Automated review request system (20 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).

21 of the 44 named sources that could be read published a usable figure; 23 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; 207 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): the published price per location per month for every vendor that publishes one, normalised across three pricing models that nobody compares side by side - the per-location software seat, the per-message or per-invite allowance (SMS and email metered on top of the seat), and the managed agency retainer - giving the median, minimum and maximum per model, the share of vendors publishing a price at all against quote-only, the real first-year bill once setup fees and annual-commitment discounts are added, and repriced-since the previous read; the setup fee is the figure the category hides and it is read from the same page as the headline price.

A one-time flat package and a results-based fee (pay only after the agreed removal) are recorded as their own pricing models beside the monthly retainer rather than forced into a per-month figure, so the bare 'reputation management' buyer sees all three shapes a managed engagement is sold in.

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