---
language: en
title: "Position-Adjusted Word Count"
description: "A GEO research metric that weights words in sentences citing a source according to their position in the response, without measuring human attention or brand presence."
canonical: "https://tropk.ai/en/resources/glossary/position-adjusted-word-count"
last_modified: "2026-10-12T16:12:04.470Z"
---

# Position-Adjusted Word Count

Reviewed on 2026-10-06.

## What it means

Position-Adjusted Word Count, or PAWC, comes from Position-Adjusted Word Count, a [GEO](/en/resources/glossary/geo) research metric. It sums the words in sentences citing a source, applies a decreasing exponential weight based on their position, and divides by the words in the response. When a sentence cites multiple sources, its words are divided equally among them for the calculation.

## Why it matters for marketing

The metric helps examine how much textual space is associated with a source and where that space appears. This complements a simple link count, which does not distinguish an early reference from one near the end of the response. To evaluate a report, the team needs to confirm that the provider actually implements this formula rather than merely using a similar name.

## Example in practice

Imagine an editorial lab comparing responses about food preservation. In two responses with the same total number of words and sentences, the same guide is cited in a 30-word sentence, and no other source shares the [citation](/en/resources/glossary/ai-citation). The sentence appears near the beginning of the first response and near the end of the second. The team identifies a higher weight in the first and keeps both responses to check the calculation.

## How to apply or verify

For sentences citing the source, calculate sum[(w/k) × exp(−p/N)] / W: w is the number of words in the sentence; k, the number of sources it cites; p, the sentence's position; N, the total number of sentences; W, the number of words in the response. Preserve the position convention, segmentation, and original citations. Manually check cases with multiple sources before aggregating results by question.

## Important distinctions

The subject is the cited source, not a brand name or its ordinal position among competitors. The weight represents a hypothesis of greater exposure for earlier passages; it does not measure actual attention or reading. The research also presents unadjusted counts and subjective evaluation. PAWC should not be confused with these measures, traffic, or a guarantee of citation.

## Sources and references

- [GEO: Generative Engine Optimization — Version 3](<https://arxiv.org/html/2311.09735v3>) — Pranjal Aggarwal e coautores, arXiv

## Continue a leitura

- [GEO](</en/resources/glossary/geo>)
- [Share of sources](</en/resources/glossary/share-of-sources>)
- [Position bias](</en/resources/glossary/position-bias>)

## About the author

[Vinicius Pacheco](</en/team/vinicius-pacheco>)

Responsible for Tropk’s technical content, focusing on crawling, indexing, structured data, and agent access to content.

---

[All pages in Markdown](<https://tropk.ai/sitemap.md>)