Ask five AI engines which RIA to recommend, and ChatGPT names different firms than the other four.

When a prospective HNW client asks AI to recommend a registered investment adviser, the engines do not agree. In our May 2026 pilot we ran six HNW buyer-intent prompts — family-office management at $50M+, multi-generational wealth, exiting founders, pension and institutional advisory, private-equity coordination, ultra-HNW tax and succession — across all five major engines. Result: 86 RIAs named, mean inter-engine overlap 0.32 (lower than commercial insurance at 0.38), and not a single buyer question produced the same firm at #1 across all five engines. ChatGPT is the systematic outlier: it tilts toward trust companies and private banks (Bessemer Trust, Rockefeller Capital, J.P. Morgan Private Bank) while the other four engines lean toward independent RIAs (Creative Planning, Mariner Wealth Advisors, Cresset Asset Management). An HNW prospect arrives at the discovery call with a different shortlist depending on which engine they used. This page describes observed AI surfacing, not investment advice or adviser-selection advice.

The shift we keep hearing

Inbound from web is softer than it used to be. The Barron’s ranking still drives reputation but no longer drives the pre-meeting research. Prospective HNW clients increasingly start advisor discovery the same way they start every other research task: by asking AI. They still close on referrals and in-person fit; but the consideration set that gets to your first call is being filtered by something opaque.

And the filter is not consistent. The Diagnostic shows that the same HNW buyer question produces materially different shortlists depending on whether the prospect uses ChatGPT, Claude, Gemini, Perplexity, or Grok. ChatGPT is the most striking outlier — it systematically surfaces a different set of firms than the other four engines. Family-office introductions still happen the old way. The shortlist of firms a prospect considers in the first place does not.

What we found when we asked AI to recommend an RIA

In late May 2026 we ran a bounded pilot capture — six HNW buyer-intent prompts, five AI engines, three runs each (90 captures, scored line by line). The prompts spanned family-office management at $50M+ in assets, multi-generational wealth and complex trust structures, regional advisors for venture-backed entrepreneurs exiting their companies, institutional advisory and pension-fund management, private-equity and alternative-investment coordination, and ultra-HNW tax-optimization and succession planning. The cross-engine divergence was larger than any other sector we have measured.

Creative Planning surfaced strongly on Claude (67% of relevant prompts), Gemini (50%), Grok (50%), and Perplexity (33%) — but ChatGPT surfaced it 0% of the time. Mariner Wealth Advisors: 67% on Gemini, 17–33% on Claude/Perplexity/Grok, 0% on ChatGPT. Cresset Asset Management: 50% on Claude and Grok, 17–33% on Gemini and Perplexity, 0% on ChatGPT. The pattern recurs for nine of the top fourteen consensus firms.

ChatGPT IS naming RIAs — it surfaced a comparable volume of firms per prompt to the other engines — but it is naming different firms. ChatGPT’s shortlists tilt toward Aspiriant, Bessemer Trust, Pathstone, and J.P. Morgan Private Bank; the other four engines tilt toward Creative Planning, Mariner, Cresset, and Mercer Advisors. Two distinct shortlist universes. Same buyer question.

Five engines · five shortlists Prompt: RIA for multi-generational wealth & complex trust structures (WMR-02)

Five different shortlists. Two distinct lean directions. The full per-prompt exhibit — including how ChatGPT systematically diverges from the other four engines — is in the Answerability Index entry for Wealth Management & RIAs.

If your diagnostic shows a similar gap, here is what we usually recommend building

For SEC-registered investment advisers facing a category where the engines name a different firm each, the typical post-Diagnostic build queue is six to eight specific assets:

• Three to five answer-shaped pages for the HNW scenarios where AI names a competitor instead of you — family-office at $50M+, multi-generational wealth with complex trust structures, exiting founders, institutional advisory, private-equity coordination, ultra-HNW tax/succession.
• Entity-graph cleanup — consistent firm identity across SEC Form ADV, Barron’s Top 100, FA Magazine RIA Ranking, RIA Edge 100, your site, Wikidata, LinkedIn, and regional business-journal listings. The engines stop fragmenting your firm across slight name variants.
• Citation assets — Barron’s / FA Magazine / Forbes ranking submissions, family-office trade-press placements (Family Wealth Report, Citywire RIA), and regional business-journal positioning. The corroboration set the engines actually weight.
• Competitor-comparison pages — your firm vs Creative Planning / Mercer Advisors / Mariner Wealth / Cresset for the wealth tiers you actually serve, written compatible with SEC Marketing Rule.
• Schema package — Organization, FinancialService, FAQPage, BreadcrumbList, named-advisor profiles with CFP / CFA / JD credentials, Form ADV linked. Ship-ready JSON-LD compatible with fiduciary-disclosure obligations.
• FAQ expansion — passage-citable answer blocks (134–167 words each) on the HNW questions your prospects actually ask, compliance-reviewed before publish.

Build it in-house from the Diagnostic, or have us build it — a Sprint engagement ($15,000 then $950/mo) is done-for-you remediation over four weeks, with output staged for your compliance counsel’s SEC Marketing Rule review before any external publish. Either way, the Diagnostic is the prerequisite.

Order the Diagnostic to see your specific build queue — $2,500 →

Why RIA recommendations fragment — and what to do about it

Wealth-management authority is fragmented in the underlying sources the engines weight. Barron’s Top 100 surfaces one canonical set of firms. FA Magazine’s RIA Ranking surfaces a meaningfully different set. SEC Form ADV is the underlying disclosure document but it is dry and structured, not narrative; engines that can parse it well surface different firms than engines that lean on awards lists or regional press. Family-office trade publications, fiduciary-fit blogs, and HNW-specific media each contribute their own corroboration set.

The result: each AI engine weights these sources differently and produces a different shortlist for the same buyer question. ChatGPT in particular seems to favor brand-recognized trust companies and private banks; Perplexity and Grok favor large independent RIAs. The Diagnostic reads where your firm sits in each engine’s weighting and locates the gap in three places.

Pillar	The question we answer	What we look for in an RIA site
Content	Can the engines extract a differentiated value proposition from your site?	Server-side-rendered firm pages with explicit specialization signals (family-office vs. mass affluent, alternatives vs. traditional, geographic focus); buyer-question pages (“wealth advisor for second-generation business owners,” not just “Our Services”); named advisors with credentials (CFP, CFA, JD); and Form ADV linked from your site.
Retrieval	Do you appear when HNW prospects actually run the questions?	Observed surfacing across the buyer-intent prompts that match the wealth tier and complexity you serve, on the engines your prospects use. Not your branded queries — the unbranded category queries where the named firms show up.
Trust	Are you corroborated in the sources these engines weight?	Presence in Barron’s Top RIA rankings, FA Magazine RIA Ranking, RIA Edge 100, regional business-journal recognition; entries in family-office trade publications; consistent SEC Form ADV registration that the engines can confirm; named-but-not-cited gaps where an engine knows your firm exists but doesn’t treat you as a source on its recommended shortlists.

HNW prospects ask AI which RIA to consider. The engines answer with five different shortlists. The first question is whether yours appears in any of them. The second is which engine is the outlier — and what to do about it.

What you receive

Built for

SEC-registered investment advisers, multi-family offices, wealth-management firms, and trust companies (typically $1B–$50B AUM) — firms already spending on partner relationships, referral programs, conference sponsorships, or content marketing, and who need to know whether the AI engines are putting them in the HNW prospect’s consideration set or naming a competitor instead.

Order the Diagnostic

Find out whether AI is naming your firm to HNW prospects — and which engine is the outlier.

Delivered in 10–14 business days. Five engines. Thirty to fifty HNW buyer-intent prompts measured for your specific firm, on the wealth tiers and complexity you serve. Confidential under MNDA. The first month is the full Diagnostic; thereafter the Visibility Intelligence subscription keeps the picture current as the engines move.

Order your Diagnostic — $2,500 See the underlying capture →

$2,500 first month (the Diagnostic) · $950/mo thereafter (Visibility Intelligence: monthly re-runs, deltas, build-queue updates)

A note on scope. The wealth-management industry is regulated under the Investment Advisers Act of 1940, the SEC Marketing Rule (Rule 206(4)-1), state-level adviser-registration regimes, and (for broker-dealer-affiliated advisers) FINRA 2210. Nothing on this page is investment advice, advisor-selection advice, a solicitation, an endorsement, or a testimonial. The Diagnostic measures observed AI surfacing — how AI systems represent firms — not adviser quality, fiduciary fit for any specific HNW client, investment performance, or recommendation quality. Surfacing patterns reflect the engines’ current behavior on the captured date and will change as the engines and the underlying public sources change. Capture: 2026-05-30. Six HNW buyer-intent prompts, five engines, three runs per engine (90 captures total). Firm names canonicalized; firms named by only a single engine (the long tail of 80+ RIAs) summarized in the Index entry, not named individually here.