Latest / Startup Stories with Fexingo: Conversations About Founders, Funding, and Building Companies from Zero / How Plaid Built the API Layer for Modern Finance
Transcript
- Lucas: So there's this number that's been stuck in my head all week: Plaid now reaches one in four Americans with a bank account. That's roughly 80 million people, and most of them have no idea the company exists. Luna: Right, it's that invisible middleware layer. Like the electrical wiring behind the wall. You only notice it when it breaks. Lucas: Exactly. And that's the whole point of today's episode. Plaid is arguably the most important fintech company that no consumer ever directly interacts with. It started in 2013 as a Y Combinator project—the founders Zach Perret and William Hockey originally built a budgeting app called Tally. Luna: Wait, Tally? I thought Plaid was always about connecting bank accounts. Lucas: Common misconception. Tally was supposed to be a smarter Mint.com. But when they tried to pull users' transaction data from banks, they hit a wall. Every bank had a different system—some had APIs, most didn't. They were spending 80% of their engineering time just plumbing data, not building product features. Luna: So they realized the actual pain point was the data access layer, not the budgeting UI. That's a classic pivot. Lucas: Exactly. They pivoted from being a consumer app to being an API company. And they called it Plaid. The insight was dead simple: every fintech app in the future would need to connect to a bank account, and if they could build that integration once and sell it to everyone, they'd own the pipe. Luna: But how did they solve the technical problem? In 2013, most banks didn't have developer-friendly APIs. They were still running on COBOL mainframes. Lucas: They used a technique called screen scraping. Essentially, they'd log into the bank's online portal using the user's credentials, pull the HTML, and parse out the data. It's fragile, it's slow, and it breaks when the bank changes its website. But it was the only option back then. Luna: And banks hated it. They saw it as a security risk. There were lawsuits, credential-sharing bans. Lucas: Right. Plaid's big early challenge was building trust with financial institutions. They started pitching themselves as a security layer—instead of giving your banking password to a random budgeting app, you give it to Plaid, and Plaid handles the connection with enterprise-grade encryption. They basically said, 'We're the professional bouncer for your bank data.' Luna: That's a smart framing. So who were their first customers? I'm guessing not the big banks. Lucas: Startups, mostly. Venmo was the breakthrough. In 2017, Venmo integrated Plaid so users could link their bank accounts to send money. That was massive—it gave Plaid legitimacy and scale. Then came Robinhood, Coinbase, Betterment, SoFi. By 2018, Plaid was the default way to connect a bank account in fintech. Luna: And that network effect is the real moat, right? The more apps integrate Plaid, the more data Plaid has, the better their integrations get, the harder it is for a competitor to replicate. Lucas: Exactly. By 2020, Plaid had integrations with over 11,000 financial institutions. That's a decade of relationship-building and engineering. No startup can wake up and match that overnight. And that's why Visa tried to buy them for $5.3 billion in 2020. Luna: Which the Department of Justice blocked on antitrust grounds. They argued it would give Visa a monopoly on online debit transactions. Lucas: Right. The deal fell apart in early 2021. And since then, Plaid has been on its own, raising money at a $13.4 billion valuation in 2022. The question is: can they sustain growth without Visa's resources? Luna: I think so, but the landscape has changed. Open banking regulations in Europe and the UK are forcing banks to build standardized APIs. The screen-scraping era is ending. Plaid needs to evolve from scraper to connector. Lucas: That's the billion-dollar question. Plaid's actually been preparing for that for years. They launched Plaid Exchange in 2019, which is a standardized API gateway that banks can use to offer secure data access. They're trying to become the switchboard, not the scraper. Luna: So they're positioning themselves as the middleware that sits between banks and apps, regardless of how the data flows. If a bank has an open API, Plaid connects to it. If it doesn't, Plaid scrapes. Either way, the app developer only writes one integration. Lucas: Exactly. And that's the durable value proposition. The app developer doesn't care if the bank uses an API or a mainframe—they just want the data. Plaid abstracts that complexity. Luna: But there are competitors now. Finicity was acquired by Mastercard. Yodlee has been around forever. And there are open-source alternatives like Salt Edge. Lucas: True, but Plaid has the distribution. They're already in thousands of apps. Swapping out a data provider is a pain—it requires re-authenticating users, re-testing edge cases. The switching cost is high. And Plaid keeps adding features: income verification, asset reporting, anti-fraud signals. Luna: Let's talk about one concrete number: Plaid processes data for 80 million consumers. How many transactions does that represent? Lucas: They don't disclose that publicly, but I've seen estimates of over 500 million API calls per day. That's a massive data moat. Every time someone uses Venmo or Coinbase, Plaid's models get smarter about detecting fraud or verifying income. Luna: And that data is the foundation for new products. Plaid's income verification product, for example, lets a lender see a borrower's payroll deposits in real-time instead of asking for pay stubs. That's a direct competitor to traditional credit bureaus. Lucas: Yeah, they're moving up the stack. They started as plumbing, now they're selling analytics. Their signal product helps fintechs prevent fraud by analyzing bank account behavior. Their consumer-permissioned data product lets users share specific data—like 'show that I pay rent on time'—without giving full account access. Luna: That's smart because it addresses the privacy concern. Consumers are more willing to share data if they control the scope. Lucas: Exactly. And it's a regulatory hedge. As open banking laws expand, Plaid wants to be the trusted intermediary, not the company that scrapes your password. Luna: So where do they go from here? I mean, they're already in one in four bank accounts. What's the ceiling? Lucas: I think the ceiling is becoming the standard infrastructure for all financial data. Not just bank accounts—investment accounts, payroll, insurance, maybe even crypto wallets. They want to be the universal API for finance. Luna: That would require massive expansion of their integration network. And it invites more competition from big tech. Apple is building its own financial services layer. Google is working on Google Pay. They could decide to build their own Plaid. Lucas: They could, but would they? Building integrations with thousands of banks is a slog. It's not glamorous engineering. And Plaid has a decade head start. Apple might just partner with Plaid rather than build it themselves. Luna: I think you're right. The moat here is not technology, it's the relationships and the integration network. And that's hard to replicate. Lucas: So let's leave listeners with this: Plaid is a perfect example of a startup that solved a boring, painful problem better than anyone else. They didn't invent a flashy new technology. They just made it ten times easier for developers to connect to a bank. And that turned into a $13 billion company. Luna: And the question for the future: as open banking becomes the norm, does Plaid become more valuable as the switchboard, or does it get commoditized by standardized APIs? What do you think? Lucas: I think they'll survive commoditization by moving up the stack into higher-value data products. The API might become a commodity, but the insights from the data won't be. That's the bet they're making. Luna: All right, that is Plaid in a nutshell. Next episode, we're looking at a company that took a very different path: bootstrapped, profitable, and still growing like crazy. Lucas: Can't wait. For Startup Stories, I'm Lucas. Luna: And I'm Luna. See you next time.