HEYTONY FOUR-REPORT EVIDENCE LIMITS — 20 October 2026

SEO score breakdown
Source: 33 passed + 8 warnings + 2 failed = 43
Finding: Arithmetic reconciles to 43; the score 86/100 cannot be independently recalculated because category weights and scoring formula are not supplied.

SEO false positives
Source: Privacy missing; only four internal links
Finding: Both claims conflict with the recorded independent site checks. The privacy endpoints responded HTTP 200, and the homepage had 52 anchors to 26 distinct internal URLs at test time.

Backlink gap universe
Source: 500 competitor links and 500 claimed gaps
Finding: The PDF prints 100 domain/homepage pairs, not 500 verifiable referring-page links. Neither the 500-link total nor absence of links to AIWebPageSEO can be reconstructed from those rows.

Backlink evidence completeness
Source: 100 displayed domains
Finding: Domain names and homepages are discovery leads, not a verified backlink graph. No exact referring page, outbound anchor, competitor target or live checked timestamp is supplied for any displayed row.

Unlinked mentions arithmetic
Source: 300 found, 193 already linking, 106 unlinked
Finding: 300 − 193 = 107, not 106. These figures cannot be reconciled from the PDF; exclusions or overlap could explain the difference but are not documented. The 27 “backlinks checked” may measure something different and should not be subtracted without a definition.

Unlinked mentions entity
Source: Search query is “sass”
Finding: The report heading itself establishes the search term is not AIWebPageSEO. Sass-related results are not evidence of a mention of the audited brand. Individual page verification is still necessary.

Unlinked mentions URL coverage
Source: 106 rows
Finding: 98 printed URL strings are complete and eight are truncated. Truncated URLs cannot be used as reproducible evidence; the report should export full permalinks.

Reddit topic sourcing
Source: 40 main titles and 10 bonus ideas
Finding: No post-by-post source citations appear beside the topics. Independent limited-sample review found 5 direct, 18 partial/contextual and 17 without relevant sample evidence. The 35 disputed cases are not proof of invented questions.

Audit verification boundary
Source: Four supplied PDFs and preserved independent review files
Finding: This audit checks the reports, their arithmetic, their printed evidence, and earlier independently recorded tests. It does not represent a new crawl of all live referring pages or Reddit posts. Missing verification remains an open test, not a proven negative.

This report does NOT assert that every external page has been independently accessed or verified. A source URL alone is not proof of the claimed link or mention.