The lead-gen middlemen sitting between buyers and you.

When a buyer asks ChatGPT or Google AI for a Hawaii wealth manager, the AI cites your firm’s own website 47% of the time and third-party content the other half. The largest single block of that third-party content is two lead-gen platforms most firms do not actively manage.

The cross-category picture

For wealth, the read is closer to half-and-half. About half the AI-citation surface for your category is on your own site (clean it up, the leverage is real). The other half is off-site (lead-gen platforms, named-founder bylines, published rankings, Wikipedia presence for principals). The full Hawaii wealth-management teardown covers the 42-firm cohort, the lead-gen middleman pattern, and Claude’s structurally thin training-data presence for the category.

The Microsoft Copilot gap, specific to wealth

Microsoft Copilot cites zero Hawaii wealth-firm websites across 787 mentions in our data. The gap is cohort-wide. In our data, Copilot answers track Bing organic results, so the firms ranking in Bing organic for the buyer questions are the ones Copilot currently cites, and absent firms are absent.

The observed condition is Bing organic visibility for the named queries. The current Bing top 5 for these questions is dominated by national publications (SmartAsset, Investopedia, Forbes) and lead-gen platforms, not Hawaii-specific firm content. We report which firms Copilot cites today and where Bing organic visibility sits for the named queries, run over run.

The Claude training-data finding (worth flagging)

Claude answers from its training data rather than searching the live web. Across 646 Claude mentions in our wealth measurement, only 14% went to firm-owned websites. The other 86% pulled from third-party content even though Claude could have named the firms. For comparison: Claude cites Hawaii law firms 51% firm-owned and Honolulu dental practices 34%. Hawaii wealth firms have a structurally thinner Claude training-data presence than the other relationship-business cohort we measured.

What that means: a buyer asking Claude “should I work with [your firm]” is unlikely to get a substantive answer drawing from your firm’s brand presence. Closing that takes longer than any other move in the engagement (Claude’s training data refreshes on a schedule we do not control), but the signals that move it are nameable: Wikipedia presence for named principals, sustained press coverage, named-author bylines in publications AI ingests.

What an engagement looks like for a Hawaii wealth firm

The standard engagement shape, scoped to a wealth firm:

• Scoping call (30 min). Lock the 18 buyer questions a real Hawaii wealth-management buyer would ask AI (an example set is published at /methodology/#query-set). Lock the cohort. Confirm the questions reflect your specific buyer (high-net-worth, business owner, physician, retiree, inherited wealth, etc.).
• Three-week kickoff. Daily measurement across the seven AI tools: five citation-grade engines that search the live web (Perplexity, ChatGPT search, Gemini grounded, Microsoft Copilot, Google AI Overviews) and two model-knowledge engines that answer from training data (Claude, Gemma).
• Research memo + clear list of fixes. Named competitors, observed gaps per question, prioritized punch list scoped to two parallel workstreams (on-site cleanup + off-site authority work).
• Monthly delta memo after the kickoff. What moved, what did not, drift alerts when a competitor moves on your category.

For wealth firms specifically, the punch list typically covers: (1) SmartAsset and getwarmer.com profile presence, (2) the Bing organic visibility that Copilot citations track in our data, (3) Claude training-data signals, (4) the named-comparison query positioning when your top 2-3 competitors are mentioned by name. We name the specific gaps in the paid engagement. The free check does not.

Honest scope

• 42-firm cohort grew through three rounds of cohort-coverage expansion. Cohort includes both Hawaii-based independent firms and the Hawaii practices of national firms (Morgan Stanley, Merrill, UBS, Edward Jones, Raymond James, etc.). Bank wealth subsidiaries (FHB wealth, BOH wealth) are double-tracked with the banking cohort.
• Hawaii-specific. National wealth-mgmt brands without Hawaii presence are not in the cohort.
• Measurement-only. We do not write content, edit pages, or change your site. Your team or your agency does.
• The initial measurement (1 run, 15 firms) reported 27% own-site / 70% third-party and suggested wealth was the "relationship-end" category. Three rounds of cohort expansion showed that was an artifact of incomplete cohort coverage. The corrected number is 47% own-site, closer to dental than to law. We document the correction publicly on the cross-category teardown and on the wealth-management teardown itself. Self-correction is part of the discipline.

NeverRanked home · Full wealth-mgmt teardown · Cross-category gradient · How we measure