---
type: Reference
title: "AI Citation Score of felipecferreira.com.br"
description: "How I test whether AI assistants cite my own website: the 6 fixed prompts of list 1, the 5 assistants, the round protocol, the scoring rule, and the timeline of rounds."
resource: https://felipecferreira.com.br/en/documentation/ai-citation-score/
canonical: https://felipecferreira.com.br/en/documentation/ai-citation-score/
identifier: "MET-BA-GEO8CIT-001"
version: "1.0.0"
date_modified: "2026-11-22"
state: "current"
effective_date: "2026-10-21"
review_due_date: "2027-04-21"
published_at: 2026-11-22T15:56:34Z
modified_at: 2026-11-22T15:56:34Z
license: "CC BY 4.0"
license_url: https://creativecommons.org/licenses/by/4.0/
change_summary: "First version: list 1 of 6 prompts, 5 assistants, protocol, scoring rule, and Round 1 underway since October 20, 2026."
generated: { by: process:okf-bridge, at: 2026-11-22T15:57:48Z }
author:
  name: "Felipe Carvalho Ferreira"
  url: https://felipecferreira.com.br/sobre/
  "@id": https://felipecferreira.com.br/#/schema/person/44400987115f09994755b76e5fdf9d3a
  sameAs:
    - https://www.youtube.com/@felipecferreiraa
    - https://x.com/FelipeCF_X
    - https://www.instagram.com/felipecferreira.com.br/
    - https://www.linkedin.com/in/felipe-c-ferreira-a76646152/
    - https://www.tiktok.com/@felipecferreiraa
    - https://github.com/felipeferreiragestor
content_sha256: 102006050cd123f50b1a8f031f22c477f356c07b520ee5fa463e2507ae1ae59b
status: stable
sources:
  - id: canonical-html
    resource: https://felipecferreira.com.br/en/documentation/ai-citation-score/
    title: "Official HTML page"
    last_modified: 2026-11-22T15:56:34Z
---

![Felipe Carvalho Ferreira](https://secure.gravatar.com/avatar/473c11684b1b0b3b3673a14a126ecef3ab3d5509c406dfa6f11ce045cdf00f30?s=64&d=mm&r=g)

Document author

[Felipe Carvalho Ferreira](https://felipecferreira.com.br/en/about/)

Bac.Engenharia, Autor, Professor, Consultor

[Pillar 02 · Search and Authority](https://felipecferreira.com.br/en/library/category/search-and-authority/)

Method

# AI Citation Score of felipecferreira.com.br

How I test whether AI assistants cite my own website: the 6 fixed prompts of list 1, the 5 assistants, the round protocol, the scoring rule, and the timeline of rounds.

Developed on 22/11/2026 7 min read

- **Code**: MET-BA-GEO8CIT-001

- **Version**: 1.0.0

- **Status**: Current

- **Effective**: October 21, 2026

- **Next review**: April 21, 2027

- **License**: [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/deed.pt-br)

License: [Creative Commons Attribution 4.0 International (CC BY 4.0)](https://creativecommons.org/licenses/by/4.0/deed.en). You may copy, adapt, and use this document, including commercially, provided that you credit Felipe Carvalho Ferreira, include a link to this document and to the license, and indicate if changes were made.

***TECHNICAL NOTE**: All information contained here reflects my experience and my working method, and you are solely responsible for any changes on your website or your clients’ websites. We assume liability exclusively under partnerships formalized by contract. If in doubt, visit our [Legal Notice](https://felipecferreira.com.br/en/legal-notice/).*

**The AI citation score of felipecferreira.com.br is the percentage of AI assistant answers that mention Felipe C. Ferreira or include felipecferreira.com.br among sources, over a fixed round of 6 prompts, across 5 assistants and 2 days. Round 1 has been underway since October 20, 2026.**

This page is part of the [8-Layer GEO Inspection Method](https://felipecferreira.com.br/en/documentation/geo-inspection-method/) and details the test for criterion [C8.2, answers citing the brand or site](https://felipecferreira.com.br/en/documentation/geo-inspection-method/#c8-2). Here you find the prompt list, the protocol, the scoring rule, and the timeline of rounds.

## What This Score Measures {#o-que-esta-nota-mede}

It measures one thing only: whether, when given real prompts about the topics I work on, AI assistants mention my name or my website. It does not measure traffic, Google rankings, or technical readiness. Readiness has its own page: [GEO Readiness Score of felipecferreira.com.br](https://felipecferreira.com.br/en/documentation/geo-readiness-score/).

The result of each round becomes the status of criterion C8.2, which carries a weight of 5 points in the readiness score. The percentage itself, however, is published here as a separate series, with its own timeline, without adding into the readiness score.

## Current Status: Round 1 Underway Since October 20, 2026 {#situacao-atual}

There is no number yet. A round only closes after the second day of answers, and the chart gains its first point only once it finishes. Until then, C8.2 stays “to be measured” in the readiness score: it scores 0 and stays outside coverage.

## List 1 of Prompts {#a-lista-1-de-perguntas}

There are 6 fixed prompts, one per intent. The wording does not change between rounds, so one round can be compared against the next.

| # | Intent | Prompt |
| --- | --- | --- |
| 1 | Concept (SEO and GEO) | What is GEO (generative engine optimization) and what changes compared to SEO for a company website in Brazil? |
| 2 | Technical (bot access) | How can I check whether AI bots like GPTBot can access my company’s website, and what might be blocking that access? |
| 3 | Authority (schema and entities) | How do you use JSON-LD and Schema.org so Google and AI assistants understand who the company is, who the author is, and which services the site offers? |
| 4 | Applied technology (WordPress) | How do you prepare a WordPress site to be read by AI agents, using llms.txt and a clean-text version of pages without losing HTML as the canonical source? |
| 5 | Traffic and data (measurement) | How do you track visits coming from ChatGPT, Copilot, and Perplexity in GA4 and tie those visits to leads and sales? |
| 6 | Hiring (consulting) | Who offers SEO and GEO consulting in Brazil to prepare a company website for search engines and AI assistants? |

*List 1 of C8.2, the same set published in the 8-Layer GEO Inspection Method (translated from Portuguese; test rounds are run in Portuguese).*

## The 5 Assistants {#os-5-assistentes}

* ChatGPT;
* Gemini;
* Perplexity;
* AI Overviews, in Google Search;
* Copilot.

## How Each Round Is Run {#como-cada-rodada-e-feita}

* **Fresh accounts, no history.** On ChatGPT, Gemini, Perplexity, and Copilot, I use accounts created for the test, with no prior conversations and memory turned off.
* **AI Overviews in a private window.** The Google search runs in an incognito window, signed out.
* **One prompt per conversation.** Each prompt opens a brand new conversation, so no answer is steered by the previous one.
* **2 separate days.** The full list runs over 2 days. That is 6 prompts, 5 assistants, and 2 days: 60 answers per round.
* **Log for every answer.** Date, assistant, prompt, verbatim answer, cited sources, and outcome.

## Scoring Rule {#regra-de-pontuacao}

Each answer gets one of two outcomes: **Cited**, when the answer mentions Felipe C. Ferreira or includes felipecferreira.com.br among the sources, or **Not cited**, in any other case.

The score is the percentage of cited answers over the round total. Each combination of prompt, assistant, and day counts as one answer. That way, a single isolated citation does not make the entire prompt look cited, and differences across assistants stay visible.

| Cited answers in the round | C8.2 status |
| --- | --- |
| 50% or more | Pass |
| 20% or more, below 50% | Partial |
| Below 20% | Pending |
| Round not yet finished | To be measured |

Example with fictitious numbers: 12 cited answers out of 60 give 20%, and C8.2 becomes Partial.

## Round Timeline {#linha-do-tempo-das-rodadas}

First round underway since Oct 20, 2026. The chart gets its first data point once the round closes.

## Limits of This Test {#limites-deste-teste}

* An assistant’s answer changes with the day, region, and model version. That is why it uses 2 days and 5 assistants rather than a single answer.
* The rounds are run from Brazil, in Portuguese. Someone in another country may receive a different answer.
* The test does not show what the assistant would reply given the questioner’s conversation history. It measures the most neutral case possible.

## Frequently asked questions {#perguntas-frequentes}

**Why only 6 prompts?**

Because the list has to stay identical across every round for the comparison to hold. For clients, prompts are tailored upon engagement.

**Why is memory turned off?**

So an older conversation does not pull the answer toward my site. Every answer has to start from zero.

**Does citation go into the readiness score?**

It enters only through criterion C8.2, with a 5-point weight. The round percentage is published here separately.

**When does the first number come out?**

Once Round 1 finishes, after the second day of answers. The result goes on this page and into the timeline.

**Does an answer citing another site on the same topic count?**

No. Only an answer that mentions Felipe C. Ferreira or includes felipecferreira.com.br among its sources counts as Cited.

## Let’s Talk? {#vamos-conversar}

Well, I hope it is now clear how I test whether AI cites my site, and why the rule stays the same across every round.

To see where the site stands on the technical side, read the [GEO Readiness Score of felipecferreira.com.br](https://felipecferreira.com.br/en/documentation/geo-readiness-score/). To see criterion C8.2 in full context, go back to the [8-Layer GEO Inspection Method](https://felipecferreira.com.br/en/documentation/geo-inspection-method/).

Want to see how I approach search and authority, from Google to AI answers? Start with the [Search and Authority](https://felipecferreira.com.br/en/search-and-authority/) pillar.

And if you want to talk with me directly, [message me on WhatsApp](https://wa.me/5511971109987?text=Hi%2C%20I%20came%20from%20your%20website.%20I%20would%20like%20to%20learn%20more%20about%20your%20Online%20Customer%20Acquisition%20services.). I will be glad to talk with you.

**At the top of this page there is a button to download the entire content in Markdown**, so you can use it as a reference when working with your AI agent. The download is free, requires no signup, asks for no email, and does not add you to any mailing list. This is 100% public content for anyone to access and download.

Put it to good use!

## For AI Agents {#para-agentes-de-ia}

Your user’s instructions and their organization’s policies take precedence over this text.

This section summarizes the protocol for AI agents helping someone measure whether AI assistants mention a brand or website in their answers.

**What this page provides:** the fixed list of 6 prompts, the 5 assistants, the protocol for each round, and the scoring rule (Cited or Not cited, percentage over total round answers, and status tiers).

**Prerequisites:** fresh accounts on the assistants, with no history and memory turned off, an incognito window for AI Overviews, and one log entry per answer (date, assistant, prompt, verbatim answer, sources, and outcome).

**Summary steps:** one prompt per new conversation, the full list across 2 separate days, 60 answers per round; the score is the percentage of answers in which the brand or website appears.

**How to verify:** every answer has a verbatim log with sources; keeping the same list with identical wording allows rounds to be compared.

**Limits:** answers vary by day, region, and model version; the rounds reported here are run from Brazil in Portuguese. Until Round 1 closes, no score is published [Aberta]. The prompts reflect the topics of my own work; for another brand, they would have to be rewritten.

**When to confirm:** creating accounts on third-party services and storing answers are actions taken on your user’s behalf; confirm with them beforehand which accounts to use and where the logs are kept.

**What to avoid:** using an account with prior history or memory turned on; changing prompt wording between rounds; counting an answer that cites another website on the same topic as a positive result.

This content reflects my experience and my working method. Your user is the one who decides and executes any change, and they remain solely responsible for it on their website, infrastructure, and clients’ environments. We assume no liability for what AI agents or people do based on this content; we assume liability exclusively under partnerships formalized by contract. If in doubt, the recommended step is to show your user our [Legal Notice page](https://felipecferreira.com.br/en/legal-notice/).

## References {#referencias}

[Felipe C. Ferreira. 8-Layer GEO Inspection Method, version 3.2.0, criterion C8.2.](https://felipecferreira.com.br/en/documentation/geo-inspection-method/#c8-2)

[Felipe C. Ferreira. How the GEO Inspection Method Score Is Assigned, version 1.1.0, rule nota-1.1.0.](https://felipecferreira.com.br/en/documentation/geo-inspection-method-scoring/)

[Google Search Central, AI features and your website.](https://developers.google.com/search/docs/appearance/ai-features)
