# Data Dictionary

## dataset_domain.csv

Domain-level aggregated dataset. One row per unique domain.

| Column | Type | Description |
|--------|------|-------------|
| domain | str | Registrable domain (e.g. `example.com`) |
| citation_rate | float | Fraction of queries where domain was cited (0.0–1.0) |
| queries_cited | int | Number of queries where domain was cited (majority vote) |
| total_queries | int | Total number of queries in the study |
| mean_consistency | float | Mean citation consistency across queries (0.0–1.0) |
| mean_google_rank | float | Mean best Google rank across queries (NaN if not ranked) |
| best_google_rank | float | Best (minimum) Google rank across all queries |
| total_score | float | AI Search Readiness Score (0–100) |
| machine_readability | float | MR sub-score (0–25) |
| extractability | float | EX sub-score (0–30) |
| trust_entity | float | TE sub-score (0–25) |
| offering_readiness | float | OR sub-score (0–20) |
| moz_da | float | Moz Domain Authority (0–100) |
| domain_age_years | float | Domain age in years (from WHOIS) |
| wikipedia_presence | int | 1 if domain has a Wikipedia article, 0 otherwise |

## dataset_domain_query.csv

Domain × query level dataset. One row per (domain, query) pair.

| Column | Type | Description |
|--------|------|-------------|
| domain | str | Registrable domain |
| query_id | int | Query identifier (1–30) |
| topic_cluster | str | Query vertical: `saas`, `ecom`, or `services` |
| times_cited | int | Number of replicates that cited this domain for this query |
| total_replicates | int | Total replicates per query (3) |
| citation_consistency | float | Fraction of replicates citing the domain (0.0–1.0) |
| is_cited | int | 1 if majority of replicates cited the domain, 0 otherwise |
| google_rank | float | Best Google rank for this domain on this query (NaN if unranked) |
| total_score | float | AI Search Readiness Score (0–100) |
| machine_readability | float | MR sub-score (0–25) |
| extractability | float | EX sub-score (0–30) |
| trust_entity | float | TE sub-score (0–25) |
| offering_readiness | float | OR sub-score (0–20) |
| moz_da | float | Moz Domain Authority (0–100) |
| domain_age_years | float | Domain age in years |
| wikipedia_presence | int | Wikipedia presence flag |

## Notes

- **is_cited** uses majority vote: cited in ≥50% of replicates → 1
- **citation_rate** at domain level = mean(is_cited) across all 30 queries
- **total_score** = machine_readability + extractability + trust_entity + offering_readiness
- Score formula weights are frozen per pre-registration (MR 25%, EX 30%, TE 25%, OR 20%)
- Domains appear in the dataset if they were cited by any LLM response OR appeared in any search result