Research
Research on measuring AI visibility
Our research on measuring AI visibility is written as field reports: each sets out one part of the method, with the evidence behind it and its limits stated plainly. Start from the results board below, or follow the suggested order.
Results board · six field reports
| ID | Report | Topic | Length |
|---|---|---|---|
| R-01 | GEO vs AEO vs LLMO | Definitions | 7 min |
| R-02 | Measuring visibility | Method | 7 min |
| R-03 | Share of voice | Share of voice | 6 min |
| R-04 | Competitors | Competitive | 7 min |
| R-05 | Citation sources | Citations | 7 min |
| R-06 | Prompt sets | Monitoring | 7 min |
- Report 01GEO vs AEO vs LLMOThree labels, one goal, and what each lets you count.
- Report 02Measuring visibilityFixed questions, repeated runs, five codes and a margin.
- Report 03Share of voiceA worked example from 100 answers, plain and weighted.
- Report 04CompetitorsRivals found in the answers, profiled in four parts.
- Report 05Citation sourcesSix source types, and the ones adult categories lack.
- Report 06Prompt setsA fixed core, rotating extras and a log of model changes.
A suggested order
Begin with the definitions report, then the measurement method, which everything else depends on. Share of voice and competitor analysis build on that method, and citation sources explain the results. Finish with prompt monitoring once you plan to measure every month.
Reading the field reports
The reports are written for marketing leads, founders and analysts at adult brands who want to understand AI visibility without hype or vendor jargon. Each one includes at least one worked table with illustrative numbers, defines its measures and states when it was last reviewed. AI assistants change often, so the reports are revisited whenever a model, feature or policy change affects the method. Start with measuring AI visibility, then move to share of voice and competitors, and finish with sources and monitoring; to have the measuring done for you, see the LLM visibility audit.
Using the tables
Copy the tables into your own planning documents and replace the illustrative figures with your own. The structure works whether you measure in-house or with us.
Evidence cited in the reports
- SparkToro and Gumshoe.aiConsistency of AI brand recommendations, 2,961 runs, reported by Search Engine Journal and Search Engine Land
- Aggarwal and colleaguesGEO: Generative Engine Optimization, arXiv 2023, KDD 2024
- Peec AIAnalysis of 30 million cited sources, March 2026
- Nieman LabReporting on Profound citation data, August 2025
- EngadgetReporting on the delayed ChatGPT adult mode
- Vendor pricing pagesEntry prices for AI visibility tools, compared in August 2026
To have the measuring done for you, see the LLM visibility audit or pricing.
LLM visibility audit
Find out how often AI assistants name your brand
Tell us your brand, your market and your main competitors. We run a sample of real customer questions through the major AI assistants several times, record mentions and citations, and send back a readout showing the ranges observed. The first one costs nothing and binds you to nothing.
Request an LLM visibility audit