Skip to content
New: the dataset now includes 2,128 company profiles and every supplied review record. See the methodology →

Step 01 of 4

Two datasets, kept separate

Company profiles and review records are supplied as two distinct workbooks. Neither is edited to make a company look better or worse — fields are mapped as-is into structured profile sections.

How It Works

The process behind this step

01
Two source workbooks
Company profile data and review/rating data arrive as two distinct workbooks and are mapped independently of each other.
02
Field-level mapping
Each field is mapped as-is into structured profile sections — nothing is rewritten, embellished, or quietly dropped because it looks unfavorable.
03
Review-only profiles
A company that appears only in the review dataset gets a clearly labeled review-only profile rather than a fabricated company page.
04
No manual curation
Inclusion is determined by dataset presence, not by an editorial judgment call about which companies "deserve" a profile.
What this prevents
Cherry-picked companies
Marketing copy replacing factual fields
Silent exclusion of unfavorable data