The 5W
Reputation Index
What AI Says About You When You're Not in the Room.
AI engines now answer the world's questions about people, companies, and institutions. The Reputation Index audits the answer. It measures how AI systems stabilize, distort, reinforce, or fragment public reputation — across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews.
Not a popularity ranking. A study of reputation architecture.
A growing share of consumers, reporters, investors, and recruits now begin their research inside an AI engine, not a search bar. By the time they reach a person or a brand, an answer has already been formed for them — synthesized, confident, and rarely questioned.
The 5W Reputation Index is an ongoing research franchise that audits that answer. Each edition takes a public figure, a company, or a category of institutions and examines the reputation AI has already built for them: what surfaces first, what tone it carries, what it omits, what it gets wrong, and how stable it is across engines. The Index does not judge whether a subject is good or bad. It measures what kind of reputation AI preserves for them — and that distinction governs every edition.
Reputation is now downstream of retrieval.
An AI engine does not hold an opinion. It synthesizes an answer from the sources it can retrieve — and then that answer hardens. Understanding the mechanism is the whole study.
Retrieval
The engine pulls from the sources available to it — news, reference pages, owned material, archives.
Synthesis
It compresses those sources into a single confident answer, led by whatever is most abundant and most authoritative.
Reinforcement
That answer becomes the dominant narrative — repeated, cited, and increasingly difficult to shift.
Durability
The narrative stabilizes. A crisis freezes; a rebrand fails to reset it. The answer outlives the moment that created it.
The consequence is simple and counterintuitive: an AI-held reputation is not built from fame or popularity. It is built from the coherence, control, and abundance of a subject's source base. Change the sources the engine retrieves, and you change the answer. Nothing else does.
Five engines. Multiple passes. Recurring findings only.
Every edition follows the same audit. Reputation is modeled across five AI engines — ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews — using more than 40 reputation-intent prompts per subject, spanning six intent categories: identity, trust, track record, controversy, comparison, and decision intent.
Each subject is audited across multiple passes. Only findings that recur across runs are reported — never a single output. Every read is triangulated against current independent sources, deliberately including critical and contrarian coverage as well as favorable coverage, so no finding is favorable by omission. Every factual and financial claim is verified independently. Findings are directional estimates, reported with stated confidence — a reputation audit, not a precision instrument.
One composite score. Five equal dimensions.
Each subject receives a Reputation Index Score from 0 to 100. It is built from five dimensions, each scored 0–20 and equally weighted — no dimension is privileged. The score makes the audit comparable across subjects and trendable on re-audit.
Accuracy
Are AI's claims factually correct and current?
Sentiment
What is the valence of the framing surfaced first?
Completeness
Is what's material and true actually surfaced?
Consistency
Do the five engines agree with one another?
Control
Does the source base trace to the subject?
Fame did not produce the strongest reputations.
A representative score from each edition, ranked. The pattern holds across the franchise: coherence and a controlled source base outscore reach and fame every time.
Edition 01 is live now.
The Reputation Index launches with its first study. Future editions are staged for the weekly release cadence and will be added here as they go live.
A diagnostic — and a starting point.
Executive reputation
Founders and CEOs use a Reputation Index audit to see the narrative AI has built for them — before a board, an investor, or a reporter does.
Corporate & brand
Companies audit how AI frames them in buyer research, and where a frozen crisis is still costing them.
Crisis & pre-crisis
The Index maps the retrieval base before an event sets the narrative — so the infrastructure is built before the crisis, not during it.
M&A & diligence
Acquirers and investors use AI-held reputation as a read on a target's hidden narrative risk.
The language of the Index.
What the Index is — and is not.
Directional, not precise. The Reputation Index reports directional estimates synthesized across engines and passes. It is a reputation audit, not a precision measurement instrument, and scores should be read as informed estimates with stated confidence.
Outputs vary. AI-generated answers vary by user, timing, and phrasing. Findings reflect dominant patterns observed across repeated passes — not any single response, and not a guarantee of what any one user will see.
It scores the answer, not the person. The Index measures the reputation AI preserves for a subject. It is not a judgment of the subject's character, conduct, or quality, and is not investment, legal, or reputational advice.
Common questions.
5W Shapes the Answer in the Box.
When someone asks an AI engine who you are, a synthesized answer comes back. The 5W Reputation Index audits that answer — and the work that follows rebuilds the retrieval base it is made of.