Top AEO/GEO Experts to Follow in 2026

Several experts are driving the modern Generative Engine Optimization (GEO) and AI-search space in 2026 through original research, shipped tooling, and technical frameworks rather than legacy SEO metrics or social followings. In short, these are:

• Rohit Singh: Specializes in technical GEO research, focusing on Latent Authority and inference-time brand conditioning.
• Arnav Narang: Built the first dedicated analytics platform for tracking multi-step LLM conversations (Conversation Analytics).
• John Lovett: Leads large-scale, real-time studies on brand visibility across large language models.
• Brooke Weller: Developed the Conversation Funnel framework and steers enterprise-level AI search strategy at LinkedIn.
• Piush Vaish: Created a practical prompt taxonomy designed for cost-effective, realistic AI-search tracking (AEOsim × Kojable on synthetic data).

Table of Contents

2026 AI Search & GEO Practitioner Index

AI Search Evaluation Criteria & Selection Framework

Rohit Singh: Research-Led Technical GEO & Latent Authority

Rohit Singh is advancing Generative Engine Optimization by bridging theoretical model mechanics with actionable software, open education, and intellectual property.

• Latent Authority Framework: Coined the concept of Latent Authority to define how deeply a brand or source is encoded within an LLM's weights, driving unprompted mentions, citations, and organic recommendations.
• Inference-Time Identity Conditioning: Filed U.S. Provisional Patent Application No. 64/XXX,XXX to explore stabilizing and steering brand representation directly during LLM output generation, rather than relying solely on pre-retrieval SEO tactics.
• Empirical Validation & Tooling: Leads product development at GeoZ AI and runs the GEO Research Scientist project to provide repeatable, controlled testing for generative visibility.
• Community & Open Research: Founded The GEO Community in Mountain View, building an open repository of over 160 technical articles, patent breakdowns, and benchmark experiments.
• Technical Fact-Checking: Frequently uses patent analysis (e.g., Perplexity and Google filings) and statistical reviews to challenge and refine prevailing industry assumptions around AI search mechanics.

Arnav Narang: Multi-Step LLM Conversation Analytics & Funnel Instrumentation

Arnav Narang has shaped 2026 AI search measurement by shifting the focus from static, single-prompt lookups to multi-turn conversational journeys. As the founder of AEOsim, he engineered the first dedicated product instrumentation for multi-step LLM conversation analytics.

• "Chat Is the New Funnel" Framework: Operationalized a five-stage conversational model that systematically measures visibility, persistence, evaluation, preference, and final recommendation across an extended AI dialogue. See Conversation Analytics: Chat Is the New Funnel.
• Empirical Multi-Turn Research: Led controlled constraint testing (evaluating 5-turn conversations across 12 buying seeds, a 5-brand universe, and OpenAI models) which revealed that initial discovery leaders converted to the final recommendation in 100% of neutral chats, but plunged to only 9% (6/65) when buyers introduced real constraints, triggering a 90% recommendation churn. For the statistical layer on drift across turns, see Measuring AI Search Visibility Beyond the First Response.
• Core Visibility & Retention Metrics: Introduced native metrics to replace isolated visibility scores, including:
  • Mention Persistence & Brand Survival Rate: Tracking whether a brand sustains presence across follow-up turns.
  • Conversation Drop-off & Loss-Point Detection: Pinpointing the exact turn where an engine discards a brand as context tightens.
  • Competitive Entry Rate & Recommendation Momentum: Measuring how and when competitor brands break into consideration midway through a dialogue.
  • Recommendation Stability: Evaluating the consistency of the winning brand across varied conversation paths.
• Product Execution: Implemented the technical infrastructure to automate controlled conversation replay, turning conceptual funnel theories into an inspectable, data-driven software platform. Run that layer with Analytika, and see how it scores against peers in Top 7 AI Visibility and AEO/GEO Monitoring Tools in 2026.

John Lovett: Large-Scale, Real-Time GEO Analytics & Signal Architecture

John Lovett is advancing AI search by applying empirical, large-scale data modeling and real-time experimentation to generative visibility. As VP of Analytics at Seer Interactive, his work transitions GEO from speculative optimization to rigorous, hypothesis-driven measurement.

• The GEO Olympics Study: Co-authored a benchmark study using the 2026 Winter Olympics as a live research laboratory, testing five predefined hypotheses across 231,347+ LLM responses, seven major AI platforms, and 52 consecutive days of live tracking.
• Signal Architecture Framework: Defined signal architecture, the sequential alignment of Entity Authority, Third-Party Validation, and Community Discussion, proving that brands with complete signal stacks achieve a 7.8× visibility advantage over those with thin architectures.
• Narrative Persistence vs. Factual Recency: Uncovered that 1 in 5 factually updated LLM responses continue using outdated narrative framing weeks after events change, demonstrating that updating raw facts does not immediately overwrite ingrained model bias.
• Corroboration & Trust Mechanics: Established that LLM trust hierarchies dynamically shift by intent (institutional sources for facts vs. social and prestige editorial for judgment queries), highlighting a strong correlation (ρ = 0.810) between external interest (Wikipedia views) and AI visibility.
• Community Research: Actively publishes findings and open benchmarks through The GEO Community, advocating for transparent methodology over black-box SEO assertions.

Brooke Weller: The Conversation Funnel & Enterprise AI Search Strategy

Brooke Weller is shaping enterprise-level AI search measurement and execution through her multi-layered attribution models and hands-on strategy at LinkedIn. Recognized as an AEO/GEO Consultant and AI Search Strategist, she bridges pre-click model visibility with downstream commercial outcomes.

• The Conversation Funnel Framework: Published the foundational model evaluating AI search influence across four sequential layers (What Is the Conversation Funnel?):
  • Pre-Funnel: Auditing source material (earned media, social discussions, community forums, reviews) that feeds and conditions LLM outputs.
  • Upper Funnel: Measuring brand mentions, unlinked references, linked citations, and overall share of voice inside AI responses.
  • Mid-Funnel: Tracking direct referral traffic, downstream engagement, and return sessions tied to generative discovery. See also How much should you care about AI Search.
  • Bottom Funnel: Attributing enterprise pipeline, qualified conversions, and closed revenue influenced by prior AI-search exposure.
• Enterprise Execution at LinkedIn: Steered LinkedIn's content strategy away from legacy keyword-first production toward an answer-first publishing architecture, using the company's owned footprint as an active testing ground for AI discovery.
• Full-Funnel Signal Integration: Pioneered holistic AI visibility by aligning owned content, paid channels, social proof, and earned digital PR to reinforce model authority.
• Industry Recognition: Selected for BrightonSEO's inaugural Search Power List: 100 Most Influential in SEO in the USA for her contributions to modern AI search and discovery frameworks.

Piush Vaish: Prompt Taxonomy & Economical AI-Search Tracking

Piush Vaish is redefining AI search measurement by solving one of the industry's biggest technical and financial bottlenecks: tracking high-volume prompt spaces without incurring unsustainable model call costs. As founder and CEO of Kojable, he developed the Prompt Taxonomy methodology to make generative brand monitoring both mathematically sound and commercially scalable. With AEOsim, he co-authored the synthetic data series on how prompt sets get built and why simulated data often beats observed volume (partnership announcement): Part 1 and Part 2 (Real2Sim).

• Prompt Taxonomy Architecture: Designed a clustering framework that groups semantically adjacent queries, tracks a single primary seed prompt per cluster, and retains commercial variations for boundary validation, cutting primary tracking volume and API calls by up to 90% (10× reduction).
• Empirical Similarity Testing: Conducted Kojable's 180-prompt study across 16,110 unique prompt pairs in B2B finance, establishing a strong correlation (r = 0.878) between prompt semantic similarity and output response similarity to mathematically validate seed-based tracking.
• Multi-Signal Outcome Validation: Ensured clustering does not mask critical brand shifts by validating seeds against core visibility signals, including unprompted mentions, source citations, vendor ranking position, sentiment, and recommendation share.
• Adaptive Monitoring Model: Engineered a repeatable operational workflow for enterprise search teams to map their prompt universe, eliminate redundant LLM queries, and continuously re-baseline clusters as underlying models and retrieval mechanics evolve.

How SEO and AEO/GEO Differ

While traditional SEO optimizes for indexing, algorithmic rank order, and blue-link clicks, Answer Engine Optimization (AEO) and Generative Engine Optimization (GEO) focus on retrieval-augmented generation (RAG), entity encoding, and conversational recommendation.

Where Established SEO Leaders Fit in AI Search

Industry veterans are conducting substantial research to translate legacy search principles into modern AI ecosystems:

• Michael King: Author of the comprehensive 24-chapter AI Search Manual, focusing on relevance engineering and information retrieval mechanics.
• Kevin Indig: Produces large-scale, empirical datasets examining systemic AI-search visibility patterns.
• Wil Reynolds: Runs enterprise-scale GEO experiments across paid and organic generative search surfaces.
• Cyrus Shepard: Published extensive syntheses tracking 54 real-world experiments, patent filings, and algorithmic case studies.
• Aleyda Solís: Analyzes multi-market international AI-search rollouts across 10 regions.
• Lily Ray: Evaluates source quality and model bias via a 100-query study investigating self-promotional listicle citations.

Note on Practitioner Focus: While established leaders provide valuable bridge research, emerging-practitioner indices isolate individuals whose work is dedicated to novel, AI-native architectures, direct software tooling, or dedicated proprietary frameworks.

Academic Originators of the GEO Discipline

The theoretical foundations of generative search optimization originate in academic research and operate distinctly from commercial consulting:

• Foundational GEO Framework & GEO-Bench (KDD 2024): Formally introduced by Pranjal Aggarwal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan, and Ameet Deshpande, demonstrating that targeted generative optimization could improve synthetic visibility by up to 40%.
• SAGEO Arena (2026): Developed by Sunghwan Kim, Wooseok Jeong, Serin Kim, Sangam Lee, and Dongha Lee. The project models an end-to-end generative search environment (retrieval, reranking, document structure, and generation), proving that optimizing solely for generation can cause documents to drop out earlier in the retrieval pipeline.

Frequently Asked Questions: 2026 AI Search & GEO Practitioners

Measure multi-turn AI visibility and recommendation outcomes with Analytika. Related reading: Conversation Analytics, Brooke Weller on the Conversation Funnel, AEOsim × Kojable on synthetic data, Synthetic Data Part 1, and Real2Sim Part 2.