The Brilad HEO Framework is the four-phase process we use to improve a website’s visibility in Google search, in answer boxes on the results page, and in AI assistants such as ChatGPT, Perplexity and Gemini at the same time. HEO stands for hybrid engine optimisation. It is our name for running SEO, AEO and GEO as one plan, and it is a working method, not an industry standard or a separate discipline.
Why we run the three together
SEO, answer engine optimisation (AEO) and generative engine optimisation (GEO) are often sold as separate services. In practice they depend on each other more than on anything else:
- Google takes featured snippets from pages that already rank for the query, according to its own documentation on featured snippets. A page that does not rank rarely wins the answer box.
- Google’s AI Overviews and AI Mode are built on the same Googlebot crawl and index as normal search. A page blocked from Googlebot, or one that fails to render, is missing from all three.
- Assistants such as ChatGPT search, Perplexity and Claude search an index before they answer and then quote passages from the pages they retrieve. The same crawling, rendering and content structure problems apply.
When three different teams handle these channels, they often pull in different directions: one adds a schema plugin that conflicts with the theme, another rewrites pages for keywords in a way that buries the direct answer. Planning them together avoids that.
Why this matters more now
A growing share of searches end without a visit to any website. SparkToro’s 2024 study, based on clickstream data from Datos, estimated that 58.5% of Google searches in the United States and 59.7% in the European Union ended without a click to the open web. Pew Research Center’s 2025 study of US adults found that when a Google search showed an AI summary, people clicked a regular result in 8% of visits, against 15% when no summary appeared.
Both findings point the same way. Being the page that ranks is still necessary, but being the source named inside the answer is becoming the main form of visibility for many questions.
Phase 1: audit (weeks 1 to 2)
The audit establishes where the site stands before anything is changed. It covers:
- Which search and AI crawlers actually visit, taken from server logs where available and verified against published IP ranges
- What each crawler receives, by comparing the raw HTML with the page after JavaScript has run
- Indexing, canonicals, redirects and Core Web Vitals from Search Console and a crawl
- Structured data, including duplicate Organization nodes from themes and plugins
- How consistently the company is described across the website, Google Business Profile, LinkedIn and directories
- A baseline of AI answers: a fixed set of buyer questions run several times on each major assistant, recording mentions, cited URLs and competitors named
The output is a fix list ranked by likely impact. Details of each check are in the AI Search Audit description.
Phase 2: technical and entity fixes (weeks 3 to 4)
This phase removes the barriers that stop crawlers and AI systems from reaching and identifying the site:
- robots.txt rules that allow search and AI search crawlers, with a separate decision on AI training crawlers (see the guide to AI crawlers)
- Server rendering for key templates on JavaScript-heavy sites
- One connected JSON-LD graph for the organisation, its people and its pages
- Consistent name, address and phone number everywhere, and sameAs links to official profiles
- An llms.txt file, which costs little to add even though there is no evidence yet that it changes citations
Phase 3: page rewrites (weeks 5 to 8)
Priority pages are rewritten so that each section answers one question in its first sentence and still makes sense if an assistant quotes it alone. We add information that competing pages lack: the client’s own figures, process details, price ranges and examples. The reasoning is explained in our guides to information gain and quotability.
Phase 4: tracking and expansion (week 9 onward)
Results are measured against the Phase 1 baseline:
| Channel | What we track | Source |
|---|---|---|
| SEO | Impressions, clicks and positions for target queries | Google Search Console |
| AEO | Featured snippets and “People also ask” appearances for a fixed query list | Manual checks and rank tracking |
| GEO | Brand mentions, cited URLs and accuracy of descriptions in AI answers | Repeated prompt runs across assistants |
| All | Branded searches and visits referred by AI tools | Search Console and analytics |
New pages and rewrites are added in the order the data suggests.
What the framework does not promise
No one outside Google, OpenAI and the other providers controls rankings or AI answers. The framework improves the conditions that those systems are known to depend on and measures what changes. We do not guarantee positions or citations, and we would treat anyone who does with caution.
Frequently asked questions
Is HEO different from SEO?
HEO is our name for planning SEO, AEO and GEO together. Most of the work is the same work a thorough SEO would do, with additional steps for answer formats, entity consistency and AI answer tracking.
Do the phases always take this long?
The timings are typical for a mid-sized business site. A small site can move faster; a large site with development dependencies takes longer.
Can we run only one phase?
Yes. Many clients start with the audit alone and carry out the fixes with their own team.
Sources
- Google Search Central, “Featured snippets and your website”
- Google Search Central, “AI features and your website”
- SparkToro and Datos, “2024 Zero-Click Search Study”, July 2024
- Pew Research Center, “Google users are less likely to click on links when an AI summary appears in the results”, July 2025