Top 100 Catering Companies in the San Francisco Bay Area — Graded

How the AI Score works

Every company carries an AI Score from 1–100, computed by one published formula from the signals measurable today — legitimacy and online action, verified across at least two independent sources (open place datasets and public registries). The formula's other factors — blended reviews, complaint record, and responsiveness — carry no weight in the score until licensed data sources for them exist. When they land, the score re-spreads across the full formula.

One disclosed exception: a score of 100 is reserved for RankMyCompany clients; every other score is computed by the same formula for everyone. Scores are not spread artificially — companies with identical measured data share identical scores — and each listing also shows its rank and its percentile within its category.

Missing-data rule: when a factor has no lawful data source for a company yet, its weight redistributes proportionally across the factors that do have data, and every score shows its data-coverage %. Coverage rises as licensed sources are added; scores are recomputed monthly and score history is kept for every company.

No score is ever adjusted by hand, sold, or removed for payment. Companies marked RMC Client are scored by the exact same formula as everyone else — the 100 cap above is the only exception, and it is disclosed here rather than hidden. If your company's underlying data is wrong, file a correction on the dispute page — disputes fix the data, and the corrected data flows into the next ranking run.

Place data: © OpenStreetMap contributors (ODbL) and Overture Maps Foundation (CDLA Permissive 2.0). Listings are refreshed monthly; a score change from a refresh takes effect only after the new score holds for two consecutive runs. Formula changes are published as a new methodology version — never applied silently.

Read the full standard — The RMC Standard →