AI将如何重塑2026年的金融科技

2018 and 2019 in fintech was late spring. You get into 2020 and COVID, and that was utter insanity of a story.

25% of all venture dollars in that period went into fintech, which is—

Wow. 25%.

The stat after that is not a good stat: starting in the second half of 2022, basically 0% of venture dollars went into fintech.

A drought, maybe.

Yeah. Yeah. Fintech winter was the second half of 2022. Most of 2023 and 2024, things started to thaw a little bit, and now we're very much back in spring.

It turns out the biggest use case for AI is fraudsters committing fraud against financial services companies. Financial fraud is growing at 18% to 20% a year, which is insane, and it's already a huge market. I mean, the cat will win long term, but the mouse is winning right now.

Zach, David, we did this podcast, I believe, 7 years ago, and it's great to have the gang back together. Thanks for joining.

Thank you for having us.

Great to be here.

Of course, a lot has happened since the last conversation in our personal lives, and a lot has happened in fintech more broadly. I was listening to the episode that we did the last time we spoke, and we were talking about what had changed in fintech from the early 2010s to just before 2020. I'm curious if we could check in or reflect back as to, since the last time we spoke to now, what have been some of the major themes in fintech?

Catch us up. If someone was in a coma after listening to the last episode and just woke up and said, “Hey, what's changed in fintech?” what would we say?

Let's see. The last time we talked was probably 2018 or 2019. Is that right?

Yes.

Yeah. A lot. There have been a bunch of different eras, or maybe we can think of it as almost seasons in some sense.

2018 and 2019 in fintech was, I guess, kind of late spring. A lot of really good growth—the industry had a name. The name probably came about—I actually think, David, you created the name, but no one will give you credit. I will give you credit. I think you created the name in 2015.

We now had a name for this industry. We had gone past, “Oh, some people are maybe building financial services products,” to, “All right, we like it. It is an industry, and there are a lot of things being built.”

You saw a million flowers bloom—to really overextend this analogy—from 2014 and 2015 up until 2019 and 2020. You saw zillions of first-time founders asking, “Hey, can I take this thing outside of a physical bank branch and deliver it to a consumer digitally?”

You saw applications like Robinhood come up and grow incredibly well. You saw all sorts of neobanks for X, Y, or Z submarket. Those were everywhere. You saw crypto—the first crypto apps really start to emerge and grow a lot.

Then, from 2019, you get into 2020 and COVID, and that was just utter insanity of a story. The first few months of 2020 were totally normal. Then you get into early COVID, where everything froze. Basically every business locked up, including all the fintech companies.

But within 2 to 2.5 months, you then had this total inversion of fintech. You went from late spring to a big, EDM-pumping summer. The EDM music turned on very loudly, very quickly.

You had this insane growth period for fintech from mid-2020 through the end of 2021 and even into early 2022. A lot of new companies formed, but every investor—whether venture or public markets or whatever it was—wanted to push money into fintech.

You had this huge boom in funding, and tons of new stuff grew. It was a really fun and very chaotic time—honestly, a hard time to manage because the feature chase, the things we had to build, were going so rapidly.

I think 25% of all venture dollars in that period went into fintech, which is insane.

25%.

It's a crazy stat, actually. I think it's a great stat. The stat after that is not a good stat: starting in the second half of 2022, basically 0% of venture dollars went into fintech.

A drought, maybe.

Yeah. Yeah. Summer went into a very, very short fall. That was kind of mid-2022 and then immediately into winter. Fintech winter was the second half of 2022. Most of 2023 and 2024, things started to thaw a little bit, and now we're very much back into spring.

Yep. Different format, but it's been a fun cycle of the seasons. Totally. I think, even to describe what drove some of the seasons, the rate cycle was a big part of that from a macro perspective. Having very low rates kind of drove ZIRP—obviously, not unique to fintech, but to technology broadly—and certainly a lot of lending volume in the space grew massively in those periods.

The one benefit that's shown up more recently in fintech, in the thaw period, is that rates went up and it sort of shifted the mix of revenues for these fintech companies from lending-driven, origination-oriented stuff to deposits.

Many of these fintech companies decided—I forget the exact timing—to go full-stack. You saw fintech companies like SoFi buy banks, LendingClub, and I think Square got an ILC charter. Robinhood, Mercury, and many of these companies are generating very significant percentages of their revenue and profits today from deposit flows as rates have gone up.

That, I think, has helped thaw the market to some degree more recently.

Yeah. In 2018 and 2019, fintech was a startup industry. Having gone through this entire cycle—some ups, some downs, but a lot of maturation and expansion—we've ended now with fintech, in my opinion, synonymous with financial services.

It goes beyond just financial services as well. You've seen a few themes emerge. One thing that we said for a long time, that Andreessen Horowitz also likes to say, is that every company is a fintech company, and that was quite common from 2018 onward.

Now you see the emergence of embedded finance. Some Plaid customers are Ford and John Deere—companies that do have captive financial services embedded within them, but you do not think of them as financial services companies. Or large billers. The category has expanded quite a lot.

Then you see the banks themselves saying, historically, “We need to be fintech companies too.” Now they're saying, “We are the biggest fintech companies.” They invest heavily in technology.

You've seen the startup industry now become mainstream and part of the fabric of financial services, but also powering experiences well beyond financial services.

Let's go deeper into where we are today and where we're going, given that we're in an exciting period. Is it still macro in terms of—are we still early, with a lot of things to be built? What are some of the spaces you're excited about?

Zach, maybe you take the first one.

Plaid, ourselves, have gone through a few phases, and we're lucky that we have this really broad view of what's happening in fintech. I'm going to keep calling it fintech, but at this point, realize that I mean financial services plus.

The things that we're seeing today are very different and much more varied than they were before. Version 1 of Plaid was, “How do we create access for everyone?” I would say largely the fintech industry was focused on the same thing.

Instead of making you walk into a bank branch to open a bank account, how can you open a bank account on your mobile app? Instead of making you carry money and go to an exchange when you're trying to cross a border, how can we create a digital way to do remittances so you can actually move money across the border a little bit more easily?

You could apply that across basically every product that the banks were building at the time. We've solved the access problem—not completely, not in every little niche, but for the most part, we as a collective industry have solved the access problem.

I grew up in a small town, and there was only 1 bank in our town. If you didn't happen to be a member of that bank, you couldn't get a loan easily. Now, if you live in that same town, you just go online and apply for a mortgage, and you get 30 mortgage offers in an hour. Or you can do it with Rocket Mortgage and be done in 5 minutes.

These are awesome experiences. That said, what we've done is taken traditional financial services and made it digital. We haven't necessarily made it excellent. That's the next horizon for us.

A lot of things that we've been investing in now are things like credit scoring. How do we make credit scoring more logical and something that a consumer can understand?

If you get a new job and your income goes up but your expenses don't go up, you were a better loan risk. However, that doesn't show up in your credit file for many years because your credit file is a long history of your repayments. It's not necessarily indicative of your free cash flow.

And so that is the next horizon that a lot of the fintech companies I'm seeing are starting to solve. That's one big area. It's kind of solving those endemic, long-lasting problems—things like fraud, credit scoring, and so on.

The second is making financial services really easily available in places where you might not have otherwise thought to find them. So, putting BNPL on everything. Yeah, or issuing a card everywhere, or issuing a wallet everywhere.

Now we're entering this phase where fintech is everywhere. Not every company is a fintech company, but every consumer is surrounded by fintech in all the places they might want to go. The future horizons are always looking at the next few things that are happening. We look at AI and agentic financial services, and right now it's mostly hype and people talking about it. There are a few interesting use cases, but fast-forward 2 years and the way that you get a mortgage is going to be by talking to an AI application, because that is just the most efficient, fastest way to do it. That's been a fascinating one to watch, and seeing what's going on with stablecoins is fascinating as well. Lots more to come.

Just on that note, is crypto basically just fintech? People said it was the new version of the internet. Maybe, hopefully, that still happens, but in terms of where it is right now, is it mostly just a subset of fintech?

Well, David, you're an investor, so you probably know better than me. My take is that, ultimately, I don't think consumers change all that much over time. The kinds of things that a consumer would have wanted to do 5 years ago are similar to the kinds of things they might want to do today, but the form factor in which they can do it is very different.

Five years ago, a consumer might have wanted to speculate. You can speculate on gold, you can speculate on a few of these other things, and Bitcoin and other coins made it very simple for consumers to speculate. Great, you can pull up an app and speculate on things. Speculation continues; the form factor has changed.

Another thing that consumers like to do is make predictions. In the past, you might make a bet with some friends. Now you might go on Kalshi or Polymarket and enter prediction markets, or you might do that via Robinhood or whatever it is.

Other things that consumers like to do are spend money, save, invest, and so on. Inasmuch as consumer behavior doesn't change, it's a question of how and where crypto and fintech fit into the existing set of consumer behaviors.

So I think if you look at what a bank does, they're roughly tailored to what consumers want. Consumers want to save money, invest, get loans, and so forth. I think the wisest product-development strategy is to take the things that consumers already do and make them newer, easier, more accessible, and so on.

I suspect that there will be a convergence of one side of crypto and core financial services—whether that's exchanging checking accounts with dollars in them for checking accounts with USDC in them, wallets with USDC in them, or something similar. I think there's a convergence that will likely happen there, but crypto also does some crazy, out-there stuff and really pushes the bounds on innovation.

I'm not sure that that's necessarily going to end up merging with banks, but who knows?

Totally. I totally agree with what Zach was saying. I think part of it is culture and how people, to Zach's point, want to interact with financial services. I think part of this has been driven from a regulatory perspective.

The more meta theme, as I've watched fintech evolve—and I think this is permeating into crypto—is just how large incumbent financial institutions are embracing innovation and technology at large. I'd refer to my crypto colleagues, who are much deeper in this space than I am. A lot of the enthusiasm here, I would say, is about the existing financial system adopting things like stablecoins or maybe even tokenizing real-world assets.

I think that's different from a lot of the more frontier stuff that the team had talked about internally, which was more purely decentralized and about owning the internet. But I think for crypto to go very mainstream and plug into the broader financial system, that probably is—and will continue to be—what happens with what Zach and the team at Plaid have done over the last 13 years. It's remarkable.

I can't take credit for creating the fintech term. You created the enabling infrastructure to create the industry in many ways. You now have hundreds of millions of accounts connected, and you're, to your point, bringing this whole ecosystem of value-added services and analytics to make financial products better.

I think while we saw different seasons over that period—high fever and long winters, and euphoria in some moments—many of these companies are now bigger than ever. Robinhood is now, I don't know, a $100 billion public company. I looked up SoFi's stock price; they're a $35 billion public company. Affirm is a $20 billion company. These are outcomes that you couldn't even imagine.

Revolut.

Yeah, I mean, Revolut is $75 billion for new investors. That phenomenon isn't just US-centric, to that point. It's become a global one. Nubank is a $100 billion company in Brazil. My good friend Pierpaolo, who runs Ualá in Argentina, Colombia, and Mexico—these companies have worked, and they've proliferated and brought access to financial products everywhere.

I think that trend will continue. They started often with point solutions, and they perfected whatever their wedge product was. Many of them have now rebundled: they want to become the full financial picture for their customers, whether that's through cards, accounts, or lending. Again, many of them have gone full-stack and actually bought banks, hold deposits, and generate significant revenue from that float.

I think the other meta theme, which has been interesting and is accelerating now with AI, is the posture of many incumbent financial institutions toward fintech and technology broadly. I saw this firsthand as an investor back at Spark Capital, as a founder, and then inside Goldman—even their own evolution and posture toward technology.

For a long time, many of these institutions were like, "If the technology wasn't built there, they weren't interested." Goldman had literally created its own email client. They didn't operate on Outlook or Gmail; they had this thing called Orbit. I don't know why Goldman Sachs needed to create its own email client, but that was a window into the psychology from a technology perspective.

Don't they still use SecDB internally? They have their own database that they built?

That makes more sense to me because it was a centralized risk system for managing all their trades. But an Outlook equivalent makes no sense.

Then I think there was this period where many of the large institutions were saying, "We want to be the fintech companies ourselves," and Goldman went very aggressively into Marcus, and others followed suit. I think there's been a bit of a humbling. Maybe I'm using Goldman as one lens, but more broadly, I think the positive impact of that experience made them more open to adopting the best technology that exists in the market, and they're no longer building everything in-house.

A lot of where I've been spending time over the past several years has been in fintech companies that lead with software, ideally have the potential for a network effect, and are selling into these larger financial institutions and solving real workflow challenges for them.

I think we're at this interesting moment where, because the software itself can actually do the work with AI, there's this bottoms-up momentum and top-down pressure that's accelerating cultural change. Many of these institutions are beginning to adopt products like Cursor, GitHub Copilot, and a broader ecosystem of AI products across their employee base. People are seeing the productivity gains, and unlike prior periods of product cycles or platform shifts, if you were the CEO of a big bank and said, "Do I need to be in the cloud?" that was sort of an esoteric question.

Now it's like any CEO or board member can plug a prompt into one of these models and intuitively understand the impact it could have on their business. I think that's broadening the aperture, at least from my vantage point, of what fintech is.

And it's really, I think, to your point: just financial services, and software in large part sold into financial services as well.

Yeah. And, David, say more about that change, around when it went from 25% to significantly less than that. What was changing in these businesses that caused that? You mentioned the macro environment. Is there anything else we could learn from it? And, more around now, where are you particularly excited to invest, or what are the different subspaces that you're looking at or excited to?

I think the 2021 period was wild for lots of reasons. Financial services is and remains one of the biggest parts of our global economy, and I think people often get overexcited, maybe by TAM. Every venture firm created a fintech team and was deploying a lot of capital to that market. Many of these companies have continued to succeed, but I think there was probably too much euphoria going into that space relative to the amount of dollars.

No, I think it was the exact right amount of euphoria. It was just the pullback afterward that was the issue.

Exactly. Again, part of that was that when rates are zero, you can lend money and grow very quickly, and there's a lot of margin to capture there. I think when rates go up, your cost of capital goes up and that margin shrinks. There's a natural ceiling on borrowing that people face, both from a regulatory perspective and from a consumer appetite perspective. So the business models of a lot of companies on the lending side compressed.

But you also have to look at the underlying growth rates of these apps. They were insane. You look at the number of consumers who were signing up to invest, signing up to take a loan, signing up to buy Bitcoin, or whatever it was.

Totally.

We just looked at the charts, and if the app was growing at 25% a month, it was actually a great venture investment. You might know that the music was going to slow down or stop at some point, but 25% monthly growth is insane.

Totally.

Yeah, totally. This was stimulus, and there was a lot of—

Helicopter money everywhere. There were a lot of reasons they were growing that fast.

100%. And look, from an industry health perspective, I think things have normalized, but the companies continue to grow and succeed. Again, the great ones are bigger than they've ever been. There was a washout, and there were a lot of fintech companies that died or shut down in the second half of 2022 and the first half of 2023. There were a lot that went sideways for quite a while, and a lot of lenders especially that basically closed up shop, merged, or did things like that.

But the ones that succeeded coming out of it, across all of fintech, were much stronger for it.

Totally. So, as you said, if you started off with a neobank and all it did was have a checking account, a savings account, and maybe a card, then in this period, if it wanted to survive, it needed to build the lending side of its offering. Or it needed to build the investment side of its offering, so it expanded there. Now you've come out with these much more full-fledged, long-lasting companies. The winners became even more so the winners, and there was an unfortunate number of companies that also didn't make it.

I'm curious how we look at the investable universe, or how we divide it. Is it that there's a certain type of form factor and each region is going to have its new banks, so to speak? Or is it by form factor or value proposition? How do we think about the universe? How do we map it?

It's been interesting. From our vantage point, we haven't made as many consumer fintech investments in recent years as we have historically. Part of that is simply that it's more expensive to acquire customers and hit the kind of scale you need to really be venture-scale outcomes. That's a function of consumer acquisition channels getting more expensive, while some of these companies started earlier, when it was easier to acquire customers and then build massive LTV with their existing customer bases.

That does change around the world. In some markets, people were entering the formal financial economy for the first time, and offering a fee-free, mobile-first bank account and a debit card literally gave them access to e-commerce and things like Netflix, Spotify, and Amazon for the very first time. Credit doesn't exist equally in every market around the world, nor do credit bureaus and credit data. So there's still tons of interesting macro opportunity from a financial product perspective, especially in emerging economies.

I think AI could be an interesting catalyst for a new resurgence of consumer fintech. There's always been this promise of self-driving money, or PFMs that actually do the work for you and help you make—not just give you advice, but actually help you earn, save, and spend better. We've yet to see as many of those companies today, but I think the technology might be ripe. I'm curious if you're seeing this on your side, to actually deliver on that promise.

Yeah. You know, it's funny: when we think about prospective apps, the app that I wish existed, I wish there were a self-driving money app that I could just say, “Hey, my paycheck goes in here. Sweep enough money into my checking account so that I can pay my daily expenses, but put all the rest into this high-yield savings account, and invest this percent of it in the market.” I wish that this thing existed.

Actually, I don't know that that's necessarily a very good app to build, because I'm a weird power user. I have insane trust in fintech companies to do all this stuff for me. I understand all the actions that the agent would take, and I have enough background in the space that the actions seem logical to me. But if I gave that to my mom, she'd be like, “Where's my money? What's going on? I don't trust this thing. Wait, why did it move money over there?” She'd have all these questions.

So I'm not sure that I'm necessarily the best person to judge this. I have all these visions of the prospective apps that should exist out there. But for us as Plaid—and, in a lot of senses, for you as an investor—certainly for us as Plaid, our job is to build the platform, figure out what emergent behavior starts to exist on it, and then optimize for that emerging behavior as new, interesting companies start to emerge.

That's how we think of our job. As it relates to AI, our job is to build tools that allow consumers to safely link their data with agents. Then let's build tools that allow those agents to take the proper actions, whether that's just analyzing data or actually moving money or something else. Let's build tools that allow those agents to take those actions.

Then let's see what happens and have a team that's constantly looking at the emerging behavior and figuring out: Is that a good thing? Do we want to optimize for that? Has that enabled some new vector of risk that we need to avoid? That's the thought process we take across all the things that we do.

A lot of it is, “If you build it, they will come.” You just don't know who will come, what they'll look like, or what exactly is going to be the next big thing. But we have to be very prepared to react when we see it.

Yeah. And I think, as a result, we've been focused on maybe more known problems. There's so much work that happens inside all these large financial institutions that's just done manually by expensive people, frankly, across risk, compliance, legal, vendor onboarding, and treasury management. I can go on and on.

That's largely where we've been spending time: companies like Moment, which had built fixed-income trading infrastructure. If you're a wealth-management client of JPMorgan today, building a bond ladder is still a manual process. You're picking individual securities one by one. That's insane. That hasn't existed for at least a decade in equities. So there's a ton of opportunity to solve basic problems like that.

And I would argue that you can build very large software and platform-style businesses on the back of that. Whether it’s a company like Salient, which is bringing voice agents to loan servicing and collections—the idea that a voice agent can speak in 50 languages, fully compliantly track UDAAP, do welcome calls and payment reminders, and actually deliver a better customer experience because it can speak their native language and get better results. It’s infinitely patient. That is a really interesting opportunity at the moment, in large part because it’s unlocking markets that were never particularly interesting to software companies because IT budgets were small, and now the TAM is largely labor. And so that’s been one of the reorientations that we’ve seen over the last few years, from largely financial product-led companies to software businesses in financial services writ large.

Zach, you wrote about your predictions for 2026. Maybe share one we haven’t gotten to yet around where things are going and what you’re particularly excited about.

I was at a dinner a couple of weeks ago, so this might not be a prediction. This might just be a recognition of current truth. Someone asked the table, “What’s the biggest use case of AI in financial services?” Some people had answers, and then it got to me. I flippantly said, “Doing fraud.”

It turns out the biggest use case for AI is fraudsters committing fraud against financial services companies. I said it jokingly and then realized, as I was saying it, that this is actually the correct answer. The entire table was like, “Yeah, okay, that’s the correct answer”—sadly.

We’re at this point in the ecosystem where AI has so much potential to change things. Who’s using it the most? It’s the fraudsters. Right now, financial fraud is growing at 18% to 20% a year, which is insane, and it’s already a huge market.

I guess, in that vein, one of my predictions for 2026 is that, unfortunately, financial fraud is going to continue to accelerate in a way that we don’t quite understand and probably can’t quite feel out or predict yet. It’s a cat-and-mouse game, but the mouse is winning right now. The cat will win in the long term, but the mouse is winning right now. It’s kind of a depressing prediction, but I think it’s likely.

What are you guys doing about it?

Well, we build an anti-fraud product suite. This—I promise—was not me teeing up a chance to brag about why, but I will gladly brag about it. It’s a hard problem to solve, but if anybody can try to figure it out, it’s us.

Well, we can’t solve it all. We can solve pieces of it.

So, we build an anti-fraud product suite. It’s called Protect. Within that, we have an analysis of every user and every user action, and we can assign a score to say, “What’s the trustworthiness of this user, this account, or this user action that they’re taking?” We pull this data and build it based on looking at every user action that’s taken across every fintech company that we work with, plus data coming from the bank account, plus device data, plus a zillion other data sets that we match it all with.

It’s the first network-linked, cross-fintech, cross-bank type of anti-fraud tool. It’s awesome, and it adds some amazing signal to the companies that we work with. But this is one of very many solutions that need to exist. We’re starting to get good at fighting deepfakes as well, as an industry and at Plaid specifically, but we’re still very early there.

Have you heard of pig butchering? For those listening to the podcast, I’ll explain it briefly because it’s kind of a gruesome term. Basically, you get a text message that says, “Hey, how are you doing?” Don’t ever respond to those. But if you do get one and respond to it, they strike up a conversation and eventually find some complex way to ask you to give them money.

When you execute that transaction, you have just sent money to a total stranger on the internet. Yes, they’ve stolen it—that is what happens in 100% of cases. That used to be done through human factories in Malaysia, where they would have people locked in rooms sending text messages to unsuspecting people in the US, mostly, but around the world. Now that’s all AI. You don’t need these human factories anymore. The AI can do all that. AI is just getting better and better and better.

How do we fight that? Because it’s a human taking an action that they think is sending money to a friend, and they’ve been tricked, but it is fraud. It’s very hard to fight that kind of fraud.

There are so many more tools that we need to build as an industry collectively, and of course at Plaid specifically. Totally.

We were talking about the different eras of fintech. I’m curious: What have been the different eras of Plaid? Of course, there was the acquisition that didn’t go through with Visa, and the ups and downs that you guys have had alongside the macro. Obviously, you’re in an incredible position right now. Talk more about the different eras of Plaid, or how the Plaid vision has evolved or stayed true to the original.

I started Plaid—I started working on a thing that wasn’t Plaid but pivoted into Plaid at the very end of 2012. We pivoted into what we were doing in mid-to-late 2013 and launched to the world in 2014. It’s been a good 11 to 13 years, depending on how you count that series of bad products that we built first.

In a brief aside, David, I don’t know if you know this: David found Plaid. He was the first investor and led the seed round at Spark Capital. He actually sourced the deal when you were an associate, I think, at Spark at the time. Then he went to Goldman Sachs around the time that Goldman invested. You weren’t involved in the investment specifically, but you were at Goldman at that time and were probably helpful in the background. Then you came to Andreessen Horowitz, and Andreessen invested. You’ve been a huge friend and supporter of Plaid over the years.

We owe a lot to David, and a huge amount of thanks. He also creates all the important industry terms, so the fintech industry owes a lot to David.

I don’t know. Anyway—

Plaid started, let’s say, in 2014. From 2014 to 2019, it was all about linking bank accounts: How do we enable you to link a bank account so that you can gain more access to financial products broadly? Link a bank account so you can pay a friend on Venmo. Link a bank account so you can get a loan on LendingClub. That was phase 1.

From 2019 to 2020, we continued to grow in that vein. In January 2020, we signed paperwork to sell the company to Visa. It was still late spring—we didn’t know that COVID was coming. We didn’t know that the EDM music would turn on.

David, I remember chatting with you, I think it was February or March—probably March, right when COVID was just beginning. You were like, “Wow, you really timed that well.” Then the business started ripping, and I was like, “Oh, that’s a very expensive free call option on the business.” So walking away from that is pretty—

For the next phase, the EDM music just started getting louder and louder, summer started happening, fintech started growing, and people were stuck at home. They needed to use digital finance to live their financial lives. At the end, a year later, we looked at it and said, “For a large variety of reasons, it makes sense for us to part as friends with Visa.” We’d go our own way and keep running Plaid as an independent entity. Then we raised a big up round and were off to the races.

Through that, you tell the company, “Hey, we’re selling.” “Okay, great.” That’s a really hard thing—to convince everybody to still be excited even though you’re selling the company. A year later, you say, “Hey, we’re not selling.” That’s another very, very hard thing, because you’re telling everybody, “You’re not going to get all that cash that you thought you were going to get. You can’t buy the house. I’m sorry, but we’ll try to do a secondary soon, so maybe you can buy a car.”

You have to really change the culture. It’s almost a refounding moment at that point. Then you go through the rest of the summer, and that was great—lots of growth. But then you get into fintech winter, and that’s another moment where we have to all come together. Our customers are growing more slowly. Yes, we’re producing great products. Yes, Plaid is growing, but it’s not the growth that we’re used to because we’re in fintech winter.

It’s nice to finally be back in spring, but there are definitely a lot of ups and downs on that journey. I think there were multiple refounding moments, or multiple crucible moments, along the way.

Was there a period in that where you found your second wind? Maybe you always had it, but from the outside, it felt like your product velocity really increased at some point in the last 2½ years.

Yeah, it has. I shifted my role quite significantly. I’m our chief product officer, so I’m involved in all of the product stuff.

A lot of it was really about building the data set to a size where we could actually run analytics on it. We build fraud scores that look at your actions relative to every other user we see on our platform and identify whether you’re anomalous. If we didn’t have enough data to identify whether you were anomalous, it wouldn’t be a relevant score for us to build. We got to, first, enough data, and then, second, we finally figured out how to build and launch products quickly. That’s been one of the most fun things for me, actually.

I think I was less happy during the period of EDM pumping, fast growth, and everybody throwing money at fintech. I was a little less happy because I didn’t think I was adding differential value. I was just running as fast as I possibly could. Maybe I made some good decisions, but it all didn’t matter because everything was up and to the right.

I think I was happier during that winter period. I was like, “Oh, man, this is where we become an amazing company long term. This is where we prove ourselves, really step up, and help our customers. We launch the next wave of products that really matter.”

But I think I felt similarly, to be honest. Having done fintech since, I don’t know, 2011, people thought that felt early—to be investing then. Then everybody found out that this thing existed. Everybody became a fintech investor from 2019 to 2021. Then some of the best fintech investors in the world came out on podcasts and were like, “Fintech is dead.”

I’m like, “Fintech is dead? Everyone should go home except for people that are building fintech and making products that people use. You guys can leave and just stop investing in fintech. We will continue. The fintech team is still here, despite the brand.”

The brand.

I think that’s actually benefited us, selfishly, but I think it’s tested the true believers. In some ways, it’s brought the community together, I would argue.

Yeah. The tourists go home. We saw it on our team, too. There were people who joined Plaid in 2020, when the music was loud and it seemed like the industry to be in. Then they chased the next trend, and the next trend. While we’ll miss them, and they’re nice people, the people who are focused on it now are the people who really want to be here in the long term. They deeply believe in the mission, and they’re in it in the way that we all want to be in it.

Where are we now in the cycle? How should we think about this moment?

Early to mid-spring, I would say. We see green shoots and lots of emergence. It’s been a pretty good year for many parts of fintech, and it’s been a shaky year for others.

If you look at the lending markets, it’s not as bad as last year, but it’s not as good as it was. There are elements of the economy that are pretty scary, and a large part of consumer spending is being propped up by a small number of people. There are all these things that are scary, but for the most part, you continue to see companies building very solid products.

You do see great startups emerging, but they look a little different than they used to. They’re thinking more responsibly about markets in the long term. They’re thinking more about profitability and growth. You’re also seeing the insanity of AI funding continue in AI land, and some of that is starting to bleed into fintech because you’re seeing these fintech AI products start to emerge.

I would say we’re in spring: lots of green shoots and lots of exciting stuff. There’s still some snow in the background, and the snowmelt is still happening, but it’s looking pretty optimistic right now.

Awesome. Let’s wrap on what 2026 and the near-term future look like. David, how are we approaching it at a16z?

It still feels like we’re in the early innings, even in AI land and in spring. I’m incredibly excited and enthusiastic about the momentum we’re seeing, again, largely from software companies selling into financial institutions. That’s been our orientation in the fintech ecosystem.

I sat on the board of a company called Moment, which we described earlier. It’s now bringing some of the largest wealth management platforms online. You’ll see them—they announced LPL—and we have a number of other large institutions that we’ll be announcing early next year.

Companies like ModernFi have built bank-to-bank deposit marketplaces that are really starting to grow and see significant volume in that network. More broadly, I’m really excited by the opportunity for AI to actually do the work within these institutions, and by the momentum and excitement there to adopt new products.

And are we excited, David, because they’re such great customers, or because they’re so underserved, or because they’re finally transitioning? Why have we narrowed in on that focus as one we’re particularly excited about?

The industry is still massive. If I look back at even just Goldman Sachs—I know I use them as an example often—the entire firm called the middle and back office the federation. These were folks living largely in Excel, not using Excel as a modeling tool, but using Excel to track work.

There’s such an opportunity to build amazing software products to solve everything from compliance to payments to treasury management, as well as all of the manual work that goes into making the financial services industry tick. AI is creating a new window and wedge opportunity for entrepreneurs to build software companies that couldn’t have existed years ago.

The appetite for adopting new products and new software to solve some of those problems is more real than ever, because the most senior people at these institutions can intuitively understand the impact AI is having on their business. There’s a lot more conversation and momentum happening at the board level, and it’s making enterprise sales cycles for many of our early-stage companies happen a lot faster than I’ve seen in my experience investing in this space.

Zach, how about you? How do you think about things at Plaid and more broadly?

This past year, we launched, as I said, the anti-fraud suite called Protect, and there’s been tons and tons of acceleration behind that. We launched a credit score—a modern consumer credit score based on your income, your expenses, and the things that you do in your daily life.

So your score goes up if you have a higher income. Your score goes down if you start having way higher personal expenses, like the logical credit score. So we launched that; it’s called Lens Score. We launched that last year. These two things are going to be major drivers for us in the coming year.

Distributing this new version of a credit score to all the lenders and, of course, on the Protect side, helping fight this AI-driven financial fraud that we’re seeing.

For us, we’re back to hiring, recruiting, and growing. Despite the fact that fintech has been through these waves, I still think that Plaid is one of the most amazing places to work. If you want to work with big data and have a huge impact on consumers’ lives, financial freedom is the core focus of what we do.

We try to think of ourselves as the most customer-centric employer, where we put engineers in front of the customer so they’re actually talking to them. We think it’s an incredibly fun way to work. So, not forward-deployed engineering, but a forward-deployed company. We’re hiring lots of people, and I think it’s going to be a great 2026.

Zach, David, you guys are pioneers in the space, in the category, and I can’t wait to have you both back in 2030 so we can talk about how the space has evolved. Thanks so much.

Do it sooner.

Yeah, it’s so far away.

Exactly. We don’t want you to wait.

That’s true. We don’t have to wait every 5 years. Zach, David, thanks so much for coming to the podcast. Great. Good to see you both.