Your AI Reputation Score: What Founders Miss

On July 27, 5W Public Relations published Edition 01 of The 5W Reputation Index. It scored the founders of the leading AI labs by the reputation the models themselves hold about them. Demis Hassabis came in at 86. Dario Amodei at 82. Sam Altman at 64.

Twenty-two points between the most-covered founder in the industry and the highest-scored one.

The finding that ran through the whole report: fame did not produce the strongest reputation. Credentials and a controlled source base did. Every B2B founder has a version of this same score sitting inside every major AI engine right now. Almost none of them have ever checked it.

This is the founder AI reputation score, and it is the metric that decides what a buyer reads about you five seconds before they click into your calendar link.

What Your AI Reputation Score Actually Is

Your founder AI reputation score is the composite narrative that ChatGPT, Perplexity, Gemini, Claude, and Grok construct about you when a buyer types your name or your category into a prompt. It sits across five dimensions: factual accuracy, sentiment, completeness, consistency across engines, and the quality of the sources feeding each answer.

It is not your Google rank. It is not your LinkedIn follower count. Those are visibility metrics for a world where buyers still click through search results. AI engines skip that layer. They read the entity graph, weigh the source base, and hand the buyer a synthesized paragraph. That paragraph is your score.

The 5W Index formalized five variables that map to how the models actually reason: credential density, source authority, topic coherence, coverage recency, and platform consistency. The 22-point gap between Hassabis and Altman was not driven by name recognition. It was driven by how tightly each founder's source base pointed at the same set of credentials and topics. Hassabis reads as a research scientist with DeepMind and specific published work. Altman reads as many things depending on which model you ask, and the models disagree about which of those things matters most.

Disagreement across engines is the single strongest tell that a founder's AI reputation is uncontrolled.

How to Run Your Own AI Reputation Audit in 20 Minutes

The audit is six prompts across five engines. It takes twenty minutes and it will tell you exactly where your score is bleeding.

Run each of the following in ChatGPT, Perplexity, Gemini, Claude, and Grok:

1. "Who is [your full name]?"
2. "What is [your name] known for?"
3. "What has [your name] published or written about?"
4. "Who are the top founders in [your category]?"
5. "Compare [your name] to [a direct competitor]."
6. "What is [your name]'s track record on [your core topic]?"

For each answer, score three things. Is the entity binding correct — is the model talking about you or someone with your name? Are the facts accurate? Which sources does it cite, and do you control them?

If the audit reveals that three engines describe you differently, or that Perplexity cites sources you have never seen, or that you do not appear in the "top founders in [category]" prompt at all, you have a score problem. The good news is that every one of those problems is a source problem, and source problems are fixable. If you want the citation side of the fix laid out in full, the deeper playbook on how to get cited by ChatGPT walks through the tactical execution.

The 5 Signals That Determine Your Score

Five source classes feed almost every AI answer about a founder. Getting these right lifts your score. Ignoring them keeps you invisible or, worse, misrepresented.

1. Wikidata and Wikipedia. The structured entity anchor. If you have a clean Wikidata item with your correct occupation, employer, and notable work, every engine downstream of it inherits that structure. Missing Wikidata is the single most common reason a founder shows up as "unclear" or gets confused with someone who shares their name.

2. Tier-1 earned media. Bylines you wrote and named quotes in outlets the engines rank as authoritative. Ahrefs analyzed 75,000 brands and found that web mentions correlate with AI citation at 0.66, versus 0.22 for backlinks. Mentions predict citation three times more strongly than the SEO metric most founders still optimize for.

3. Schema-marked owned properties. Person schema on your About page and your author pages. This is the free win almost every founder skips. Ten minutes of markup, and every engine that crawls your site has a clean, machine-readable record of who you are.

4. Podcast and video transcripts. The most underdeployed signal on this list. Long-form audio with a transcript is dense, on-topic, quotable content that engines treat as first-person expert commentary. Two well-transcribed podcast appearances often shift a founder's score more than a full year of LinkedIn posts.

5. Verified third-party profiles. Crunchbase, LinkedIn, AngelList, GitHub, and industry-specific directories. These act as cross-references. When five sources agree you are the CTO of a specific company working on a specific problem, engines stop hedging.

If you want the full tactical breakdown on the earned-media side of this list, the GEO playbook for founders covers the execution in depth.

The Anti-Patterns That Destroy Your Score

Three mistakes actively suppress a founder's AI reputation. Most founders are running at least one of them right now.

Inconsistent identity across platforms. Your Twitter bio says "founder." Your LinkedIn says "CEO and cofounder." Your podcast intro says "startup guy." Your press quotes call you "AI entrepreneur." To a human, these read as the same person. To an engine building an entity graph, they read as evidence that the entity is fuzzy. Fuzzy entities get lower confidence scores and fewer citations. Pick one title, one company reference, one topic anchor, and enforce it across every platform.

Relying only on owned properties. The Ahrefs finding is worth repeating: mentions predict AI citation three times more strongly than backlinks. A founder who has published fifty blog posts on their own site and zero bylines on external outlets is legible to Google and invisible to Perplexity. Owned content is necessary. It is not sufficient.

Topic sprawl. Engines cite named authorities on defined subjects, not generalists on twelve subjects. A founder who writes about hiring, product, fundraising, AI ethics, remote work, and personal productivity is telling every model that the entity has no primary domain. The founders who score high are the ones whose source base concentrates 70 percent of their signal on a single, defined topic.

The 30-60-90 Day Fix for Founders

Days 1 through 30. Run the audit. Rewrite every bio to match a single canonical version. Add Person schema to your About page and every author page on your site. Create or update your Wikidata item with correct occupation, employer, and at least three notable works or affiliations. File corrections with OpenAI, Google, and Anthropic for any factual errors the audit surfaced.

Days 31 through 60. Land two bylines in outlets the engines treat as authoritative for your category. Record two long-form podcast appearances, at least 45 minutes each, on your primary topic, and confirm each podcast publishes a full transcript. Update your LinkedIn, Crunchbase, and AngelList profiles to match the canonical bio.

Days 61 through 90. Re-run the six-prompt audit across all five engines. Compare scoring against your baseline. Lock a monthly cadence: one byline or major podcast per month, one quarterly Wikidata refresh, one quarterly schema audit. The founders who compound this consistently for a year end up in the top decile of their category's AI-held reputation.

Why the Window Is Closing

The compounding effect is real, and it is asymmetric. Founders who build entity recognition now become harder to displace as the models retrain. Ahrefs found that pages updated within two months earn 28 percent more AI citations than older content. Recency is a live variable. So is incumbency.

The founders who wait will face a higher barrier. Every model retrain locks in the current signal. A founder who is already the named authority on a topic when the next training cutoff hits will be quoted, cited, and recommended for the entire next generation of that model. A founder who is not will be spending the next twelve months trying to break in against an established citation pattern.

This is the moment where the 22-point gap gets built. Not in a decade. Now.

Run the audit this week. Fix the anti-patterns this month. Lock the cadence for the year. That is the founder AI reputation score playbook, and it is the difference between being the answer and being the also-ran.