Market researcher & executive translator. I study how people—and AI systems—discover, evaluate and choose brands. @Semrush · Global keynote speaker.
Barcelona, Spain
- Tweets 22,062
- Following 4,890
- Followers 12,283
- Likes 16,072
Discovery is moving from ranked links to synthesized answers, citations and agent-mediated decisions.
I research how people—and AI systems—discover, evaluate and choose brands.
Current thesis: GEO is the new ranking. I research who gets cited.
fernandoangulo.com
1
97
At Future Ready Digital: North today, I’m discussing the discovery shift with one goal: give marketing leaders a defensible way to separate structural change from platform noise and decide what the team should build next.
1
112
One useful test for an AI-search strategy: can the team explain what would make the brand more likely to be mentioned, cited or recommended—and why those are different outcomes?
If not, the dashboard is ahead of the operating model.
175
Agent-ready content still has to persuade people. It also needs to be readable, attributable and actionable for machines.
The first audit is simple: can a system identify the entity, verify the claim and understand what action is possible?
132
Where AI gets built. And I’ll be speaking there.
I’ll be at AI Summit Barcelona, 22–23 September at the World Trade Center Barcelona.
If you’re going too, let’s connect.
#AISB26 #AISummitBarcelona #AI
1
2
141
AI visibility is cross-platform by definition. ChatGPT, Google AI and Perplexity do not select the same sources in the same way.
The decision is not “Which engine matters?” It is which audience, query set and outcome your measurement needs to represent.
1
89
Good research communication moves in three steps:
1. Identify the signal.
2. Explain the mechanism.
3. Name the decision.
Skipping step two creates hype. Skipping step three creates an interesting presentation that changes nothing.
68
Zero-click does not mean zero influence. It means traffic is no longer sufficient evidence of visibility.
Leadership reporting should separate reach, mentions, citations, referral traffic and the decisions each signal can support.
1
1
197
Entity optimization starts before the keyword.
If an AI system cannot resolve who you are, it cannot reliably connect your research, role and claims.
Use one canonical description across your site, profiles, bylines and schema.
130
Before publishing another page, ask a harder question: does this topic already have one canonical source strong enough to update?
In AI search, scattered expertise creates ambiguity. Consolidation often improves the signal more than volume.
158
Mentions and citations are not the same metric.
An AI answer can name a brand without linking to it, or cite a source without making the brand prominent.
Measure them separately before assigning one blended “AI visibility” score.
1
2
125
AI search did not make SEO irrelevant. It changed the unit of competition.
A page can rank without becoming a cited source.
The practical decision: keep technical SEO, but add entity clarity, attribution and extractable evidence.
1
185
Today at UK SEO Summit, I’m focusing on the question behind GEO: what makes a source clear, credible and useful enough for an AI system to cite?
The answer is not one tactic. It is a system of entity, evidence, structure and distribution.
71
We go live shortly.
The first stop on my 2026 AI Search & GEO calendar is about to begin: Winning AI Visibility with Stephen Adams and Emplifi.
Nine events. Seven cities. Four countries.
Here’s where we can meet ↓
1
1
1
120
Next stops: Leeds, Barcelona, Vilnius and London.
We’ll get practical about LLMs in the purchase journey, evidence for AI visibility, ecommerce in AI-led discovery and what 47,000+ marketers reveal about GEO.
Slides 5–8 ↓
1
1
54
Then: Digital Marketing Europe and the GEO Conference in San Francisco.
Save the thread and tell me which room—or screen—I’ll see you in.
Dates, cities and registration:
fernandoangulo.com/speaking
#AISearch #GEO
23
Most teams do not need another AI-search trend recap. They need a shared language for what changed, why the old metric is incomplete and which decision comes next.
That is the standard I use for every research briefing.
135
5/
Both are good. They answer different questions.
Big rooms give the leadership layer a frame. Small rooms give the practitioner layer a method.
1
22
6/
So I mapped the whole circuit.
28 events across the US and Europe. 16 still ahead this year. Dates, venues, prices and a screenshot of every landing page as I checked it.
Six are stages I am on. Two of the best cost nothing to attend.
fernandoangulo.com/blog/ai-s…
33