FERNANDOANGULO.COM AI SEARCH · 2025
Practical Framework

The AI
Visibility
Framework

A step-by-step system for brands navigating the shift from search rankings to AI citation.

Fernando Angulo
Senior Market Research Manager, Semrush (an Adobe company)
Global Keynote Speaker  |  50+ Conferences  |  35+ Countries

What you will walk away with
A diagnostic to assess your current AI visibility
The 4-pillar GEO implementation system
An AEO content checklist you can apply this week
A proxy measurement dashboard you can build today
74%

of branded search queries now trigger an AI answer before organic results. Most brands have no system to respond.

— Semrush Research 2025

Based on Semrush global search behavior research fernandoangulo.com
THE AI VISIBILITY FRAMEWORK fernandoangulo.com
How to use this document

Read it as a system,
not a summary.

This is not a research report. Each section builds on the previous one. Start with the diagnostic to understand where you stand. Then work through the four stages in order. Use the checklists at each stage before moving on.

The Five Stages
01
Diagnose

Assess your current AI visibility and identify where you are losing ground.

02
Restructure

Rewrite and restructure your content for machine extraction.

03
Signal

Build the authority signals that AI systems use to decide who to cite.

04
Answer

Build a question-first content architecture that captures AI query intent.

05
Measure

Set up the proxy measurement system to track AI visibility over time.

Fernando Angulo  |  Semrush 1  ·  © 2026
THE AI VISIBILITY FRAMEWORK fernandoangulo.com
Stage 01 Diagnose

Diagnose Your AI Visibility

Before optimizing anything, you need to know your current position across three dimensions: whether your content is being extracted, whether you are being cited, and whether your authority signals are strong enough to compete.

The Core Problem

Most organizations have no visibility into which tier their content sits in for any given query. That absence of measurement is itself a strategic risk. You cannot fix what you cannot see.

The Three Visibility Tiers
TierStatusWhat It MeansPriority
1 Cited Source AI extracts and attributes your content with a link or brand mention. Protect and expand
2 Silent Extraction Your content is used but not attributed. No credit, no traffic. Force attribution via schema + E-E-A-T
3 Not Extracted Your content is ignored. A competitor fills the gap. Full GEO restructure required
Stage 01 Checklist — Complete before moving to Stage 02
  • Run your 20 highest-traffic keywords through Google, Perplexity, and ChatGPT. Record whether your brand appears.
  • For each query where you rank on Google but do not appear in AI answers, flag as Tier 3.
  • For each query where your content is paraphrased without attribution, flag as Tier 2.
  • Audit your top 10 pages: do they state facts explicitly with sourced claims in the first 100 words?
  • Check whether your site has Article, Organization, and Author schema implemented.
  • Verify your author bylines link to a professional profile with external mentions.
Fernando Angulo  |  Semrush 2  ·  © 2026
THE AI VISIBILITY FRAMEWORK fernandoangulo.com
Stage 02 Restructure

Restructure for Machine Extraction

AI answer engines do not read your content the way humans do. They mine it for structured assertions, factual claims, and entity definitions. Content written for human narrative fails machine extraction. This stage fixes that.

The Four GEO Pillars
Pillar 1 · Explicit Factual Density

State the claim, then the source, then the implication — in that order, in the first sentence of each section. Vague language ("studies suggest", "it is believed") signals low confidence to extraction models. Replace every hedged claim with a precise, attributed assertion.

Pillar 2 · Structured Entity Definition

Every key concept, person, product, or organization in your content should be explicitly defined in a standalone sentence. AI knowledge graphs map named entities to attributes. If your content never defines what something is, the model cannot reliably attribute claims to you.

Pillar 3 · Schema Markup as Citation Signal

Implement Article, FAQPage, HowTo, Product, and Organization schema on every relevant page. Schema is not just an SEO tactic — it is a direct signal to AI crawlers about content type, authorship, and factual structure. Pages without schema are systematically deprioritized.

Pillar 4 · Content Hierarchy and Scannability

Use H2 and H3 headings that state conclusions, not topics. "Voice search adoption grew 43% in 2024" extracts cleanly. "Voice search trends" does not. Every section heading should be a complete claim that stands alone as a citable fact.

Stage 02 Checklist — Complete before moving to Stage 03
  • Rewrite the first sentence of every major section to lead with an explicit, sourced claim.
  • Add a one-sentence definition for every named entity (brand, product, person, concept) on the page.
  • Implement Article schema with author, datePublished, and organization fields on all content pages.
  • Rewrite H2/H3 headings to state conclusions rather than announce topics.
  • Remove or rewrite any section that relies on implied context or assumed reader knowledge.
  • Add FAQPage schema to any page that contains Q&A content.
Fernando Angulo  |  Semrush 3  ·  © 2026
THE AI VISIBILITY FRAMEWORK fernandoangulo.com
Stage 03 Signal

Build Your Authority Signals

AI systems learn who to trust from the same signals that build human credibility: institutional affiliation, third-party citation, consistent publication, and documented expertise. These signals cannot be faked or fast-tracked — but they can be built systematically.

E-E-A-T as an AI Extraction Variable
Experience

Document first-hand involvement. AI models weight content from practitioners over commentators. Add case studies, personal data, and direct observations that cannot be generalized from elsewhere.

Expertise

Make credentials explicit and linkable. Author bios must reference institutional roles, publication history, and verifiable professional context — not just a job title.

Authoritativeness

Earn external citation. Conference speaker listings, industry publication bylines, podcast appearances, and institutional backlinks are the signals AI training data uses to rank sources.

Trustworthiness

Audit your technical trust layer: HTTPS, clear authorship, update dates on all content, and correction policies. AI systems deprioritize sources that cannot be verified as current and maintained.

Stage 03 Checklist — Complete before moving to Stage 04
  • Create or update a detailed author page with institutional affiliation, credentials, and external links.
  • Add publication and last-updated dates to all content pages.
  • Identify 5 external publications where you can place a byline or data citation this quarter.
  • Add Organization schema with sameAs links to LinkedIn, Wikipedia, Wikidata if applicable.
  • Request that conference and event organizers add your speaker page with a link to your site.
  • Add first-person data or proprietary observations to your top 10 pages.
Fernando Angulo  |  Semrush 4  ·  © 2026
THE AI VISIBILITY FRAMEWORK fernandoangulo.com
Stage 04 Answer

Build Your Question Architecture

Users addressing AI interfaces do not type keywords. They ask full questions in natural language. Answer Engine Optimization (AEO) is the practice of restructuring your content around the questions your audience actually asks — not the keywords they used to type.

The Question Architecture Process
1
Map Your Question Universe

Use Semrush Keyword Magic Tool with the question filter, run your core topics through "People Also Ask" on Google, and prompt ChatGPT with "What are the 20 most common questions someone asks about [your topic]?" Compile a master question list by intent category.

2
Prioritize by AI Extraction Likelihood

Sort your question list by intent: informational questions have the highest AI extraction rate and should be addressed first. Commercial investigation questions are second priority. Assign one content page or section to each high-priority question.

3
Write the Answer in the First 100 Words

Every question-targeted section must contain the direct answer within the first 100 words. The format is: state the question as an H2 or H3 heading, then answer it in one to three clear sentences, then expand with supporting evidence. Never bury the answer.

4
Implement FAQPage Schema

Every page that answers structured questions needs FAQPage schema with the question and answer text matching the on-page content exactly. This is one of the highest-ROI schema implementations for AI citation because it tells the model exactly what question you answer.

Stage 04 Checklist — Complete before moving to Stage 05
  • Build a question map of at least 30 natural-language questions for your core topics.
  • Assign each high-priority question to a specific page or section.
  • Rewrite section headings on priority pages to mirror question phrasing.
  • Verify that every targeted question has a direct answer in the first 100 words of its section.
  • Implement FAQPage schema on all pages with Q&A structure.
  • Test your priority questions in ChatGPT, Perplexity, and Google AI Overview. Record results.
Fernando Angulo  |  Semrush 5  ·  © 2026
THE AI VISIBILITY FRAMEWORK fernandoangulo.com
Stage 05 Measure

Measure Your AI Visibility

Native AI attribution data does not yet exist in standard analytics tools. Until it does, you need a proxy measurement system — four signals that correlate reliably with AI citation presence and allow you to track progress over time.

Your Four Proxy Signals
Signal 1 · Branded Search Volume Trend

What it measures: Whether AI mentions of your brand are generating downstream search behavior.
How to track: Google Search Console branded query volume, month over month.
What to look for: Rising branded search volume not explained by paid campaigns or PR events is a strong proxy for AI citation driving direct discovery.

Signal 2 · Rank-to-CTR Ratio Degradation

What it measures: Whether AI interception is reducing clicks even as rankings hold.
How to track: Google Search Console — compare average position against CTR for the same queries, tracked monthly.
What to look for: CTR dropping more than 15% on stable-ranking queries signals active AI interception.

Signal 3 · Direct Traffic to Deep Pages

What it measures: Whether AI answers are driving users directly to non-homepage URLs.
How to track: Google Analytics — direct traffic channel, filtered to pages 3+ levels deep.
What to look for: Any deep page receiving consistent direct traffic with no referral source.

Signal 4 · AI Mention Audit Frequency

What it measures: Whether your brand and content are appearing in AI-generated answers.
How to track: Monthly manual audit — run your 20 core queries through ChatGPT, Perplexity, and Google AI Overview. Log cited/uncited/absent for each.
What to look for: Trend from absent to cited over 3–6 month windows.

Stage 05 Checklist — Complete to establish your measurement baseline
  • Set up a branded keyword volume tracker in Google Search Console. Export monthly.
  • Build a CTR vs. Position comparison report for your top 20 queries. Update monthly.
  • Create a direct traffic segment in Google Analytics filtered to pages 3+ levels deep.
  • Build a spreadsheet: 20 queries × 3 platforms (ChatGPT, Perplexity, Google AIO) × cited/uncited/absent.
  • Run your first AI mention audit this week and record the baseline.
  • Schedule a monthly review — set a calendar reminder to run the full audit on the same day each month.
Fernando Angulo  |  Semrush 6  ·  © 2026
THE AI VISIBILITY FRAMEWORK fernandoangulo.com
Quick Reference

The AI Visibility Framework at a Glance

StageActionKey OutputDone When...
01 Diagnose Audit AI presence across core queries Tier classification per query Every priority query is classified
02 Restructure Apply the 4 GEO pillars to content Schema + factual density per page Top 10 pages fully restructured
03 Signal Build E-E-A-T and authority signals Author page, external citations, schema All 6 checklist items complete
04 Answer Build question-first architecture Question map + FAQPage schema 30+ questions mapped and answered
05 Measure Set up 4 proxy signal trackers Monthly measurement dashboard First baseline recorded

Hear This Live

This framework is built on ongoing Semrush search behavior research. If you want to see the full data, live demonstrations, and category-specific applications — invite Fernando to speak at your event or conference.

fernandoangulo.com
Fernando Angulo  |  Semrush 7  ·  © 2026