Payments, banking, and wealth are all raising budgets
In 2025, US financial-services digital ad spending grew 23% in payments, 20% in banking and lending, and 14% in wealth (eMarketer, 2025). The subverticals holding the deepest first-party data are the ones raising budgets fastest. The category reached $33.8 billion in 2024, behind only retail and CPG, which is where its top-three standing was last measured (eMarketer, 2024).
Finance is also building media of its own. Financial media networks, the bank and payment-company answer to retail media, are forecast at roughly $640 million in 2025 and toward roughly $1.22 billion in 2026 (eMarketer, 2025). Institutions that hold consented transaction data are beginning to run it as advertising infrastructure, on their own properties and under their own governance.
That is the tension defining the vertical. The data that makes finance targeting precise is the data the law protects. Fair-lending law, privacy law, and the ad platforms themselves have taken the most precise targeting tools out of finance marketers’ hands, and the removals are mechanical. The plan that wins in 2026 reaches the right customer without prohibited proxies, on supply it can defend to a compliance reviewer.
Five gates, checked before creative
Five compliance gates govern a US finance campaign before a single placement is named. ECOA and Regulation B bar targeting, or proxy targeting, on a prohibited basis: race, sex, age, national origin, and close stand-ins for them. UDAAP and TILA govern the claim itself, where terms like “free” and “0%” pull in mandatory disclosures. GLBA and FCRA wall off customer financial and credit data from ad targeting outside the law’s narrow path, with opt-out suppression built into the path. Platform Special Ad Categories are the fourth gate, enforced mechanically by Meta and Google. FINRA 2210 and SEC 206(4)-1 add the fifth: principal pre-approval and archiving for regulated creative.
The platform gate is the one buyers feel daily. Since January 2025, US finance ads on Meta have run inside a Special Ad Category: no age, gender, or ZIP narrowing, no lookalike audiences, and a 15-mile minimum location radius (Meta Transparency Center, updated May 2026). Credit, housing, and employment ads on Meta and Google carry the same restrictions (Google personalized-advertising policy). In May 2026, Meta’s Credit category expanded to capture Buy Now, Pay Later ads. What remains permitted is specific: household income, homeownership, in-market intent, broad geography, contextual placement, and consented first-party data.
The rules tightened and shifted in 2026, and the constraint did not go away. State fair-lending law is untouched, and privacy and truth-in-advertising rules apply regardless of platform policy. The safe default is to build the audience without prohibited proxies in the first place, and to check each gate at the buying layer, in order, before creative. Compliance is part of the buy.
| Gate | Rule | What it removes or requires |
|---|---|---|
| Fair lending | ECOA / Regulation B | No targeting, or proxy targeting, on a prohibited basis: race, sex, age, national origin, or close stand-ins |
| Truth in advertising | UDAAP · TILA / Regulation Z | Claims must be substantiated and disclosed; "free," "0%," and rate terms pull in disclosures |
| Privacy | GLBA · FCRA · state law | Customer financial and credit data stay out of targeting outside the law's narrow path; every opt-out honored |
| Platform Special Ad Categories | Meta · Google | ZIP, age, and gender narrowing and lookalikes removed from credit and housing ads; BNPL added to Meta's Credit category May 2026 |
| Approval and recordkeeping | FINRA 2210 · SEC 206(4)-1 | Principal pre-approval on broker-dealer creative; strict disclosure on RIA performance and testimonials; everything archived |
Three compliant paths to precision
Context is the first path. Ads placed against finance content (retirement, investing, mortgages) draw their relevance from the page. After permanent signal loss, it is the most defensible path available, and it doubles as a brand-safety control, since choosing the page also means choosing what the brand sits next to.
Permitted signals rebuild the precision the platforms remove. Household income, homeownership, in-market intent, and broad geography are all allowed inside the Special Ad Category limits, provided each variable is business-justified and documented, and none stands in for a prohibited basis. These are modeled audiences assembled entirely from signals a regulator can review. The documentation is the point, because an audience that cannot be explained cannot be defended.
Consented first-party data is the third path: CRM and clean-room audiences, built to screen against GLBA, FCRA, and state-law opt-outs, matched without exporting customer records, with the clean room returning aggregate answers. One set of guardrails shapes all three paths: no prohibited-basis targeting, no ZIP, age, or gender narrowing on credit or housing, opt-out screening built into the audience build, and principal pre-approval on regulated creative.
Trust is recovering while impersonation sets records
Two trends ran opposite each other in 2025. US consumer confidence in banks recovered to roughly 62%, a post-2008 high (MediaPath Insights, Financial Services 2026, citing Gallup, 2025), and banking has been the most-trusted financial subsector since 2023 (MediaPath Insights, Financial Services 2026, citing Edelman, 2025). At the same time, US imposter-scam losses hit a record $3.5 billion, with bank impersonators driving the most (MediaPath Insights, Financial Services 2026, citing FTC, 2026).
The impersonation pattern predates the record. In 2024, Akamai measured financial services drawing 36% of all counterfeit-domain traffic, the most-impersonated industry online that year (MediaPath Insights, Financial Services 2026, citing Akamai, 2024). A recovering bank brand is also a standing target, and the open programmatic environment can put a real institution’s ad one impression away from a scam built on its name. Where the brand name is the thing being counterfeited, clean supply is brand protection as much as media efficiency.
The supply chain itself is improving. The ANA/TAG TrustNet Benchmark for Q2 2025 found $439 of every $1,000 entering a DSP now reaches consumers, up 7.9 percentage points, with private-marketplace share climbing from 64.5% to 87.8% and made-for-advertising spend down to a 0.8% median (MediaPath Research, The State of Digital Media 2026, citing ANA/TAG TrustNet Benchmark, Q2 2025). For finance, the reading is direct: curated, verified supply keeps the brand away from scam adjacency and puts more of each dollar in front of a real person.
Where the 2026 finance customer already is
The channel mix rewards under-investment. US CTV ad spend reaches $37.95 billion in 2026, up 14.5%, and finance still leans more linear than retail does, which leaves CTV room to grow (eMarketer, 2026). Podcasts reach 58% of Americans 12 and older monthly, 167 million people, and money and business shows over-index on $100K+ households (Edison Research, 2026). Search captures the rate-shopping moment, and social reaches the youngest cohorts while the Special Ad Category limits how narrowly it can aim.
The consumer behind the plan is soft-switching. In 2024, McKinsey put the average US consumer at about three banking relationships (McKinsey, 2024), so the practical job is to be chosen as the household’s next one. First-time buyers fell to 21% of all buyers, a record low, with a median age of 40 (MediaPath Insights, Financial Services 2026, citing NAR, 2025). 38% of Gen Z have used Buy Now, Pay Later, the highest-adoption cohort (MediaPath Insights, Financial Services 2026, citing Bankrate, 2025). And in 2024, FINRA’s NFCS put the new-investor entry rate at 8%, down from 21% on the same measure (FINRA Foundation NFCS, 2024), a contracting funnel that raises the value of every relationship won.
The advisor is a separate lane with its own physics. Advisors research inside a tight set, 6.4 websites per month on average (MediaPath Insights, Financial Services 2026, citing Escalent, 2026), and endemic finance media concentrates the audience: CNBC alone reaches 45.7 million people monthly across platforms (MediaPath Insights, Financial Services 2026, citing comScore, December 2025). In B2B finance, the environment carries more of the targeting than the ID does, so the lane runs on endemic media and the professional graph. Consumer and advisor campaigns plan together and run governed apart, with no prohibited data bridge between them.
Measure to a funded outcome
Finance measurement carries a confidence gap. 85% of marketers say they are confident in their ROI measurement, while only 32% measure holistically across channels (MediaPath Insights, Financial Services 2026, citing Nielsen, 2025). The gap matters more in finance than in most verticals, because the outcomes that count sit weeks downstream of the impression, past the application and past the approval, at the funded account.
The entry condition is verified media metrics: human traffic, viewability, and completion. Applications and approval rates are the mid-funnel an account passes through. The outcome layer is where a finance plan is judged: cost-per-funded-account and incremental net-new-money for banking, net-new-assets stripped of market appreciation for wealth, and cost-per-qualified-advisor with AUM influenced for the advisor lane. Real proof at that level is incremental, established by a holdout inside a GLBA-safe clean room, not by a match-back report.
Four moves follow. Build the compliant audience first, classifying every credit and housing campaign into the right Special Ad Category and rebuilding precision from permitted signals and consented first-party data. Move budget to where the customer already is, since CTV is under-used by finance, audio over-indexes on the affluent, and endemic media concentrates advisors. Protect the brand on clean, verified supply paths. Then measure to a funded outcome, which is what turns a media report into a number a CFO recognizes.
The takeaways
Finance grows fastest where first-party data is richest: payments +23%, banking and lending +20%, and wealth +14% in 2025, while financial media networks are forecast from roughly $640 million toward roughly $1.22 billion in 2026 (eMarketer, 2025).
Five compliance gates govern a US finance campaign before a placement is bought: ECOA/Reg B, UDAAP/TILA, GLBA/FCRA, platform Special Ad Categories, and FINRA 2210/SEC 206(4)-1. Meta's Credit category captured Buy Now, Pay Later in May 2026.
Precision gets rebuilt on three permitted paths: contextual placement, documented permitted signals such as household income and homeownership, and consented first-party data matched inside clean rooms.
Consumer confidence in banks recovered to roughly 62% while imposter-scam losses hit a record $3.5 billion, and $439 of every $1,000 entering a DSP now reaches consumers, so clean, verified supply is brand protection in finance (MediaPath Insights, Financial Services 2026, citing Gallup 2025 and FTC 2026; MediaPath Research, The State of Digital Media 2026, citing ANA/TAG TrustNet Benchmark, Q2 2025).
Only 32% of marketers measure holistically against 85% who say they are confident, so the finance plan is judged on funded accounts, net-new-money, and AUM influenced, proven by a holdout inside a GLBA-safe clean room (MediaPath Insights, Financial Services 2026, citing Nielsen, 2025).
Common questions
How is financial services advertising regulated online?
Five compliance gates govern a US finance campaign before a placement is named: ECOA and Regulation B on fair lending, UDAAP and TILA on the claim itself, GLBA and FCRA on customer and credit data, platform Special Ad Categories enforced by Meta and Google (Meta Transparency Center, updated May 2026), and FINRA 2210 with SEC 206(4)-1 on approval and recordkeeping. Each is checked at the buying layer, before creative.
What is a Special Ad Category on Meta?
It is a platform restriction that strips targeting from regulated ads. Since January 2025, US finance ads on Meta run inside one: no age, gender, or ZIP narrowing, no lookalike audiences, and a 15-mile minimum location radius (Meta Transparency Center, updated May 2026). Credit, housing, and employment ads on Meta and Google carry the same restrictions, and Meta's Credit category captured Buy Now, Pay Later in May 2026.
How do finance advertisers target without prohibited data?
Three permitted paths rebuild precision. Contextual placement draws relevance from the finance page itself. Permitted signals (household income, homeownership, in-market intent, broad geography) are allowed inside Special Ad Category limits (Meta Transparency Center, updated May 2026) when each variable is business-justified and documented. Consented first-party data runs through clean rooms that return aggregate answers without exporting customer records, screened against GLBA, FCRA, and state-law opt-outs.
What is cost-per-funded-account in finance marketing?
It is the consumer North Star of a finance media plan: the cost of an account that is opened and funded, weeks downstream of the impression. It sits above verified media metrics and above applications and approvals on the measurement ladder. Proving it takes an incremental holdout inside a GLBA-safe clean room. Only 32% of marketers measure holistically across channels today (MediaPath Insights, Financial Services 2026, citing Nielsen, 2025).
Sources
eMarketer, US financial services and industry digital ad spending, 2024-2026
eMarketer, Financial Media Networks forecast, 2025
eMarketer, US CTV ad spend outlook, 2026
Edison Research, The Infinite Dial 2026
MediaPath Research, The State of Digital Media 2026 (citing ANA/TAG TrustNet Benchmark, Q2 2025)
MediaPath Insights, Financial Services 2026 (citing Gallup, Trust in Banks, 2025)
MediaPath Insights, Financial Services 2026 (citing Edelman Trust Barometer, 2025, financial services)
MediaPath Insights, Financial Services 2026 (citing FTC imposter-scam data, 2026)
MediaPath Insights, Financial Services 2026 (citing Nielsen Annual Marketing Report, 2025)
MediaPath Insights, Financial Services 2026 (citing comScore, December 2025)
MediaPath Insights, Financial Services 2026 (citing Escalent / Cogent advisor media study, 2026)
MediaPath Insights, Financial Services 2026 (citing NAR Profile of Home Buyers, 2025)
MediaPath Insights, Financial Services 2026 (citing Bankrate BNPL survey, 2025)
Meta Transparency Center (updated May 2026); Google personalized-advertising HEC policy
ECOA / Regulation B · UDAAP · TILA / Regulation Z · GLBA · FCRA · state law
FINRA Rule 2210 · SEC Marketing Rule 206(4)-1
MediaPath Insights, Financial Services 2026 (citing Akamai, 2024) — historical narrative only
FINRA Foundation NFCS, 2024 — historical narrative only
McKinsey, 2024 — historical narrative only
Every figure is as reported by the named source, and forecast values are directional. Pre-2025 figures appear only as explicitly dated historical narrative and never in a stat tile or figure: eMarketer 2024 ($33.8 billion), Akamai 2024 (36% of counterfeit-domain traffic), FINRA Foundation NFCS 2024 (8% entry rate) and McKinsey 2024 (about three banking relationships). Issuers carried inside MediaPath's Financial Services 2026 report are cited through that report rather than bare. The financial media network CAGR published with the $640 million and $1.22 billion forecast endpoints is not reproduced, because it cannot be reconciled with those endpoints as a single-year rate. MediaPath holds no isolated measured outcomes, and none is claimed here.