Will ChatGPT recommend your firm?
Your next client may ask an AI for a shortlist before they ever ask around. So it's a fair question — and unlike most things in marketing, you can get a real answer to it in about twenty minutes.
Maybe — but probably not for the reasons you'd hope. AI assistants now hand buyers a shortlist of firms before anyone makes a call, and whether yours is on it depends less on your reputation than on whether the engines can resolve and corroborate you. The good news: you can check exactly where you stand today, yourself, for free.
Your buyers are already asking
The first move used to be a referral or a Google search. Increasingly it's a question to a chatbot: "who are the best [practice] firms for a situation like mine?" The model answers with two or three names and a confident sentence on each — and that shortlist is now the first filter, applied before anyone visits a website or picks up a phone.
This is already normal among the buyers worth having. One 2025 study found a quarter of affluent households shopping for a financial advisor were starting that search inside ChatGPT or Gemini.1 If your firm isn't named, you don't get a rejection — you simply never enter the conversation. The most expensive losses in this channel are the ones you never see.
Check it yourself in 20 minutes
You don't need a tool to get a first read. You need a fresh browser and twenty unhurried minutes. Here's the procedure we'd run:
- Open a clean session. Use an incognito/private window so your own history doesn't bias the answer. Do this in each engine — at minimum ChatGPT and Perplexity; ideally add Claude and Gemini.
- Ask the shortlisting questions, not your name. Use the queries a buyer would actually type: "best [practice] firms for [situation] in [region]," "who should I hire to [job-to-be-done]," "[Competitor] vs alternatives." Aim for 6–8 prompts.
- Record what comes back. For each answer, note the firms named, in what order, and — this is the part most people skip — which sources the engine cited. A simple sheet with columns for Prompt, Firms named, Sources cited does the job.
- Compare across engines. Don't average them. Look at where they disagree — you may be named in Perplexity and invisible in ChatGPT, which are two different problems with two different fixes.
- Read the sources, not just the names. If the engines are leaning on a directory, a "best firms" listicle, or a competitor's page to build the list, that tells you where the citation is actually coming from — and it's rarely your homepage.
Three patterns show up almost every time: the lists barely overlap between engines; the firms named tend to be the ones easiest to corroborate (a directory listed them, a ranking included them) rather than the ones with the deepest track record; and the behavior is unstable — ask again next week and it may shift. A single screenshot is a snapshot, not a verdict.
Why the engines pick the firms they pick
Here's the uncomfortable mechanic. An AI deciding which firms to name isn't ranking websites; it's asking around, and it trusts what it can corroborate from independent sources. That's why unlinked brand mentions — other people writing about you — correlate with AI citations far more strongly than backlinks or domain authority do.2 Your own site is one source, and the one most obviously incentivized to flatter you.
For most firms the binding constraint isn't talent or content — it's whether the engine can resolve you to a real, corroborated entity at all. We get into the full mechanic in why AI recommends some companies and ignores others.
Why professional-services firms are uniquely exposed
Most categories that worry about AI visibility sell a product with a website built to be read by machines. Professional-services firms are the opposite case: high authority, low machine-readability. The reputation is real — but it lives in private referrals, closed-door deals under NDA, and long-form thinking published as PDFs and narrative essays rather than as clean, extractable answers. The signals the engines weight most — a resolvable entity, third-party corroboration, dated and structured content — are exactly the ones a partner-led firm has never had a reason to build.
The result is a widening gap between a firm's standing in the market and its standing in the answer layer. You're not losing because the work is weak. You're losing because the work is illegible to the system now doing the first round of vetting. That's the case we make in depth on the professional-services diagnostic page.
You're not losing because the work is weak. You're losing because it's illegible to the thing now doing the first round of vetting.
What to do once you've looked
If your twenty-minute check turns up a problem, resist the urge to "publish more content" at it — that fixes the wrong gate if the issue is corroboration or crawlability. The durable moves are making your expertise legible to machines and giving the rest of the internet enough true, consistent signal to vouch for you. Slow, but it's the work that holds.
If you'd rather not run the audit by hand — or you want it across all five engines, scored, with the competitor citation landscape and a sequenced fix list — that's exactly what we do.
Get the firm-grade version of that 20-minute check
The Diagnostic runs your real buyer questions across all five engines, captures every answer, scores each cited URL, and maps which competitors get named instead of you — with a sequenced roadmap to close the gap. One-time, $3,000, confidential under MNDA.
References
- Affluent households beginning advisor search inside AI tools (2025 study, cited in financial-services coverage). See WealthManagement.com, "The AI Search Reckoning in Financial Services" (2026).
- Ahrefs study of 75,000 brands: unlinked brand mentions correlate with AI citations at 0.664 vs 0.218 for backlinks; domain authority predicts under 4% of citations. See "Backlinks vs brand mentions: the 2026 AI visibility playbook" and AuthorityTech (2026).