> ## Documentation Index
> Fetch the complete documentation index at: https://docs.getcited.in/llms.txt
> Use this file to discover all available pages before exploring further.

# Prompt Engineering

> The practice of crafting specific inputs to LLMs to produce desired outputs — in AEO/GEO, the discipline of writing prompts that reflect real customer language.

**Prompt engineering** is the practice of crafting specific inputs to [large language models](/glossary/llm) to produce desired outputs. In the context of [AEO/GEO](/glossary/geo) measurement, prompt engineering is the discipline of writing prompts that accurately represent how real customers ask questions — consumer-authentic language, specific intent, and natural phrasing rather than keyword-style searches.

## Why it matters

The quality of AI visibility measurement depends entirely on the quality of the prompts used. A prompt phrased as "best CRM India 2026" (keyword-style) produces different AI responses than "which CRM do Indian startups actually use" (consumer-authentic). Cited's measurement validity rests on prompts that mirror real customer conversations, not SEO-style keyword constructions. Poor prompt engineering produces measurements that do not reflect actual customer discovery.

## How Cited uses it

Cited's [prompt-generation pipeline](/methodology/query-generation) is essentially an automated prompt engineering system. It generates consumer-authentic prompts from brand intelligence, filters out marketing jargon, and classifies each prompt by [intent type](/glossary/intent-type) and [query archetype](/glossary/query-archetype). The anti-jargon filter and consumer-authenticity scoring are the prompt engineering quality gates.

## Related concepts

* [How we generate prompts](/methodology/query-generation) — Cited's automated prompt engineering pipeline
* [Query archetype](/glossary/query-archetype) — the structural categories of generated prompts
* [Intent type](/glossary/intent-type) — the purpose classification applied to each prompt