Skip to content
Science-backed data on why AI says what it says
Citation signals

Answer position

Answer position is BLURSOR's estimate of where the first heading followed by substantial body text occurs in the extracted content. It is a structural reading aid, not a judgment that the passage answers a customer's question or will be cited.

Checked against BLURSOR’s measurements ·

What it means

A useful section usually names its subject and then explains it. Heading and paragraph markup makes that relationship easier to navigate and inspect. For a business owner, the goal is to help visitors reach relevant information without searching through an unrelated introduction.

The checker uses a simple pattern to draw attention to where an explanation might begin. It does not understand the visitor's question. A heading followed by generic marketing prose can satisfy the pattern, while a useful answer inside a layout container the detector does not recognize can be missed.

Illustrative example

Hypothetical example: reading 7 of 69 blocks

Imagine a fictional delivery guide producing 69 extracted blocks. Its seventh block is a heading, Delivery areas, followed by a paragraph of at least 40 characters explaining the places served. A result of 7 of 69 means the detected heading starts at block seven; it does not mean seven answers were found or that only seven blocks are useful.

The owner still needs to read the paragraph. If it says only that the company offers excellent service, the early position has not established that customers can find their delivery area.

How BLURSOR measures it

In the report: First heading followed by body text

BLURSOR extracts heading, paragraph, list-item, and blockquote elements from the raw HTML after removing common code and marked navigation, header, footer, aside, and form sections. It excludes blocks shorter than eight characters and uses a bounded scan. Fewer than five extracted blocks leaves the citation measurements unavailable.

For each heading, it looks within the next four blocks for a paragraph, list item, or blockquote containing at least 40 characters, stopping at another heading. The first matching heading's one-based position is reported. The early-position flag uses the first 30% of blocks, rounded up. A separate early-text percentage describes the character share in that opening group, not how much of the page answers a question.

What to do next

  1. Read the first detected section and check whether it explains something a visitor actually needs to know.
  2. Put essential explanations near descriptive headings, using ordinary paragraph or list markup where appropriate.
  3. If no pair is detected, inspect the page and extraction before adding headings or rearranging useful content solely to change the result.

What it cannot tell you

  • Block position is document order in the extraction, not screen position, scrolling distance, or reading time.
  • The 30% window and text-length cutoff are checker heuristics. They are not validated thresholds for answer quality, search visibility, or AI citation.

Sources and further reading

The definitions below describe the underlying concepts. The measurement section describes BLURSOR’s own checks.

Related explanations