The public record of who gets named, and who does not.
A page per vendor per category, pinned to the release it was measured in, with the count, the base and a 95 percent interval on every record. Free to read, free to dispute, free to claim.
Observed answers on stated dates. Not a ranking, and not a judgment of product quality.
The Index exists because the absence is the finding. A vendor that never appears in a category's answers has a problem that no amount of traffic reporting will show it, and until somebody writes that down in public, nobody can argue with it or correct it. The record is published so it can be argued with.
A page per vendor per category
Every checked vendor has a record in every category it was checked in, carrying the count of answers that named it, the base those answers came from, the percentage and its Wilson 95 percent interval. A record that a release did not supply reads as unavailable rather than as zero.
Each record is also pinned. The live path follows the current release; a release path holds the record as it stood in that month, permanently, so a figure quoted in a deck in August still resolves in December.
Who gets named
The explorer runs across the whole published release: every question, every checked vendor, the counts and their bases and intervals. It answers the only question a buyer in a category actually asks first, which is who the engines name when nobody mentions a brand.
Claims and badges
Disputes and re-runs
Any named company gets one re-run per release, under the same configuration, at no cost to it. Both results are published side by side, the original first, and the original is never replaced or removed. The dispute policy and the log of disputes are public, and only the date, the vendor, the category, the release and the status appear in the log.
Dispute a record or request a re-run. A company that has not claimed its page can still dispute a record by writing to index@broadcastwell.com.
What the record does not claim
The current release is 2026-08.1, deposited as 10.5281/zenodo.22721962, with 10.5281/zenodo.22150321 as the permanent all versions citation. Its configuration is stated on the record and is distinct from the current five engine method used in client work.
Capture dates in this release are the dates the sources were captured. Separate answer capture dates are not available and are not presented as though they were. Per engine counts are not inferred from combined counts, a missing dimension is published as unavailable, and no answer text from these releases was retained.
Questions
Is the Index a ranking?
No. It is a record of observed answers on stated dates. Companies are not ordered by merit and nothing in it is a judgment of product quality. A vendor that is not named in a category was not named in those answers, which is a fact about the answers rather than a verdict on the vendor.
How does a company claim its page?
Where the record carries a canonical domain, by proving control of that domain with a DNS TXT record, and no email verification is needed. Where the release supplied no canonical domain, the claim opens by email instead. A claimant never chooses the domain. A claim verifies control and nothing else: it does not change an observation, endorse a vendor or add a vendor's words to the record.
What can a company do once it has claimed?
Dispute the record, request its one re-run for the release, request an alias correction, and take the badge code for its own site.
What are the re-run terms?
Any named company gets one re-run per release, under the same configuration, at no cost. Both results are published side by side with the original first. A re-run never replaces the original.
What do the badges say?
A badge states an observation on a dated release and links to the record it came from. The named form prints the count, the base, the percentage and the 95 percent interval. Text names only, no logos. Do not alter a badge, and keep its link and its visible category and release month.
What does the Index not know?
Where a release did not supply a dimension, the record says so rather than printing a zero. Per engine counts are not inferred from combined counts, capture dates are the dates the sources were captured, and answer text from the historical releases was not retained.
Start with a measured baseline
See which buyer questions leave you out, the sources visible in those answers, and the three changes worth testing first. The Category Audit covers ten questions; the Diagnostic extends to 35 questions and 525 scheduled observed answers, adaptive runs extra. AI answers vary, so repeated observations and disagreements are reported.
Start with the $490 Category AuditResults appear in your account; a sign-in link arrives by email. Credits in full against the Diagnostic within 30 days.
Get the Diagnostic, $990Results appear in your account; a sign-in link arrives by email. Credits in full against the program.