CHAPTERS
- 0:00 – 1:42
Fintech’s seasonal cycles: spring, summer, winter—and back again
Zach and David frame the last decade of fintech as a set of "seasons," from the rapid growth period of 2018–2019 through the COVID-era boom and the sharp pullback in 2022. They discuss how venture appetite and macro conditions amplified these cycles. The conversation sets up why the industry today feels like a renewed “spring.”
- •2018–2019 described as "late spring" with strong growth and clearer category identity
- •COVID triggered a brief freeze followed by a rapid, euphoric boom
- •2022 brought a fast transition from “fall” into a deep fintech winter
- •2023–2024 saw gradual thawing; the current moment feels like early-to-mid spring
- 1:42 – 2:55
2014–2019: digitizing banking and the first wave of breakout apps
The hosts revisit the rise of fintech as an identifiable industry, when new products moved core banking experiences from branches to mobile. This era produced iconic consumer apps and a proliferation of “neobank for X” offerings. Crypto’s first mainstream consumer apps also emerged during this period.
- •Fintech became a named, cohesive industry with many new entrants
- •Digitization of branch-based services drove mass product experimentation
- •Examples: Robinhood’s growth, niche neobanks, early crypto consumer apps
- •A “million flowers bloom” phase created many point solutions
- 2:55 – 4:45
2020–2021: COVID inversion, stimulus tailwinds, and the funding frenzy
They describe how fintech went from a short COVID shock to explosive adoption as consumers moved their financial lives online. Investor demand surged across venture and public markets, making the period both fun and chaotic to operate through. Rapid feature shipping and “growth at all costs” dynamics defined the era.
- •Early 2020 freeze was quickly replaced by massive demand for digital finance
- •Mid-2020 through 2021/early 2022 became an intense hypergrowth period
- •Funding flooded in from multiple capital sources, not just venture
- •Operational pressure: feature chase and rapid scaling challenges
- 4:45 – 5:48
Macro reality check: rates, lending compression, and the shift to deposits
David explains how the interest-rate cycle reshaped fintech business models, especially lending-driven companies that relied on cheap capital. As rates rose, deposits and net interest income became more important, pushing many fintechs to go “full stack” via charters or acquisitions. This revenue mix shift helped stabilize parts of the ecosystem in the thaw period.
- •Low rates fueled lending volume and amplified growth across tech/fintech
- •Rising rates increased cost of capital and reduced lending-driven margins
- •Many fintechs moved toward deposits as a core revenue and profit engine
- •Full-stack moves: SoFi buying a bank; Square’s ILC charter; others following
- 5:48 – 7:37
Fintech becomes ‘financial services plus-plus’: embedded finance goes mainstream
Zach argues fintech is no longer a niche startup category but effectively synonymous with modern financial services. He highlights embedded finance as a major expansion vector, where non-financial brands integrate financial capabilities. Even banks now position themselves as major technology builders rather than fintech outsiders.
- •Fintech matured from a startup wave into the mainstream of financial services
- •Embedded finance expands fintech into unexpected verticals (e.g., Ford, John Deere)
- •Banks’ posture evolves from “we need fintech” to “we are the biggest fintechs”
- •Fintech increasingly powers experiences beyond traditional financial services
- 7:37 – 10:40
Beyond access: making financial services excellent (credit, scoring, endemic problems)
Zach reflects that the industry largely solved the “access problem” by digitizing financial services, but hasn’t made them truly excellent. He points to foundational problems like credit scoring and fraud where better data and product design can materially improve outcomes. The next horizon is making systems more logical, transparent, and aligned with real consumer cash flow.
- •Access improved dramatically (mortgage shopping online vs. single local bank options)
- •Digitization didn’t automatically create excellence or fairness
- •Credit scoring is a key frontier: incorporating income/expenses and real-time conditions
- •Other endemic problems: fraud, risk, and core infrastructure gaps
- 10:40 – 15:32
Is crypto just fintech? Consumer behavior, convergence, and the ‘frontier’ edge
The conversation explores where crypto overlaps with fintech versus where it remains distinct. Zach emphasizes that consumers’ underlying desires (speculation, prediction, spending, saving) remain stable while form factors evolve. David adds that mainstream adoption likely comes through integration with incumbents via stablecoins and tokenized assets, distinct from purely decentralized visions.
- •Crypto overlaps with fintech when it maps to familiar behaviors (speculation, payments, investing)
- •Prediction markets and new platforms reshape how consumers express old behaviors
- •Likely convergence: dollars ↔ stablecoins (e.g., USDC) within mainstream accounts/wallets
- •Frontier crypto innovation may remain separate from bank-like systems
- 15:32 – 19:03
From product-led fintech to software sold into incumbents—especially with AI
David describes a structural shift toward fintech companies that lead with software and sell into large financial institutions. Banks and incumbents have become more open to adopting best-in-market tools instead of building everything in-house. AI accelerates this transition by making productivity gains visible from the boardroom down to individual teams.
- •Incumbents historically built internally (e.g., bespoke tools) and resisted external software
- •A humbling period led to more openness to buying vs. building
- •AI creates bottoms-up adoption (employee tools) and top-down urgency (executive understanding)
- •Fintech’s scope broadens to ‘software for financial services’ as a major category
- 19:03 – 23:45
Post-boom lessons: euphoria, washouts, and why the winners got stronger
They unpack why venture dollars poured in and then disappeared, tying it to macro conditions and to temporarily extraordinary growth rates. Zach notes many companies shut down or stagnated during winter—especially lenders—while resilient players expanded their product suites to survive. The downturn functioned as a stress test that strengthened the long-term winners.
- •2021 was driven by massive TAM excitement and capital abundance
- •Rate increases and normalization reduced lending economics and growth narratives
- •Significant washout in late 2022/early 2023; consolidation and shutdowns followed
- •Survivors rebundled into fuller suites (lending, investing) and emerged stronger
- 23:45 – 25:50
Consumer fintech’s trust gap vs. ‘self-driving money’ ambition
David raises the promise of autonomous personal finance, but Zach questions whether average consumers will trust agents to move money without anxiety. They discuss Plaid’s platform role: enable safe data linking and actions for agents, then watch emergent behavior and risks. The key constraint becomes trust, transparency, and safety—not just capability.
- •Potential: AI-driven PFMs that execute, not just advise
- •Reality: many users may not trust automated money movement and need explanations
- •Plaid’s approach: build safe linking + action rails; observe emergent behavior
- •New capabilities introduce new risk vectors that require active mitigation
- 25:50 – 27:43
Where investors see opportunity: AI automating manual ‘back office’ labor
David explains a16z’s recent focus on known, high-friction operational problems inside financial institutions—work still done manually by expensive teams. AI shifts the TAM from software budgets to labor replacement and productivity, enabling new wedges. Examples include modern market infrastructure and AI agents for servicing and collections with compliance guardrails.
- •Huge internal workflows remain manual across risk, compliance, legal, onboarding, treasury
- •AI makes it economical for software to ‘do the work,’ not just track it
- •Examples: fixed income tooling (bond laddering) and voice agents for servicing/collections
- •AI agents can improve CX (languages, patience) while enforcing compliance constraints
- 27:43 – 32:03
2026 prediction: AI accelerates fraud—the ‘mouse’ is winning for now
Zach’s near-term outlook is stark: the most impactful AI use in financial services today is enabling fraudsters. Fraud is growing rapidly and evolving in ways that are hard to forecast because it’s a cat-and-mouse game with fast-moving adversaries. He highlights emerging attack patterns that blend social engineering with AI automation.
- •Claim: biggest AI use case in finance is fraudsters using AI against institutions
- •Financial fraud growing ~18–20% annually, on top of an already large base
- •Adversarial innovation is outpacing defenses in the short term
- •Examples include scalable scams and increasingly convincing deepfakes
- 32:03 – 39:44
Plaid’s evolution: from account linking to network-scale analytics, fraud, and credit
Zach walks through Plaid’s phases: early focus on account linking, then navigating the Visa acquisition attempt, and later expanding into analytics-driven products. He describes the cultural and operational strain of announcing a sale and later reversing it, followed by renewed product velocity. The company’s scale now enables networked signals for fraud and new approaches to creditworthiness.
- •2014–2019: core mission was bank account linking for fintech apps
- •2020–2021: Visa deal signed, prolonged regulatory scrutiny, and eventual split
- •Multiple “refounding” moments: morale, compensation expectations, and strategy resets
- •Product acceleration required data scale + improved internal launch velocity
- 39:44 – 40:58
What ‘spring’ looks like now: green shoots, more responsible startups, AI bleeding into fintech
Zach characterizes the current environment as early-to-mid spring: innovation is returning, but with lingering macro risks and uneven conditions (notably in lending). Startups appear more focused on durability—profitability and sustainable growth—while AI funding dynamics begin to influence fintech. The overall tone is optimistic but cautious.
- •Current phase: early-to-mid spring with new startup formation and product building
- •Lending remains challenged, though better than the prior year
- •Macro risks persist, including uneven consumer spending dynamics
- •Startups now prioritize responsible scaling; AI-fintech hybrids are emerging
- 40:58 – 45:37
Near-term roadmap: a16z’s AI-for-institutions thesis and Plaid’s Protect + LendScore
David summarizes a16z’s enthusiasm for AI-driven software sold into large institutions, citing faster enterprise adoption and strong demand for productivity. Zach details Plaid’s key product bets: Protect (anti-fraud) and LendScore (cash-flow-informed credit scoring), plus continued hiring and platform expansion. The episode closes with a forward-looking view toward 2026 and beyond.
- •a16z focus: early innings for AI in financial institutions; faster sales cycles and board-level urgency
- •Software opportunity spans compliance, payments, treasury, and operational workflows
- •Plaid priorities: Protect anti-fraud suite and LendScore modern credit scoring
- •Plaid positioning: platform enabling safer agentic finance; growth and recruiting for 2026
