What question does this page answer?

Question identification

Every page is analyzed to extract the primary question it answers. Not the keyword it targets — the actual question a person would ask to land on it. That question becomes the anchor for everything downstream: SERP analysis, fact research, rewrite strategy.

Question matching

Two pages answering the same question split authority instead of building it. Question matching runs across your entire site — not keyword overlap, but structural question duplication — and surfaces which page should own the topic and which is competing with it.

This is one of the most common silent problems on growing sites. Content accumulates. Nobody tracks what each piece was for. Question matching makes the duplication visible.

Alignment — URL, title, content

Is the URL, title, H1, and body content all answering the same question? Misalignment is one of the most common reasons pages don't rank — the signals compete instead of reinforce. Every page is checked for alignment across all three layers.

CTA alignment

Does the call to action match the question the page answers? A guide that ends with a hard sell conflicts with the intent that brought the visitor there. A service page with no CTA loses the conversion it was built for. CTA alignment checks whether the ask matches the answer — and flags pages where they're working against each other.

Page model extraction

Every page is parsed into a structured model — sections, content blocks, semantic structure. The page model is what the content generation pipeline works from. It's not the raw HTML, it's a clean representation of what the page actually says and how it's organized.

Pre-generation scoring

Before a rewrite is queued, the page is scored against Google's Search Quality Rater Guidelines — expertise, authoritativeness, trustworthiness, and experience signals. The score sets the baseline. Post-generation scoring measures the delta. You can see what changed and by how much.

The foundation for content generation

Question identification, page model, and pre-generation scoring feed directly into the content generation pipeline. Before a page gets researched or rewritten, we know what it's supposed to answer, what it currently says, and how well it's saying it.