01 / SEO
Discovery
Can a search engine discover, render and index the right page? We check access, canonical URLs, internal links, content and the declared crawling policy.
QUAERION RESEARCH / 2026.10
Answer Intelligence.
From clear information to verifiable visibility.
Quaerion is a Brazilian Answer Intelligence company. We connect SEO, AEO, AIO and GEO through AI Referral Engine™: a practical framework for making brands easier to find, understand and evaluate.
01 / THE FRAMEWORK
Four disciplines. A connected method.
Answer Intelligence is Quaerion’s organizing framework for this work, rather than a certification or a standard imposed by search platforms. The definitions below describe our service scope. In our terminology, AIO means AI Optimization; Google AI Overviews is a separate product feature.
02 / AEO
A useful answer
Does the page answer a real question with enough context to remain understandable? We review the answer, its evidence, conditions and limits.
03 / AIO
Entity understanding
Is the company consistently described across its official pages and attributable sources? We reconcile names, offerings, ownership and structured data.
04 / GEO
Generative visibility
Which sources and brands actually appear in a defined set of generative responses? We preserve the query, engine, collection date, response and cited URLs.
02 / AI REFERRAL ENGINE™
Map. Structure. Position.
The method operates through Diagnosis, Strategy, Execution and Monitoring. A diagnosis establishes the initial state; strategy prioritizes the audience and information gaps; execution improves the agreed pages and assets; monitoring checks whether the work changed the measured outcome.
- Map the entity and demand: official identity, actual offerings, buyer questions, relevant markets and attributable sources.
- Structure the evidence: stable URLs, readable HTML, ownership, dates, definitions, source links and consistency between the page and its metadata.
- Position with useful work: publish original explanations and documented deliveries, improve distribution and measure a predefined query set.
An implementation record shows what we changed. A source citation shows where a response obtained information. A recommendation shows the direction of the response. These observations require different evidence and should remain separate in reporting. Full operational method (PT) ↗
03 / MEASUREMENT
A score needs a denominator.
Before collecting results, define the questions, language, region, models or product interfaces, search settings, repetitions and valid-response rules. Keep unbranded questions separate from prompts that name the company. A failed response is missing data, not a measured zero.
| Observation | Required evidence | Interpretation |
|---|---|---|
| Technical access | HTTP, canonical, robots, HTML, JSON-LD | Published and retrievable; does not prove indexing. |
| Brand mention | Preserved response and query; valid responses as denominator. | Name appeared; may be neutral, negative or prompted. |
| Source citation | Visible cited URL, engine and collection time. | That source was cited in that collected response. |
| Business outcome | Attributable analytics or CRM evidence, period and baseline. | Observed visits, leads or sales under the stated attribution rules. |
An API model is not automatically equivalent to the consumer ChatGPT, Gemini or Perplexity interface. Versions, retrieval, personalization and interface behavior can differ. Cross-engine percentages require explicit comparability checks.
04 / PLATFORM GUIDANCE
Useful foundations, with clear limits.
Google’s current guidance keeps useful content and sound SEO at the foundation of generative search. It does not require special AI markup or llms.txt. We use structured data to describe supported facts and Markdown to offer an additional readable representation; neither artifact guarantees citation.
Platform reports also have distinct scopes. Search Console’s Generative AI report measures visible link impressions from AI Overviews and AI Mode. OpenAI documents separate roles for its search, training and user-initiated crawlers. Respecting those distinctions improves both implementation and interpretation.
- Google Search Central — Optimizing for generative AI search · Official platform guidance.
- Google — Generative AI performance report · Measurement scope and limitations.
- OpenAI — Crawlers and bots · OAI-SearchBot, GPTBot, ChatGPT-User.
- Aggarwal et al. — GEO: Generative Engine Optimization · Research experiment, not a universal commercial uplift.
This is Quaerion’s synthesis of its operating method and the cited documentation, reviewed on 4 October 2026. It does not establish worldwide leadership or a guaranteed position in an AI answer.