Publish your licensing
Name your banking partner, licence numbers and deposit protection in plain HTML, not a PDF. This one page turns an unverifiable brand into a nameable one.
Optimise Content →Solutions · By industry
AI will not recommend what it cannot verify.
Money questions get a higher bar. AI leans on regulators and finance publishers, hedges on anything it cannot check, and skips products whose fees, licence or FDIC status are unclear. Prefer shows which prompts skip you, which sources decide them, and what to publish to get named.
“best business bank account for startups”Category prompt
Per engine
ChatGPTAbsent
PerplexityAbsent
ClaudeNamed
ChatGPTAbsentPerplexityAbsentAI OverviewsAbsent
ClaudeNamedNamed in the answer
Mercury
Brex
NovoWho decides it
nerdwallet.comA regulated-finance publisher outranks every vendor page here
YMYL category prompts return few names and heavy hedging. Engines name only providers whose licence and deposit protection they can verify.
The trust math
Money answers are held to a higher bar.
How AI engines answer financial questions, across 310 tracked fintech domains.
63%of money prompts carry a hedge or disclaimer
Engines qualify financial answers far more than software ones.
4.1×regulators and finance publishers out-cite fintech brands
FDIC, FCA, NerdWallet and Bankrate carry the answer.
58%of fintechs cannot be confirmed as licensed by an engine
Unverifiable licensing reads as risk, and risk reads as absence.
3-8 wksfrom shipped fix to measurable movement
Slower than SaaS, because trust sources re-crawl less often.
Prefer benchmark · 310 fintech domains · 6 engines · Q2 2026.
The trust bar · 01
Caution is the default when money is involved.
There is money attached to the answer, so AI plays safe. Anything it cannot confirm, a licence, a fee, a deposit guarantee, it hedges or leaves out. Leaving out is the common case.
“best business bank account for startups ”
ChatGPT, answered this morning.
For most early-stage companies, Mercury is the usual recommendation: no monthly fees, FDIC insurance through partner banks up to $5M, and built-in treasury. Brex suits venture-backed startups that want cards and spend management together. Novo is a simpler option for small teams. Do check current terms directly, as fees and partner banks change.
Sourcesnerdwallet.commercury.com
fdic.gov
3providers named1hedge attached0times Northpay appears
Who decides · 02
Regulators and publishers write the answer.
Fintech citations land where you have no control: government registers, finance media and review platforms. Your own pages come last, and only if they carry real numbers.
Where citations come from
One tracked category, 210 prompts, last 30 days.
- Finance publishers
- 29%
- Regulators & registers
- 22%
- Review & complaints
- 17%
- Competitor-owned
- 14%
- Community
- 11%
- Your owned pages
- 7%
DomainTypeCitation share
1nerdwallet.comNot listed in 3 relevant roundupsPublisher13.8%
2fdic.govPartner bank listed, brand not associatedRegulator11.2%
3
bankrate.comPublisher9.4%
4
trustpilot.com3.9 stars, one recurring themeReviews8.1%
5mercury.comTheir comparison pages get quotedCompetitor7.6%
6
consumerfinance.govRegulator6.3%
7
reddit.comr/smallbusiness threadsCommunity5.9%
8
northpay.coFee page not machine-readableYour domain3.1%
Domains cited across 210 tracked money prompts, last 30 days. Regulator citations cannot be earned with content, only with correct registration data.
The prompt map · 03
Six families decide a financial product.
Two of them do not exist in software: licensing and safety. Both are answered from sources you do not own, and both end deals before a demo.
Northpay, per family per engine
Status of the best prompt in each family.
CitedNamedAbsent
FamilyExample promptChatGPTPerplexityAIOClaudeWhat wins it
Categorybest business bank account for startupsPublisher roundups + a citable product page
Feescheapest way to send money to IndiaA public rate table with structured data
Licensingis Northpay FDIC insuredCorrect register data + a plain disclosure page
ComparisonMercury vs Brex vs NorthpayYour own honest comparison pages
CompliancePCI DSS platform with SOC 2A public trust center, not a sales PDF
Safetyis Northpay safe to useComplaint themes resolved at the source
Live from the Northpay demo workspace. Licensing and safety prompts run daily, because a wrong answer there compounds.
The playbook · 04
Six plays win financial categories.
Each play maps to the Prefer tool that runs it, so this is a queue, not a PDF.
Put your fees in the open
Rate tables, transfer fees and FX margins as data AI can read. Fee prompts are decided by numbers, and a number behind a login does not exist.
Create Content →Get into the roundups
NerdWallet, Bankrate and Investopedia outweigh any brand page in finance. Prefer finds the roundups you are missing from and briefs the pitch.
Action Center →Run a public trust center
SOC 2, PCI DSS and GDPR proof on a page AI can crawl. Compliance prompts are the enterprise gate, and an incumbent answers them today.
Optimise Content →Fix complaints at the source
One unresolved Trustpilot or CFPB pattern becomes the warning attached to your name in every safety answer. Prefer traces it and tracks the repair.
Answer Engine Insights →Watch the incumbents
Every rival named, per prompt, per engine, per week. In regulated categories incumbents hold answers for years, so any movement matters.
Competition →Consumer vs B2B · 05
Consumer and B2B lose different answers.
Consumer fintech is judged on safety and fees. B2B fintech on compliance and integration. Prefer ships a prompt set for each.
Consumer fintech
One person, deciding fast, asking AI whether you are safe. Where you are exposed:
Safety and licensing“is X safe, is X FDIC insured”
Fees and rates“cheapest transfer, hidden fees”
Alternatives“X alternatives, better than X”
Complaints“X reviews, X problems, X scam”
Exposure: one complaint theme quoted as the caveat on your name.
B2B fintech
A committee with legal in the room, and AI briefing each member. Where you are exposed:
Compliance“PCI DSS, SOC 2, ISO 27001”
Platform fit“works with NetSuite, Shopify, SAP”
Payout and settlement“settlement times, payout rails”
Vendor diligence“X vs Y for enterprise, X uptime”
Exposure: the security reviewer's assistant cannot find your attestations.
Proof · 06
From unverifiable to recommended.
“We assumed the problem was awareness. It was that no engine could confirm we were a real, licensed product. We published one disclosure page and a rate table, and within a month assistants started naming us instead of hedging.”
0 → 31%mention share on category prompts58% → 4%answers that hedge on licensing8 weeksto first measurable move+34%assisted signups from AI referrals
Questions
Asked plainly.
How do fintech companies get recommended by AI assistants?
By being verifiable. Engines apply a higher bar to financial answers, so they name providers whose licensing, deposit protection and fees they can confirm from a regulator, a finance publisher or a machine-readable page on your own site. Unverifiable claims are hedged or omitted.
Why is AI visibility harder in financial services?
Because money questions fall under what search guidelines call YMYL, your money or your life. Assistants hedge more, cite regulators and established publishers more, and rely on vendor marketing pages far less than they do in software categories.
Which prompts should a fintech track?
Six families: category (best account or card for X), fees and rates, licensing and insurance, comparison, compliance, and safety or complaints. Licensing and safety are unique to regulated categories and can end a deal before any demo.
Which sources do AI engines cite for fintech answers?
Finance publishers first, then regulators and registers, then review and complaint platforms. In Prefer's Q2 2026 benchmark of 310 fintech domains, publishers took 29% of citations, regulators 22%, reviews 17%, and brands' own pages 7%.
Can AI-generated content be compliant in financial services?
Only with grounding and review. Prefer drafts from your own approved documents and live sources, attaches a citation to every claim, and refuses to write a number it cannot source. Nothing publishes without your compliance sign-off.
How do we stop AI saying we are not insured?
Publish the disclosure in crawlable HTML: name your partner bank, licence numbers and coverage limits on a page with structured data, and make sure your register entries match. Engines check the register first and your site second.
How long until AI answers change in finance?
Three to eight weeks, slower than software, because regulator pages and finance publishers are re-crawled less frequently. Your own pages move fastest, which is why disclosure and fee pages are the first fixes.
Does this work for embedded finance and B2B payments?
Yes, with a different prompt set. B2B fintech is judged on compliance, platform fit and settlement rather than consumer safety, so the tracked families and the pages that win them differ.
Get your free AI visibility report
in about 10 minutes.
See how answer engines describe your brand today, and where the openings are to outpace the competition.