR_KEY/blog/seo-articleOur API looks up the page content and returns fully-rendered HTML with SEO meta tags
Browser displays content at yourdomain.com/blog/seo-article
Server-Side Only
Key Components
Project Key
A unique identifier (like abc123xyz789) generated for each project. This key is used in the proxy URL to identify which project's pages to serve.
Rewrite Rules
Configuration that tells your hosting platform which paths to proxy. Each rule maps a source path on your domain to a destination URL on Nexting.
{
"source": "/blog/article",
"destination": "https://api.nexting.ai/proxy/YOUR_KEY/blog/article"
}Verification Endpoint
A special path (/__verify__) used to confirm your rewrite rules are working correctly. When you click "Verify Configuration" in the dashboard, we check if this endpoint is accessible through your domain.
URL Structure
Understanding how URLs map between your domain and Nexting:
| Your Domain | Nexting API | |
|---|---|---|
| example.com/__verify__ | → | api.nexting.ai/proxy/KEY/__verify__ |
| example.com/blog/article | → | api.nexting.ai/proxy/KEY/blog/article |
| example.com/seo/page-1 | → | api.nexting.ai/proxy/KEY/seo/page-1 |
Path Prefix
/blog) in your project settings, all your pages will be organized under that prefix, keeping them separate from your main site content.What Gets Served
When Nexting receives a request, we return a complete HTML document with:
- Your page content (formatted as HTML)
- SEO meta tags (title, description, keywords)
- Open Graph tags for social sharing
- Canonical URL pointing to your domain
- Clean, responsive styling
<!DOCTYPE html>
<html lang="en">
<head>
<title>Your SEO Title</title>
<meta name="description" content="Your meta description">
<link rel="canonical" href="https://yourdomain.com/blog/article">
<meta property="og:title" content="Your SEO Title">
<meta property="og:url" content="https://yourdomain.com/blog/article">
</head>
<body>
<article>
<!-- Your page content here -->
</article>
</body>
</html>Search Engine Friendly
Verification Process
When you click "Verify Configuration" in the dashboard, here's what happens:
Dashboard sends verification request
We make a request to https://yourdomain.com/__verify__
Your host processes the rewrite
If configured correctly, your host proxies the request to Nexting
Nexting returns verification data
Our API returns a JSON response containing your project key
{
"success": true,
"projectKey": "your-project-key",
"timestamp": "2024-01-15T10:30:00Z"
}We confirm the project key matches
If the returned project key matches your project, verification succeeds and your project becomes active
Common Verification Issues
Security Considerations
Project Key Privacy: Your project key is a random 12-character identifier. While it appears in configuration files, it only allows read access to published pages.
HTTPS Only: All communication between your host and Nexting happens over HTTPS, ensuring data integrity and privacy.
Published Pages Only: Only pages you explicitly publish are accessible through the proxy. Draft pages remain private.
No Direct Access: Users cannot access your Nexting dashboard or modify content through the proxy—it's read-only.
Performance
The proxy adds minimal latency to page loads:
- Nexting API is deployed on edge servers worldwide
- Responses are cached at both Nexting and your hosting platform
- HTML is pre-rendered and served instantly
- Typical additional latency: 50-150ms
Caching
Cache-Control: public, max-age=60, s-maxage=300, meaning edge servers cache content for up to 5 minutes. Content updates may take this long to propagate.