Reputation management white label (6 sources)

Every row is a verbatim sentence read from the named source's own page through our fetch service on the date shown, hash-pinned to the stored read (the sha of the page's visible text; the raw page is mirrored to R2).

7 of the 21 named sources that could be read published a usable figure; 14 published none and are recorded as not published, never filled in; 1 could not be read (blocked, dead or unreachable) and count nowhere.

The headline figures (median, minimum, maximum, quartiles) are over ONE figure per source in USD (a source's median where its page yielded several accepted sentences), so N counts sources, never sentences; 33 candidate sentence(s) were refused by a reviewer with the reason recorded.

The fuller derivation this dataset is chartered to grow into (not yet what the figures above are; they are each source's own published figure as read): one row per programme joining the wholesale price the agency pays to what the rebrand actually covers - custom login domain, logo and colours, branded review-request email and SMS sender, branded reports, the agency's own Stripe billing, an API and help centre under the agency's name - scored the same way for every vendor, beside the seat minimum and setup fee that decide the real floor; the derived figures nobody publishes are the wholesale-to-retail margin per programme (wholesale seat price against the $500-$900 the market resells at), the share of programmes that rebrand the CUSTOMER-FACING surfaces rather than only the dashboard, and the number of client seats at which a per-seat programme beats a flat one

33 candidate sentences on these pages were refused by a reviewer; each refusal and its reason is recorded in the review ledger.