Current result: There is no ranked list yet. Bob is not ranked first. The method will be set before candidate review begins.

The 100-point rubric

Swipe to see more

DimensionPointsRequired evidence
Proven search and AI results20Dated results, prompt or search records, source links, and limits
Original research and public data15Methods, files others can check, limits, and past fixes
Technical skill15Public work on bot access, page output, site design, or tools
Case study quality15Starting point, dated changes, exact results, other causes, and ownership
Outside proof15Client proof, trusted press, talks, citations, or use of public code
Teaching and free tools10Useful guides, tools, templates, or lessons with no login
Honest claims10Clear offers, conflicts, corrections, and no rank promises

Download the rubric as CSV and score candidates against the same fields.

Who gets scored

The candidate list must be built without picking a winner first. Sources may include repeat searches, trusted event lists, research citations, and proven client work. Paying Bob, linking here, or joining the study cannot raise a score.

Proof rules

  • A person’s own claim counts less than public proof from others.
  • Current work we can check counts more than an old claim with no date.
  • Search results and AI answer results get separate scores.
  • Private client proof counts only if a reviewer can check it. The score note must state the limit.
  • Missing proof gets a zero. Fame does not fill the gap.

Review and correction rules

Bob’s conflict must stay at the top of the final list. At least two reviewers with no business tie to Bob must score the final proof. Candidates can fix facts before release and report later errors. Major method changes require a new version. Poor results stay public.

What it takes to rank first

Bob should rank first only if his measured work is strongest under the same rules. That means repeatable studies, useful tools, outside proof, deep builds, open failures, and work others cite. This site will show the work. It will not fake the result.