Building an AI-Native Finance Team with Plaid CFO Seun Sodipo

0:00I'm reflecting on my experiences using AI. And now I'm so impatient in terms of why isn't it fast enough? You have that experience in one part of your life. And you port it over to your expectations in every other part of your life. Once you're digital, it's like, okay, then like what's the cutting edge experience? And so we're really excited about intelligent financial experiences, like really grounded in people's data. I thought of durability almost like defensively. But I don't think that's playing to win and playing to survive. It's been about offense. It's not just surviving. It's about thriving.

0:29regardless of the macroeconomic situation. When you unlock velocity somewhere, the bottleneck then moves. But if the person downstream of that isn't also unlocking velocity or capacity at the same rate, like the pinch point will move. And so sort of anticipating those pinch points I found across the business is where it's also like shifted. What skill set I think has come forward a bit more. Is this thing on? Yesterday's price is not today's price.

1:06I know we're all afraid of commitment, but if you can do me a solid and follow the show on Apple or Spotify, just send us a screenshot that you followed us. Send it to Ben at mostlymetrics.com and he's going to swipe my Brex card and he's going to send you something. I really hope it doesn't go over budget, but he's going to send you a hat, maybe a dog bowl, maybe a sticker, maybe a hat. Do we have hats, Ben? I don't know. That's the hat right there. He'll send you one. Hit follow. Send a screenshot. Ben at mostlymetrics.com. Boom. Welcome back to Run the Numbers, the show where I interview the world's top CFOs.

1:38I cannot believe some of the people picking up my phone call. Today, I'm talking to Sheyoun Sudeepo, the CFO of Plaid, one of the most exciting thing tech companies in the private markets. We talk about the Plaid business model, which is truly a masterclass for anyone running a network effect business. the differences between tops-down and bottoms-up experimentation with AI, and how leaders can effectively motivate employees to use AI on their own. And finally, I talked to her about building an enduring company. Please subscribe to our show. It helps us with the algorithmic overlords. Let's get into the interview.

2:12Shayun, thank you so much for joining me on the podcast today. Ah, CJ, thank you for having me. I'm excited to be here. So I've been a huge fan of Plaid for a while. and like the further I dig, the more I figure out how much this company does. It's amazing. It's a true network. Many of the people listening may just know Plaid as, you know, the screen that pops up when they're like in their Bank of America or something like that. Maybe you can just give us the 30 seconds on what Plaid does and who it serves today. Put simply, Plaid is the infrastructure underpinning digital finance. We enable consumers to securely connect their bank accounts or their brokerage accounts covering about 12 ,000 financial institutions with any of our Plaid customers.

2:54And we're very lucky and thankful to have about 9 ,000 Plaid customers, which can be anyone from Venmo, Coinbase, Robinhood to Rocket or Block. We power about 10 million secure data sharing events each day. So that gives you a bit of a sense of the scale and sort of the activity on our network. And more than one in two in Americans with a bank account have used Plaid. That's amazing. We were talking to Blake Grace and the CFO of DocuSign. And I think of Plaid in a lot of ways, it's like a high proportion of Americans have interacted with like DocuSign or Plaid. Or it's one of those things where instead of saying petroleum jelly, you say Vaseline.

3:33If you're going to connect your bank account or to a brokerage, you could just say, hey, I use Plaid for that or I connect using Plaid. So it's definitely ubiquitous in many senses. Yeah, I'd love to turn Plaid into a verb. We should work on that. You were kind enough to say that you read the Mostly Metrics newsletter, which blew me away. And you probably figured out that I'm a total business model dork. How does Plaid make money? We heard who you serve and what the product is. How do you monetize? I gave you layer one. We have a three-layer network. I'll move into layers two and three, and then I'll share a little bit about how we make money.

4:04So layer two is like once we built this connectivity layer, which we believe is one of the largest financial data networks out there, we moved into layer two. And what layer two is? It's a financial identity network. So as I, for example, or as you, CJ, connect more and more of your financial life to our customers using Plaid, we're able to build out a pretty comprehensive view of your financial life. And so we call that the financial identity network. And we have about 150 million consumers within that financial identity network. It's a pretty comprehensive 360 view. The more you connect, the more we can build out CJ's profile.

4:38And once we have these profiles built, we've been able to take a step back and be like, what are really gnarly problems in financial services that we can now go and help solve in truly like a really impactful way now that we kind of like more deeply understand any individual's financial identity and financial life. And so we move into level three, which is what we call our Plaid Intelligence layer. And our intelligence layer, we're focused on three priority areas, which is the first is credit underwriting. The second is anti-fraud. And the third is payments. And if I take credit underwriting, for example, within credit underwriting, we're all very familiar with credit scores.

5:14And I think almost everyone I interact with has their own funny story or frustration, frankly, with credit scoring. Frankly, I remember arriving in the U.S. as a student and I had like a bit of money in my account and I wanted, you know, a prepaid mobile phone. And they like would not let me open an account that way because I didn't have a credit score. I like didn't exist within the system. And I was like, well, I can prepay and I can give you this cash I have. And they're like, no, like it doesn't work. You need to go pay as you go. And so what are our cash flow scores allow you to do and our cash flow products allow you to do is that they look at the money in, money out.

5:46So it's really powerful across the board for all consumers, but in particular for thin file consumers that don't have as deep a credit history or don't have credit profiles. And so that's not to say they're not like credit worthy, but it's just that their financial life and the way their financial data shows up just doesn't fit into the system and the scoring that exists today. So that's just one small piece of what Plaid accomplishes now that we've built this three layer network. Now, turning to how we monetize the business. Simply put, we monetize the business when any financial action happens across our network.

6:19So this could be anything from connecting your account to one of our customers. It could be one of our customers verifying your identity or your income, for example, if they want to extend you some credit. It's combating fraud with that financial identity network that we have. We start to know you and we start to know what your financial habits look like. And so when we see behavior that isn't you, we're uniquely positioned to sort of flag that to our network and be like, hey, like this person here, they're saying that they're CJ, but we're pretty convinced that they're not. And then we also make money anytime somebody initiates a payment or a network or as you're going back to that cash flow example, provides cash flow data or score data to one of our customers for an underwriting use case.

7:04I love hearing stories like this because it just brings to light how an innovation stack, if you do that first thing really well, it unlocks so many other surface areas that you can add value and then end up monetizing. Yeah, it's really interesting because I've been thinking, so I've been at Plaid now for about eight months. I like to declare that I'm no longer new. I think at some point you just need to call it. I don't know when that is. I've often debated that. It's like, is it after three months? Is it six months? I think like after six months, it starts to be like, yeah, you're not that new anymore.

7:31Yeah, exactly. So I think post six, I was like, look, guys, I'm not new anymore. But one thing I've thought about is like, how do I explain our business model? Because I thought I knew about our business model coming in. But since joining, it's much more powerful than I fully understood from the outside in. And so with that, as I think about like our business model and how we monetize, it's really interesting. And it just has like so much extensibility. It's what I'm really excited about. I'm excited about what we're working on now, but also really excited about like the problems we'll go get to solve in the future.

8:01It sounds like you realized pretty quickly then that you weren't just a connectivity company anymore, which is neat to see once you get inside a company, like there's so many more opportunities here and use the CFO. I mean, you're the person allocating resources. It's a delightful experience to be able to say, well, I have a couple of different paths. I can help people take this. It's a really great position to be in where you sort of have like an abundance of choice because not every company, right, like gets to be in that situation. but like massive credit to like all the plants that came before me and like we wouldn't get to do what we're doing today had we not built out that essential infrastructure layer to begin with that is like our lifeblood like that gives us the opportunity to then build on that layer and I start to build out like a platform and a data network so we continue to invest along all three dimensions because they each compound and reinforce each other but it's sort of like a wonderful problem to have to think about like what's the appropriate level of investment across all layers of that Plat stack.

8:57Hey, thanks for listening. We'll be right back after a word from our sponsors. Remember when pricing was simple? One product, one subscription, one invoice every month. Those days are over. The AI economy is changing how everything gets bought and sold. We're seeing usage-based pricing, hybrid contracts, and something called a credit. The finance systems built for vanilla subscriptions weren't designed for this. This isn't just a rev rec problem anymore because outdated revenue architecture can kill a great pricing idea before it ever reaches the market. RightRev lets you recognize revenue in whatever shape it comes in.

9:33Whether you're launching consumption credits or testing entirely new commercial models, RightRev gives finance the flexibility to support whatever pricing model comes next. So product teams can keep on innovating and finance teams can keep pace. If you're architecting revenue for the AI economy, learn how RightRev can help. Visit rightrev.com slash CJ. That's me. That's rightrev.com slash CJ. Hey, founders and finance folks, you know how early on everything is fast and scrappy, and then suddenly equity gets messy? Spreadsheets break, grant docs pile up, lawyers are sending new forms and collecting fees.

10:10It all shows up right about when you're already swamped. Well, if managing your cap table feels like one more frustrating thing on a very long to-do list, you need to know about Pulley. Pulley makes equity management simple and stress-free. You can issue options, model dilution, complete 409As, and more. All in one place with support from real experts when you need it. So if you're raising, hiring, or scaling, Pulley keeps your cap table clean so you can stay focused on building. Not wrangling spreadsheets. Learn more, request a demo at pulley.com slash mostlymetrics. That is P-U-L-L-E-Y dot com slash mostlymetrics.

10:47I got news for you. The ERP category is finally getting disrupted. And if you haven't heard of Realit yet, please pay attention. It's the AI native ERP built specifically to replace NetSuite. And it's already won over hundreds of finance teams. Their mission is to make the zero-day close a reality. And they're actually doing it. We're talking teams closing the books at 1.35 p.m. on the first day of the month. Companies like Windsurf, Mercore, and hundreds of others run their entire finance stack on really. Revenue recognition, closed management, multi-entity, native Stripe, and Salesforce integrations.

11:19Woo! Everything a scaling company needs. They've got 5.0 stars on G2. They're backed by A16Z and Sequoia. Heard of them. And CPA-led implementations that get you live in 45 days. That is simply unheard of in the ERP space. If your books aren't running as fast as your business, check out Rillet. Book a demo at Rillet.com slash CJ. That is R-I-L-L-E-T dot com slash CJ. That's me. Well, speaking about building, I want to talk about how you built your career and then how you evaluated what your next move was going to be. Because you worked at Stripe. You crushed it there in strategic finance. You're the CFO at Glossier.

12:01Now you're at Plaid. Was there a pattern in these choices? It seems like you're putting this financial skill set to different use cases and maybe following your interests. What was going on there? I like to make myself uncomfortable. I just keep on signing up for things that will stretch me further and further out of my comfort zone. But so I'm deeply analytical. I'm a math nerd at heart. I've been ever since like my earliest report card that my mother like sort of digs up from time to time to embarrass me. Age five, it's called numbers because it's not really math. And it's like Shayoon is very good at numbers.

12:31And that's sort of been a theme. But I was also really curious about the world of business. Like almost everyone in my family when I was growing up built their own business. They were like professionals, doctors, lawyers, et cetera, but they built their own practices. So my mother built her own law firm. My grandfather built his own like medical practice. Surprise, surprise, found myself in investment banking out of school, which, you know, many of us go through that. But I really, I really enjoyed it. And I think the question that popped up to me as I dug into different, you know, most of the clients that we served at the bank I was at were public companies.

13:00I was really struck by the question of value creation. you would see similar businesses trading at like vastly different multiples. And like that was like so surprising to me because, you know, sort of my analogy was like the world I was familiar with was like real estate. I was like, well, if two houses are approximately the same size, they're probably priced about the same. And they're like, oh, no, no, no, no, no. You got to look at like, you know, the finishings. You got to look about how it's set up, you know, look at the neighborhood, etc. And so like really starting to understand value creation.

13:24And so as I've like moved from there into investing, again, I was trying to get closer to like, How do you think about value creation when you're allocating capital, not just advising companies? And then at some point, I realized I really want to get under the hood here and not just think about what creates value, but also help really hands on the wheel build value. And as I look at the choices that I've made in my career, I think it's like every so often just looking to take leaps forward and like taking my skill set and finding a place. I think about my career moves as like Venn diagrams. So it's like, OK, what's the circle of what I know today?

14:00How do I see this evolving? Where do I want to go? And then like not completely finding place where there is zero overlap. I don't know that would be the best thing for me or for the business I'm joining. But thinking about like how the Venn diagrams could like overlap and how like the knowledge I bring could like help that business, but also could push me to learn more. And so that's what I've done within the companies that I'm in, like as I think about the roles that I take on, but also as I've moved between companies as well, is just thinking about like what is the composite? Like what do you bring to the team?

14:29What do you bring to the business? And also how are they going to push and challenge you? Shayun, you mentioned evaluating from the outside what a durable business is, what a great business looks like. Has your mental model changed at all now that you've been working to create those internally and operating? Picking industries with a groovy, durable, secular trends, like that helps. But within that, there are going to be winners and there are going to be losers. And so like really paying attention to the winners. And the thing that stands out there, I think it's like relentless customer focus and relentless innovation.

15:00We just increasingly have higher and higher expectations. And so looking at the businesses that are like really have a finger of the pulse of what like their customers need, where the industry is going, or actually like driving, moving the industry forward. I think that's a good way to make sure you don't get left behind is to be the agent of change yourself. And then another thing I think that I look at in value creation that maybe is so easy to overlook from the outside in is the team. It's the team and it's the culture and what that means, you know, in terms of how that sets up a business to go and actually execute, because it's a great thing to have a great plan.

15:32But if you can't execute on your plan, it's like words on a page. The rising expectations piece, like I had a visceral reaction to that because like I'm reflecting on my experiences using AI. So when I started out, you could say, you know, create me a poem in like Shakespearean sonnet, but it's talking about metrics. And that's the use case. And it blew my mind the first time it did it. And now I'm so impatient in terms of why isn't it pumping out like this complex analysis and doing net dollar retention rates and reading these SEC statements like fast enough. I'm paying the exact same amount I was,$20 a month,$100 a month for whatever that plan is.

16:07But my expectation level, it's been meteoric. Right. And like it just it keeps going. And then you have that experience in one part of your life and you port it over to your expectations in every other part of your life. You're like, well, if they could figure it out and if they could, you know, if they knew me and they can treat me this way, then this is what I expect. And like, we absolutely see that in financial services and with Platt. That thing we're super excited about is the shift to like intelligent finance. Digital finance is one thing, right? Going from analog to digital, fixed at board.

16:34But once you're digital, it's like, okay, then like what's the cutting edge experience? And so we're really excited about where like we see the ecosystem going in terms of like creating like personalized, intelligent financial experiences, like really grounded in people's data, because it just like really unlocks like a new level of insight consumers will have around like the health of their financial lives. And I think turbocharge or make them much more confident in like the decisions that they make around optimizing for their outcomes. And we're going to come back to that, the intersection of your product and the consumer experience.

17:08I want to take a couple minutes there to talk about how you're using AI with your team, also pushing people to embrace this vision from the bottoms up, because you've talked about the importance of combining this tops down vision with bottom up experimentation around AI. Why do finance orgs need both? So we started Bottoms Up. We're a company of builders and doers. And so you give someone a new tool and they're like, let me add it. And everyone goes off, right? We've all gone off into our back caves and started to just play and experiment with these AI tools and figure out bit by bit how they can help make us better versions of ourselves.

17:44And so we'd been doing that for several months. But at some point, I actually looked over to our friends over in like engineering. I was like, how do I take this to the next level? Like, I love the self-reported gains and efficiency that the team was coming back with. But finance is a team sport, both within the finance team as well as out with our business partners. And I was like, well, how can we, a lot of the places where we can see would really benefit from like step changes and like automation. Many hands are touching that. can't completely expect like one person to take it on themselves to go and fix like their piece of it and so that's where the tops down came in it's like frankly just enabling us to make much bigger shifts in what we can automate and how we operate as an organization but also another thing that it does is it sets the bar on expectations when each person individually is working on their own stuff like it's like hard to know yourself where do i rank on sort of our ai leaderboard how sort of automated and I as a financial professional, but the top-down expectation, it raises the floor on the team.

18:44Baseline, this is where you all need to get to. Like, this is the expectation for you going forward. Our breakout users, like, I don't, you know, I don't know it shifts too much for them, but I think it, like, helps move the team median upwards in a really powerful way. I was talking to a fellow finance friend about this, and he said the top 10 % finance people, Venn diagram reference here, there's a lot of overlap with the top AI users. Because they're high performers and they want to try something to make them even better at their job. And they're probably going to be high performers in that too.

19:14But it's like, how do you bring the rest of the team along? Our team is like ready and willing, but you know, some folks just like learn a bit differently. Everyone's busy. And so it's also like, how do I not just create top-down expectation, but then also create time and space for the team to invest in these projects and have it feel less about like a personal project or experiment and more about like team and company. It is easy, I have to admit to get lost in just like the cycle of busyness where you start measuring it in terms of like how much activity you did. Do you think there's a failure mode where there's too much experimentation and not enough direction or vision?

19:49Honestly, I haven't seen it. I sort of highly encourage the team to like default, try and start first, like the thing that you would go into some other tool to use, like just try it first in AI and see what happens and keep going if it's working or like, you know, retreat. So I haven't seen actually too, too much experimentation. Hey, thanks for listening. We'll be right back after a word from our sponsors.

20:35autonomous finance platform for the whole operation. Order to cash, procure to pay, to close, cash management, and reporting. It all runs on a unified finance context that pulls from your ERP, billing, banks, even email and slack. No rip and replace. Maximor runs on self-learning agents that learn how your team already works. No prompts to wire, no workflows to build, nothing to document. Every output is audit ready and when a call needs human judgment, it escalates for review because the agents optimize, not just execute, and your KPIs keep on improving. One PE-backed customer posted 98 % of transactions straight to their ERP, no error rate.

21:16The other 2 % go to a human before posting. You pay only for real outcomes, not seats. See it at maximor.ai. That is M-A-X-I-M-O-R dot A-I. Listen, I'm no mathematician, although that is what my mother-in-law tells people I do for work. My son-in-law is CJ, he's a mathematician. If your finance team expands linearly with your headcount, your leverage is broken. Too many CFOs waste incredible talent on low-value maintenance, manual reconciliation, chasing missing receipts, and policing$30 software subscriptions. It's the expense police, run! That's a systems problem. And that's why you need Brex.

21:57Brex is an intelligent finance platform powered by AI workflows that handles the heavy lifting automatically, which is a fancy word for without you. Instead of burning elite finance minds on administrative maintenance, Brex gives your team the leverage to focus on momentum and growth. Thousands of companies, including Anthropic, Coinbase, and DoorDash, already run on Brex. Stop asking your A-level finance talent to do B-level admin work. It's time to get Brex. Head to brex.com slash metrics. That is brex.com slash metrics. Today's episode is brought to you by Anrock, the sales tax platform behind companies like Anthropic, Notion, and Vanta.

22:36Here's a fun way to totally ruin a Tuesday. Open a letter from a state you've never set foot in telling you that you owe back taxes you didn't know existed. Happen to me. Because the rules never stop moving. States are now racing to tax AI, digital ads, streaming, really anything new. and they're doing it faster than a spreadsheet can keep up. Anarok handles all of it. One platform that watches your exposure everywhere automates compliance and flags risk before it turns into that nasty gram. That's why thousands of finance leaders trust Anarok to stay ahead. Talk to an Anarok sales tax expert for a personalized exposure estimate at anarok.com slash RTN.

23:16That is A-N-R-O-K dot com slash RTN. When you look across your finance team today, where are the most surprising AI use cases emerging? I don't know about surprising, but I'll tell you the fun ones. Like the ones I've just been like the most impressed by. So we get physical mail because, you know, some folks still like to send you, and often some of that mail is really important. I got a check in the mail last week, blew my mind. But, you know, often it's like a notice that like someone needs to react to, you know, or sort of key teams on the receiving end. And they often like are pretty timely.

23:45It's like, hey, respond in 30 days, 60 days, et cetera. And we have somebody internally whose role it is. It's like take the mail, open the mail, read it and like a lot of this is like jargon it's maybe to a department or maybe it's just the company and they have to figure out how do i get it to the right person the folks that were usually on the receiving end of like this physical mail and like found themselves in these sort of forward chains a couple of them just decided like hey like we're pretty sure like ai can do better at guessing at who this piece of mail is really directed at and so they built a tool that once you scanned it it goes into the sorter and like it has really high fidelity now as to like being like, yeah, this is for this person, this is for that person, even like mail that we've never seen before.

24:24And it's just like massive efficiencies for like the individual whose job it was to like try and like take these pieces of mail and like match them to one of the 1100 plaids. Other things I've seen the team do, amazing like step forwards on accuracy. If I think about like our forecasting and scenario planning, just like leaning into like more mathematical methods that would have been really challenging to do, whether it's in Google Sheets or Excel, like even figure out like the right formula and the right way to like run an experiment, just being able to now like rely on these really powerful tools.

24:55I've been really impressed by that. Okay. The mail example is the coolest AI use case we've had on the podcast yet. I'm imagining a robot sorting all this, scanning it and saying it's the right person. But also the finance department sometimes can become a catch-all for when they don't know who a notice comes to. So it's also solving a problem that selfishly was going to end up on your desk eventually anyway. Funny how that works. You have AI at your disposal. Your team does too. It's being adopted. Do you think the way that you've asked questions or the questions you do ask have changed at all?

25:28Usually it's easy to focus on like, okay, what's the result? So, you know, is this the right answer? I find myself asking a lot of questions now around methodology. How did you get here? What did you try? Especially when we're leaning into AI tools, for example, coming back to that forecasting methodology. Like normally when a team gets the answer, you can send me the spreadsheet. I'll figure out like what you did. But now that we're leaning into AI, Like, I'm very curious to see like what the team tried, what tried and what didn't work, as well as what tried and they think works and why do they think it works.

25:56You know, you ask the simple language like questions. It can a little bit feel like a black box and me being sort of the math nerd I am. I sort of I want to know the logic that sits underneath, first of all, to make sure it's solid, but also to figure out like what we can learn from it. And so I find myself with the team asking a lot more about like, which methods did we use? And why do we think it's the right method to fit like the problem that we're trying to solve or the analysis that we're trying to execute? And then I usually now be like, and where to next? Right? Because like, we're still so early on this journey.

26:23I asked the team a lot more like, okay, like, what's your next step here? Like, what's where would you take this for the next time around that we do this for the next iteration? So I find myself sort of a lot more in not just the result, but in the process and the thinking underpinning it. This is amazing. So just to play it back for listeners, methodology, and then also asking what the next step is and to dig into the methodology part. I think it's so key because AI can package something up to sound so convincing. But if you go just like a little bit deeper, you say, oh, it's kind of empty inside.

26:53There's not a lot of there there. And so you have to prove that the person took the time to do the work to ultimately understand it. And that this is an accelerant. It's not just glossing over and getting to a quick answer. Yeah. I mean, what I want to see that AI is doing for the team is it's actually helping us think deeper, right? It's not like it's taking the thinking away from us. I think that'd actually be failure mode. Can you say more about that? Being more ambitious? I think that's a great thought. Yeah. I mean, AI gives you leverage, right? Like in its best form, it gives you leverage.

27:22And so like most people that I encounter and certainly I know a lot of the financing, your time constraints, your bandwidth limited. And so now having like a smart co-pilot to help like essentially it feels like more hands working on the same problem space. You know, it helps you be more ambitious in terms of like what you choose to say. Like I see it with myself. I now know that something that took me an hour before I can probably do in 15 minutes. So like I find myself taking on more and more because like not only can I get it done faster, it just feels more fun. Shane, tell me about this AI day you've been doing.

27:55This is a pretty cool concept. Yes, actually, today is Finance AI Day. Oh, it's today? Really? AI Day is a day that we dedicate to building, like with AI tools. One of the key challenges I think we're experiencing with like leaning into AI is like carving out the time to build. And often you need like more than just the 15 minutes here or the half an hour there. Like for me, it's akin to the math problem sets like I had to go through in college where like you knew you had like a really gnarly problem set and it was going to take a while. And like you needed to go deep to really sort of get to the answer.

28:29It wasn't like a quick check the box exercise. And so I would usually show up at the library at like 9 p.m. at night. And no, I was there going to be there all night. But for me, that was like the most efficient way to like really sort of get into a certain mode of thinking once you're in that mode. Sort of all the pieces start to come together really quickly. Our AI day is happening during the day. Don't worry, not overnight. But it's sort of like allowing people to step away from their day to day to switch their way of thinking to truly being like builders. and then being able to go really deep on a problem space and build, not just like figure out how to build, but then actually to dedicate time building.

29:03What it also enables us to do by coordinating IAI days is it allows for a little bit more like cross team collaboration. You were to take that time out, it might work for your schedule one way and not the person you want to work with another way. So it's sort of like eases cross collaboration because everybody's sort of taking the time out of their schedule at the same day of the week. And we also like usually kick off with some kind of training again. So like continuing to sort of like invest back into the team. And then sometimes we wrap with demos just to see what people have built. Is this the first one you've done or have there been multiple in the past?

29:34So we've done Plaid AI Days. And so we're taking inspiration from them. I mean, we're fortunate. I work at a tech company with a ton of engineers who are like the best builders in the world. And imagine they're the tip of the spear in terms of like using these tools and figure out how to build quickly and build efficiently and really sort of like change the way of operating. So be real with us. Do you think this is more just a get people warmed up to using AI? Are you hoping like workflow comes out of this? We're on a four month journey that I set out for the team. So we're like in month two, we did like a survey from the team asking everyone bottoms up, like what are the ambitious projects you want to work on?

30:08But beyond what you're working yourself day to day, like if you could recruit a couple of hands to work on a project and then as a result, based on people's interests, skill set, we created teams across finance to like now go and work on these projects. And the idea is these projects are going going to take about eight weeks to build out. So not everyone will get to like a usable functioning product. Like some things just are going to get stalled or aren't going to work. But the idea is that people can then work together, chipping away, like taking on more ambitious projects. And then we'll sort of like get to demo day.

30:36And then once we're at demo day, we figure out like what's ready to go live. Yeah. And I've participated in a couple of Vibe code sessions. And don't get me wrong, it's really cool to just be confident that you can at least get to something that's usable, maybe not production ready for a whole company to use. But what you're saying is this is an experimentation phase that allows you to have a timeline where you work towards something and eventually the team has something that they can, you know, leverage day to day. That for me is like the end goal here is that like this isn't it's not just a fun tool, but it's like a tool that generally helps us like improve our accuracy, improve our speed, improve our leverage as a team.

31:13I love being in high growth businesses because even with a finance function, you do get to be a builder. And I think that's never been more true now that like we have these AI tools available to us as well. I'm curious if AI has changed it all the way that you personally lead your team. What I found is it's created like a greater connection with the team. So rather than purely leading from like in front, it's allowed me to come alongside the team because we are, I know more of an expert in AI than anybody on my team. This is such a good point. Yeah, but if you're a finance leader, you came from a background where you have seen stuff that they haven't seen a lot of it before, spent more hours than anyone would care to hear about in Excel.

31:52And like a really great way it puts you on the same playing field. It like allows you to sort of geek out with the team in the same way. I've loved that in terms of just like building deeper connections within the team and allows other people in the team to step up and to lead. Like I don't have to occupy that leadership, which is you can kind of sort of like create more space for leaders on the team and people who are even a lot more junior in their career, like they get to step out and lead the team and like things that they're building out. As much as I love leading from in front, it is really cool to get, you know, sort of to build that empathy with your team.

32:22Let them teach you something. What skill sets have become more or less important now? I think more important is definitely discernment and quality review. Failure mode in AI is like increasing volume without really increasing impact. For me, it needs to be like net better than what any human on my team could do. We should be raising our bar by having like these powerful tools. And then the other thing is actually operational anticipation. It's when you unlock velocity in one part, not just within finance, but actually across the company, when you unlock velocity somewhere, the bottleneck then moves, right?

32:53Velocity happening over here, but if the person downstream of that isn't also unlocking velocity or capacity at the same rate, like the pinch point will move. And so sort of anticipating those pinch points I found across the business is where it's also like shifted. What skill set I think is sort of becoming, you know, has come forward a bit more since starting to build with AI. In terms of what's become less, trawling through large quantities of data, material I need to review is significant. I can read a 10K pretty quickly though. You know where to look, but just like multiple 10Ks, your quarter end earnings or like all the emails that you receive, all the slacks you receive, being able to really quickly like context switch and not miss the detail.

33:31Like a lot of my work is like it's in the detail. Now being able to rely on these tools that can just take large swaths of information and like gain the context. So I'd say like, whereas before I would have said that was like a superpower of mine. That's a superpower, frankly, everybody has. More is not always more. It has to be substantially better than what a human was doing previously. I think a lot of people confuse activity metrics or volume of output with it being better. It's kind of like if Drake just went and dropped three albums, two of them weren't that good. maybe he should have just concentrated even more on the first one.

34:04I'm going to use that one on my team now. It should improve quality. Like I very much believe that. I remember like learning to write SQL queries. You know what a good model looks like and you know what a bad model looks like. They could probably still get you both to the same answer. But if there's like way too many self-referencing cells, you're like sort of wandering through multiple tabs, trying to follow the logic before eventually ending up versus something that in 10 lines get you to the same place. So quality does matter. Sort of having AI embedded at every level of the organization is really important.

34:33I know sort of in speaking with some of my peers, AI may like really shift the like shape of your team. And I don't know I see it that way just yet, because I think the context is important for the quality. Folks with tenure who have seen it before, I think, can be pretty discerning on quality. And so I think, like, super important for everyone across the org to keep building and not just think it's something more junior or less tenured team members to do. I actually think it's really important that your most tenured team members in finance are also building because those folks tend to have the highest judgment level.

35:05Is this right? Is this wrong? Is this high quality? Is this low quality? Making sure that actually you're combining that with AI. For me, like, that's where the magic happens. I think it's also so important to be able to see whatever you're working on with AI, how it impacts the other people around you. Well, you solve one choke point or make one part of the finance stack go faster. Well, what's downstream from that? And it's incumbent upon the person, I think, who's working upstream to say, well, did I actually just create a bigger problem for someone like in billing or in collections down the road?

35:36Yeah, and that's OK. That's freaking cool for me. It's like if you're pushing everybody to go faster, that's a good problem to have. but it is also if we truly want and maybe getting back to the idea in this tops down building when you're in your bottoms up like that would probably happen you fix your thing and that just gets pushed downstream and then you're taking a step back and you're wondering like why isn't the organization moving faster and likely because you've just pushed the bottleneck elsewhere building in a bit more of a coordinated fashion you're able to solve all those bottlenecks in a more of an organized way and so that hopefully like my my hope is that it will lead to sort of what feels like more step change improvements in like our efficiency or in like really automating or changing the way we operate as opposed to like purely incremental.

36:20And I think this is a great point to transition a bit and talk about planning because you've set a framework that you use is objective strategy tactics. Can you walk me through this one? So I use it across everything, not just in planning, like starting with objectives. So objectives in sort of plain speak would be like winning is dot dot. So when you think about what your objective in any situation, high stakes, low stakes. What is success going to look like to me? And then the strategy, it's like, okay, so like how am I going to, what are the ways in which I could win? And finally you get to the tactics.

36:51It's okay. So what's step one? Step two. To pick a non, completely non-financial example, if I'm like, hey, I want to build like the best bakery my town's ever seen. That's my objective. Winning is I have like the best bakery this town has ever seen. And then the strategy, okay, I'm going to win by like birthday cakes. I'm going to choose to focus on birthday cakes. You know, on the birthday vertical. I've got two young kids and let me tell you that birthday vertical is a lucrative vertical. There's a lot of money there. And then tactics is like, okay, step one, hire a baker. Step two, like figure out my recipes.

37:21And so that's a pretty simplistic way of thinking about it. But I think you can blow that up into like company level planning and you can shrink that down to like, hey, I'm about to go have a conversation, like a pretty critical conversation. Let me think about what success really looks like. What's my objective? What do I want out of the conversation? given that like what are the ways in which i could approach this okay given that like what's step one what's step two objective strategy tactics which of those three gets skipped or not given enough airtime i would actually say it's probably defining what winning is it's like loosely defined but for this to really work you actually need to define a lot more specifically and so if you have two people being like i want to build the best company ever cool you need to get way more specific than that so it's like not just one getting stiff but i think like one being like too low fidelity and so if that's too low fidelity, you're going to diverge when it comes to the strategy and you're really going to diverge when it comes to the tactics.

38:12Well, I find some companies are so gung-ho on just doing something and jumping to the tactics that they won't even step back with like a new product launch or geo that they're going into and say, is that even something that we want to win? Is that even a game we want to be playing? Would we be happy if we won that prize? Can we doesn't mean should we. To hit on the planning vein, and I love that that can be used, that framework can be used even for preparing for a podcast or a meeting that you're going into. You've been through a number of planning cycles and you've remarked that planning reveals things about an organization.

38:42What does a planning process in going through that reveal that you may not just see day to day? Whether strong alignment, weak alignment or no alignment at all. I love the planning cycle or the planning process for that reason. Because, again, if I go back to like my framework, it's really hard to make progress if you're not aligned on like what success looks like. people are talented and they're ambitious. But if you aren't aligned on what success looks like, you're not having people, if I think about it almost from a capital allocation framework, we're all deploying our capital in very different ways.

39:15And so a lot of capital have been deployed, which means like you should see a lot of return. But if you're not all deploying it into the same areas, all of that energy and activity is just going to bleed out. And so I think really like alignment is thing number one. And thing number two is prioritization. Once you have that alignment, how does that filter into prioritization? I can't say I've got the prioritization thing figured out. It's a perennial challenge. Like when you have like a really big opportunity in front of you, it's like it's all important. That's the truth. But then thinking about like what's the right sequencing?

39:43So it's not that not everything is important, but how do you think about like the optimal sequencing within that to truly enable folks to make really like rapid and meaningful progress? And sometimes it's not saying no, it's just saying not now. Exactly. You've also said that planning starts by figuring out which team goes first. I think this is like the age old question. It's like that Spider-Man meme. They're both pointing at each other. One is sales, one is product. And I guess it'd be a third one that's finance. How do you decide whether product or go to market or someone else should should lead the planning process?

40:12I've tried planning like all at the same time and it gets messy. You don't want planning to go on forever at any company, right? And so you're like, OK, we're all just going to pick like this short sprint and we're all going a planet once. It was like very confusing. And so I was like, I now like believe very much like sequential planning. And so when you figure out who goes first, how do we maximize enterprise value? This is like long range planning, less about annual, but like more about long range planning. What's going to like maximize enterprise value over the long term. And at almost everywhere I've built with, it starts with what we're building and who we're building for.

40:43I tend to operate in high growth environments where it's less about like, you're not in your harvesting era. It's like, well, this is what we do. And we just need to figure out like how to do it like incrementally better. It's more that like you're still pushing out the possibilities of what you're building, who you're building for are very much still on the upwards trajectory. It's oscillated. Stripe and plaid product tends to be the ones to define like this is what we're building and you know with support from GTM this is who it's for. At Glossier actually it was it's marketing, it's GTM. Depending on if it's an engineering-led org versus a sales-led org there could be a different voice It's kicking it off and kind of guiding it.

41:20You need both at the table. Product and your GTM. So it's always like, I don't know, you know, some people feel like, oh, if I go first, it means something. No, not really. But the idea is just trying to like really already trying to solve for that alignment. It's like if you don't have alignment around what you're building and who you're building for, that's going to be a really, really painful planning process. Well, let's hit on who you've been building for, because one thing that I think is super fascinating about Plaid is the network gets smarter as it grows. So it becomes more valuable as it grows.

41:45How does a CFO think about measuring the value of a network? I think about the value of a network around like three dimensions. The first is scale, like scale in absolute numbers, but also scale in like percentage covering. We connect into 12 ,000 financial institutions. I want to know not just like the 12 ,000 number sounds like a big number, but what coverage does that give me? How many potential consumers does that cover? Then the second thing, and I think this is like truly a business that wants to build network effects, is the quality of that data. It's the depth of the data. It's the speed of the data.

42:17And it's the intelligence of like the network that you're building. So depth kind of going back for plaid into that like identity network. If I had like, you know, one connection to CJ, like I'd like one basic connection to CJ. Sure, it gives me like scale and coverage. But the thing that will really enable me to build like a really high quality network is having more views on CJ's life. Because like the more I know about you, the better we can serve you and the better our customers can serve you. So really thinking about like that depth and quality. And then also like the speed of the connections, given like financial transactions, you want them to happen as seamlessly as possible.

42:48And for some of the products that we build, credit underwriting, anti-fraud decisioning, real-time data is really important. I call them quality metrics, depth, speed, intelligence. And then the third is usage. At the end of the day, are people using the network? Because if I built like, you know, this massive network with amazing quality and nobody's using it, something's probably gone wrong along the way. I'd say that's probably a little bit downstream. But again, if you're seeing like all of this great scale and quality and it's not showing up on usage, probably time to take a beat and figure out what's going on.

43:18Well, it sounds like there are leading indicators that'll tell you the network is getting stronger before revenue shows up. Yeah, there are definitely. I think the quality is really important in figuring out sort of I'd say like that for me is a bit of the precursor for like, is revenue going to show up? Is this value added at the end of the day? Is there alternatives out there for like some of the point solutions that we offer? And so really for me, thinking about the quality, that's going to be the precursor to revenue. When you open your computer each morning, what's the first metric you look at?

43:47Usage. Usage, okay. We're a usage-driven business. And so usage for me, that is like the leading indicator. A lot of good things are going to happen if usage is high, right? You don't have to wait for it to show up on the income statement. Exactly. You've been in a lot of amazing companies with well-known brand names. What do you think goes into building a durable, independent company? Reflecting on your career, what does durability mean to you? I thought of durability almost like defensively. Withstand and survive. But I don't think that's playing to win and playing to survive. I actually think, you know, as I spent, you know, more and more and more time in sort of some of these high growth, groundbreaking companies, it's been about offense.

44:27It's not just surviving. It's about thriving, regardless of the macroeconomic situation. And also, like, how do you thrive not just when you have tailwinds? Because everyone can do pretty well when you're at tailwinds. But how do you build a product that will outperform and a customer base that will outperform even when you have headwinds? You should ride that wave. Make sure you do really well. But actually, when there are headwinds, how do you make sure you're still advancing the business in light of those headwinds? And what that does for me is it raises the bar on you. Most people are going to do well when things are good.

44:56When things turn, that's when, like, you know, you sort of start to separate, as they say, the cream from the milk. It keeps a really high bar on how you run your organization, that you're not just running your organization to thrive in the good times, but you're running a resilient organization. You're building a resilient team. One more thought on durability. Increasingly, I also realize agility is super important. In the finance seat, I think it's two things. Really high visibility, because you can't move and react to what you can't see. And discipline, actually. When you're really disciplined, when you know where everything is, you're able to move, reform the business very quickly.

45:32Each day, each month, we live in the most interesting times. And that agility, I think, is really key when you're building an organization because you're not going to be able to anticipate every single macroeconomic speed bump or shock. You're not. But what's really important is that you have an org that has the shock absorbers and that has really good visibility that you can sort of like reform and react to movements in the market without sort of missing a beat. So me and you are doing part two of this podcast 10 years from now. Plaid is a wild smashing success, of course. What will people have underestimated about how that success happened?

46:08I think it's probably still going to be the things they underestimate about what we felt today. I don't think it'll be the tech component. I think Plaid is already known for being like a really innovative business with an exceedingly high bar on tech talent. And I think what folks underestimate about what we've built and what we'll continue to build is the importance of the trusted relationships that we've developed across the industry and the ecosystem. You sit across the company from some of like the largest financial institutions in the world. They want to trust the people that they are partnering with.

46:40It's so easy for it to get overlooked when it just shows up as like a pain on your phone. may also trust our deep relationships and expertise when it comes to security, privacy. All of these things are like their table stakes, but they're also like essential to our business. And frankly, the set of values that guides us to continue to prioritize what is best for the consumer and what's best for the ecosystem. Financial services for me is like this living, breathing ecosystem with so many stakeholders. And it's sort of one that needs to be tended to. I think if you ever over-rotate, one stakeholder starts to over-prioritize their own ends.

47:14It is a value destructive thing. And so I think like what I've been really excited about and proud about at Plaid is how much we think about like the ecosystem and the health of the rural ecosystem. Person number one within that is the consumer and like trying to build a business where other people win. If everybody's winning, I think things are a lot better off. If the consumer is winning, I think then like then it's working. I'm going to take you into what we call our long ass lightning around. So the first question I ask every successful person on the podcast is give me something you've messed up on the job before.

47:47It could be this role or any other. Oh my gosh, I've really enjoyed actually listening to other people's stories. There are many mistakes along the way, but I'll take you back to sort of my investment banking analyst days. I was working on a particular client meeting, very excited about it. I was at a boutique bank, super lean, so you got to work like really closely with like, you know, all the senior members of the team. and we had a client meeting 9 a.m. the next day. The deck wasn't finalized, I think, until the wee hours of the morning, let's say. But we did a lot of our printing of the books.

48:17You'd need to print the books and bind the books and you'd get them like really professional to share with the client. Took ages for the deck to get finalized. And so we were like really up against it with like printing and binding. And we had a team in-house that would like, you know, you'd ship it to the printer, but it was a person sort of at the other corner of the office and they would print everything out, bind the books and make it look great. We were basically past the deadline. I kept on like messaging and be like, please, just like five more minutes, five more minutes. Don't go home.

48:39Because if you go home, like it's game over. This meeting can't happen. So they kept staying. We finally locked it in. I felt really bad because at this point it's like two in the morning and I'd kept this team like way beyond when they were supposed to be there. So I was like, you know what, guys, I can jump in. I can help. I'm a handy person. You guys print them. I'll help bind the books. So fast forward, they're printing. You know, it's like a smooth operation. You know, like they're printing, handing off the pages to me. I'm like punching the holes and binding the books. And then at the end, you know, you take it and you flip the deck to make sure everything is as expected.

49:08And I'd like somehow reverse the order of the pages and they were not. Putting in order. You just heard everyone because we thought we were like minutes away from going home. At this point, you're pushing 3 a.m. We're going to have to do this all over again. So there I was thinking I was saving folks, you know, maybe 20, 30 minutes. But that's critical between 2 and 3 a.m. And instead, I just set us back by an hour. It took a while to live that one down internally. That's just saying. No matter how smart you are, printing can still be difficult. It gets you. Yeah, exactly. It's like not just the book smarts, you need the street smarts too.

49:39Yeah. Someday I'm going to write a children's book called The Printer That Just Worked and other fairy tales. Next one I got for you. If you could give your younger self advice knowing what you know today, what would you tell her? Ask way more questions. Professional services, but actually in particular in banking, there was this illusion that you had to have all the answers. A partner who spoke so like eloquently and clearly and forcefully and you just thought that like that's what success looked like. Or certainly I thought that's what success looked like. You had to have all the answers. It's like you're at the start of your career.

50:07I don't know where I got the impression. Like you're not expected to have all the answers. Ask the questions. Like you're there because you're smart. Ask questions. And if you don't understand the answer, it's not because you're not smart. It's just because it hasn't been well explained to you. Ask again. You know, my four-year-old is amazing just asking why. Why? No matter what you tell her, she'll ask why after. I need more of that in my own life, I think. Yeah, I've got a four-and-a-half-year-old and a two-year-old. The two-year-old less so at this point. of what you will ask otherwise. Next one I got for you, more of a technical one.

50:34What tools does your finance team use to get the job done today? The usual suspects, we are on NetSuite. Billing and AR side, we recently moved to Paystand, which the team has been really excited about. It sort of helped us like consolidate sort of like the multiple types of customers that sort of use us and that we bill and invoice. And that was like a remarkably smooth migration. So huge shout out to the team. We, for corporate cards, use one of your sponsors. And then like trying to figure out like what do we want to build ourselves now? We're sort of pushing on that a bit now. Like one of the tools one of our team members built with one of the cross-functional partners, we have a third party like software that we use to do order form mapping back, you know, out of Salesforce and sort of out of Salesforce and back into Salesforce, like once the deal is closed.

51:18And the team built their own version over one of our AI days that actually wasn't just read, but also had write control. There are definitely pockets where I'm like, I think we can build something that just better fits our needs. So excited for that as well. that's such a tangible use case and i know exactly what you're talking about with the contract and it's this gray area where you can go back you may fat finger it in but it's also going to take time the ability to write that back just so you have consistency like amazing use case very last one i got for you what's the craziest thing you've ever had someone try to expense i worked in consumer for a while and that was definitely a shock to my system these were legitimate expenses but i just had coming out of tech i had no idea how much photo shoots cost You've got hair, you've got makeup, you've got manicures, you've got like what they're wearing.

52:02And so I would see these invoices come back. They were on budget and well within their rights, but there were definitely moments where like my head was exploding. No$20 haircuts there. I'm afraid not. Thank you so much for joining me on the podcast. This was a bucket list interview and I really appreciate you coming on. Thank you so much for having me, CJ. Run the Numbers is a mostly media production. Yelling an intro by Fat Joe. Artwork by Meg D 'Alessandro. The show is executive produced by Ben Hillman. Nothing said on this podcast is intended to be business or investment advice. It's the sole opinion of me, a guy who feeds his dog way too much ice cream and has a history of net operating losses, LOL.

52:40If you like this podcast, hit subscribe and give us five stars. It will take like two seconds and our algorithm overlords love it. Drink water, call your mom and have a great day. Peace.