ELI RESEARCH
The Eli AI Search Index
Fix the questions, save the answers, expose methodology and measure the same thing repeatedly.
The first Eli AI Search Index benchmark is currently being expanded across a fixed set of commercial B2B buyer questions. Methodology and field definitions are public now. Results will only appear after the observation set is complete and quality-checked.
RELEASE STANDARD
Methodology is publishable before findings are.
Eli can publish the question universe, schema, definitions and quality controls before a benchmark release. It will only publish an aggregate when the underlying observations are complete, reviewable and described with their coverage and limitations. Until then, this page makes no performance, ranking or customer-outcome claim.
What the Index measures
Recommendation rate, mention rate, citation rate, source coverage, cross-engine coverage and recommendation stability are related, not interchangeable.
The prompt universe
The fixed commercial prompt universe is versioned. Each prompt has ID, intent, paraphrase version, language and market. Changing it creates a new benchmark version.
Every observation stores raw evidence
Store prompt, engine, model/product surface where known, date, raw answer, brands mentioned/recommended, recommendation context, citation URLs/domains, market, language and notes. Aggregates trace back to observations.
What the Index does not claim
It is not universal ChatGPT rank, revenue predictor, proprietary ranking-system explanation or proof one source caused a recommendation.
Versioning
Every public release includes version, field dates, prompt/observation count, engines, markets/languages, methodology version, dataset link and change log.
FAQ
Questions about the research design and release boundary.
What is the Index?
Fixed-method research on B2B recommendation, mention and citation behavior.
Is it a ranking?
No. It measures repeated outcomes, not fixed position.
How many prompts?
Publish exact count only from current release metadata.
Can score be manipulated?
Versioned prompts and methodology reduce opportunistic changes.
Will raw data be public?
Inspectable tables will be published where legally and operationally appropriate.
CONNECT YOUR WEBSITE
See how your brand appears in AI search
The Eli AI Search Index is a research program measuring how often B2B companies are mentioned, recommended and cited across a fixed set of commercially relevant AI-search questions. It tracks recommendation rate, mention rate, citation behavior, source coverage, cross-engine overlap and recommendation stability.
See how your brand appears in AI search