Keith Rabois & Eric Glyman: The Tools, Tips, Secrets and Process That Drive Efficiency | E1148

1. 0:00 – 0:54

   Intro

   1. KRKeith Rabois0:00

      Second framework that Jack Dorsey taught me is editing, writing metaphor. You know when you're editing or when you're actually writing. And you want to be editing people's work as the CEO, you don't want to be writing. And if you feel like you're using a lot of red lines consistently in the same area of a company, that's a really bad signal. And if you're simplifying, that could be a good thing. But if you're constantly asking clarifying questions and not really getting, you know, to the root, that's a problem.

   2. HSHarry Stebbings0:25

      Ready to go? (instrumental music) Guys, I am so excited for this. You are two of my favorite people. So first off, thank you so much for joining me today.

   3. EGEric Glyman0:45

      Thanks for having us, Harry. It's awesome to be back.

   4. HSHarry Stebbings0:48

      It's so nice, you being in the same room as well, it's so much better for the chemistry. I wanna start off with a little bit of the context.

2. 0:54 – 3:09

   Founders Fund & Khosla Deals

   1. HSHarry Stebbings0:54

      How did you guys first come to meet? And how did the first deal come to be with Founders Fund and then the Khosla deal?

   2. KRKeith Rabois1:00

      Well, actually, uh, Eric's co-founder, Karim was playing video games with one of my colleagues at Founders Fund, and we had been looking for something interesting in the space. We thought there's a great opportunity to reinvent finance and the finance of companies. And Delian s- walked into my office one day and said, "I think I found your founders."

   3. EGEric Glyman1:18

      (laughs)

   4. KRKeith Rabois1:18

      (laughs) I was like, "Okay."

   5. HSHarry Stebbings1:22

      How was that first meeting? Take me to it.

   6. KRKeith Rabois1:24

      It was actually pretty amazing. So they actually flew out, uh, to SF, like, almost like the next day.

   7. HSHarry Stebbings1:29

      Yeah.

   8. KRKeith Rabois1:30

      And presented off notes. So it was really like, o- on the fly. But three minutes in, I was like staring at Delian because I was like, "Oh my God, he's actually right. He found the perfect founders." Because as Eric was walking through the vision and how they were gonna approach the market, it was dead on target of everything we believed. So, like those notes actually still exist.

   9. HSHarry Stebbings1:50

      Yeah.

   10. KRKeith Rabois1:50

       I've actually, um, I think you had an intern taking notes.

   11. HSHarry Stebbings1:53

       Mm-hmm.

   12. KRKeith Rabois1:53

       I've actually seen the notes from the meeting.

   13. HSHarry Stebbings1:55

       Yeah, yeah.

   14. KRKeith Rabois1:55

       And if you read the notes today, it is literally just like the board meeting we just finished. With the exception of AI. We didn't really talk about AI.

   15. HSHarry Stebbings2:01

       It's true.

   16. KRKeith Rabois2:02

       But everything else in the board meeting today is like right on target from those first notes in maybe May of 2019.

   17. HSHarry Stebbings2:08

       Keith, you don't know this. I had drinks with Eric in New York just after Paribus, before he started Ramp. And all I can think is, "Fuck, I should have asked if he was doing something great next." (laughs)

   18. KRKeith Rabois2:21

       (laughs)

   19. HSHarry Stebbings2:22

       But that's on me. How did the latest deal come to be with Khosla?

   20. KRKeith Rabois2:27

       Well, for a variety of reasons, um, we were interested, um, at, uh, Khosla Ventures in Ramp. I mean, obviously it's a high profile company, but my involvement, you know, definitely piqued people's interest, but because of the possible intersection of AI in finance, my colleagues are very interested in AI. As you know, Vinod is like a pioneer in AI thinking. They understood immediately why Ramp had a lot of the secret sauce to be successful in using AI in finance, and that nobody else on the planet does. And so they were really excited to meet with actually Eric and Karim, almost independent of investing. It was more just like, "This is the company that's gonna be successful in AI as the future of finance."

   21. EGEric Glyman3:08

       Yeah.

   22. HSHarry Stebbings3:08

       I, I'm too interested there.

3. 3:09 – 5:28

   Ramp's Secret Ingredients for Future Finance

   1. HSHarry Stebbings3:09

      Why does Ramp have the secrets of ingredients to be successful as the future of finance?

   2. EGEric Glyman3:13

      So, I, I think the first thing that's misunderstood about Ramp is that, uh, it's a money movement and a fintech company. I think one of the deceiving kind of Trojan horse ways in about Ramp is that, you know, we're a corporate card and, and we are the fastest growing corporate card in America, but we're actually a productivity company, uh, and a workflow company. People use Ramp, uh, not to buy things, but to automate expenses, um, you know, not to have the tools to close your books, but to have it be automated. And so really what's going on, uh, inside of, of the strategy is trying to understand all the ways that companies are spending money and time, and surface up data to show companies where they can spend less. Maybe they're buying 200 seats on Asana, but only 100 people are logging in, and we can detect that based on kind of Okta data and show people ways to cut. Maybe, uh, companies are, um, you know, have 20,000 transactions. With Ramp, we're able to suggest accounting categories, um, certainly more quickly, but also more accurately than the vast majority of customers. Uh, and when you think about really where, where AI is excellent and has unique capabilities, it's really around, uh, a few things. Um, you know, doing work that previously was knowledge work, it required context, required vast amount of data and in order to form an output. Um, and because we connect to so many data sources, because we automate so many functions, um, it's much easier to kind of see how it can get applied already today, um, for Ramp customers and ladder up to kind of the simple ways people think about Ramp, which is you save 5% on your card expenses, your books are closed, uh, you know, much more quickly. And as Ramp is touching, not just cards, but bill payments, but procurement, um, you know, and more aspects of how companies move money and spend time, um, really the benefits of, of AI, um, you know, it's not just to write better poems or, or make great art, but it can actually automate all of these workflow processes. And so that's what's, um, was interesting. And, you know, I think beyond it, um, you know, my, part of my rationale is, you know, Keith and, and Ramp are, uh, I, I think... I, I can't really remember, um, the time when, when (laughs) we weren't working together. I think for 97, 98% of the days we've existed, we've worked together.

   3. HSHarry Stebbings5:16

      (laughs)

   4. EGEric Glyman5:16

      And so to re-formalize, to, um, get the Node and the team's great advice, um, and to invest not just in delivering more value, but accelerating the business is, is what made us very excited. So it's been a very fun month.

4. 5:28 – 7:51

   Eric's Strengths & Weaknesses

   1. EGEric Glyman5:28

   2. HSHarry Stebbings5:28

      I mean, speaking of that partnership, I wanted to see like an all-star partnership quiz with you both. You didn't get the schedule beforehand, so I had fun with this one.

   3. EGEric Glyman5:35

      Cool.

   4. HSHarry Stebbings5:35

      Keith, I wanted to start with you. What are the one, two elements that you think make Eric truly world class that maybe not enough people see?

   5. KRKeith Rabois5:43

      So the focus from the very beginning was on talent. Um, we spent literally the first board meeting probably reviewing LinkedIn profiles for 40%.

   6. HSHarry Stebbings5:50

      Yeah.

   7. KRKeith Rabois5:51

      So the team... The Node expression that I borrow all the time, "The team you build is the company you build," took very seriously from day one. And it shows. And now that there's 754 employees or something like that-

   8. HSHarry Stebbings6:00

      Yeah.

   9. KRKeith Rabois6:00

      ... it's just compounded. So the talent advantage has amounted. Secondly, um, cutting through an initial aperture in a market that was perceived as crowded. Eric's marketing instincts are first rate, and I think if the CEO doesn't have first rate marketing instincts, you kind of never get there. And so now the company's doing things that other people don't do. We're gonna do things that the rest of the world can't do. But in the beginning, how to cut through the clutter, and the CEO has to be able to do that, and Eric's phenomenal at it.That's actually-

   10. HSHarry Stebbings6:28

       All right.

   11. KRKeith Rabois6:28

       ... why we preempted the series A, was these two dimensions, actually. It's very tangible. The second board meeting I think we had, that was in th- their original New York office, the emphasis on talent and the emphasis on being able to frame the value proposition in a compelling, powerful way, and one that was strategic, was shockingly great. And, uh, actually Delian prompted me and said, "Do you think we should preempt the series A?" And the company hadn't really even launched. Basically, I think we had $30,000 of spend maybe at the time.

   12. HSHarry Stebbings6:53

       I think that month, yeah. (laughs)

   13. KRKeith Rabois6:54

       Yeah. So $30,000 of spend, uh, but we preempted the A because it was so obvious that Eric and Karim really collectively together had those traits, um, that this was the company we really wanted to double down on right away.

   14. HSHarry Stebbings7:06

       All right, you feel pretty great hearing that. Just pause for a s- just-

   15. EGEric Glyman7:08

       I'm- I'm- I'm blushing. You- you can't do that to me.

   16. KRKeith Rabois7:10

       It was a pa- (laughs)

   17. HSHarry Stebbings7:11

       Pause for a second here. Keith, on the flip side, you've known Eric for a long time. What do you think is one of Eric's biggest weaknesses that he could maybe work on?

   18. KRKeith Rabois7:20

       (laughs)

   19. EGEric Glyman7:20

       (laughs)

   20. KRKeith Rabois7:21

       Um, all found- all found... No- no founder's perfect, but, um, I- I think the company's results (laughs) kind of speak for themselves. I think we just finished a board meeting where it's pretty clear that this is the best-run private company on the planet. Um, so I wouldn't (laughs) really critique him too much.

   21. HSHarry Stebbings7:37

       (laughs)

   22. KRKeith Rabois7:37

       (laughs)

   23. EGEric Glyman7:37

       I can take it though. P- please, yeah.

   24. HSHarry Stebbings7:41

       Okay, I'll- I'll- I'll take that. Keith, well, well done. Very, very politically answered. Eric, if I were to apply-

   25. KRKeith Rabois7:45

       It was actually accurate though. (laughs)

   26. HSHarry Stebbings7:46

       If I were to apply the same to you, Eric, and you know Keith so well now for so

5. 7:51 – 9:10

   Keith’s World Class Brilliance

   1. HSHarry Stebbings7:51

      many years, what do you think makes Keith the world-class partner that he is?

   2. EGEric Glyman7:55

      I mean, it- it's- it's a few things. I mean, first, one of the things that people talk about Ramp at being great at is just the efficiency and focus of our operations. Uh, it's very fast-moving. People know it's fast-growing. Um, and I- I think it started a lot with some of the lessons we internalized from- from Keith. He talks about this concept of can you really articulate cleanly your business equations? Um, what are the outputs you're trying to drive? What are the inputs in your business? Can you list that very clearly and map that? And that exercise of mapping how does our business really work created focus for our company. And that's something I- I recommend to every founder. It- it certainly changed my ability to operate and- and- and helped, you know, Ramp reach the- the scale that we're doing quickly is- is this talk he gave a decade ago called How to Operate. It's- it's on YouTube. It's free. Um, it is worth a lot more than that. (laughs)

   3. KRKeith Rabois8:45

      (laughs)

   4. EGEric Glyman8:45

      Um, and- and I recommend everyone watch it. And- and- and I- I think that- that his focus on, um, you know, clarity of thought, excellence in operations, central casting, finding extraordinary talent, uh, at the end of the day, it's... All a company is, is it's a collection of people. Um, and if you, if you hire great people, you give clear focus, and you just execute vigorously, (laughs) um, you know, it's not much more complicated

6. 9:10 – 12:13

   Eric’s Inputs & Outputs

   1. EGEric Glyman9:10

      th- than that.

   2. HSHarry Stebbings9:10

      Eric, how do you think about your inputs and outputs today? What are they, if you were to output them?

   3. EGEric Glyman9:15

      Uh, s- we can certainly send... From the very first board deck, we- we- we tried to write this business equation out, um, and it clarified there- there were two cuts that led to the early Ramp strategy. The first, um, if you were to distill, okay, we had no revenue. We really wanna make revenue. Um, we asked h- where does it come from? And there was, at the time, three basic variables. Purchase volume, how much are people spending on the cards? Interchange rate of what people spend, how much we're able to keep. Uh, and then last, uh, funding cost, um, uh, for it. It was pre-rewards. Um, and if you looked at every one of those variables, well, if purchase volume went up, revenue went up. That was straightforward. You grow purchase volume. Interchange rate, um, you could see it very clearly. The more, um, purchase volume you had, the more interchange we would keep. So it was, okay, if you want to grow, um, interchange, grow purchase volume. Funding costs, similar thing. If you go to a bank and say, "I'd like to borrow $100 million, please," (laughs) um, they will certainly give you a lower interest rate than if you, um, say, "I'd like to borrow one." Uh, um, you will be better at predicting fraud, all that. And so every single one of these variables, we mapped out pretty clearly, what were the drivers of them, and it turned out that there was only one variable that moved everything, um, and it was just purchase volume. And so the whole question was, how do you make a product that people want to use and use deeply in the organization? Uh, and then there was... So that was the equation. There was a second cut, and- and I actually think it- it- it surprised a lot of people, which was, it was this very crowded field. How did Ramp grow out of it? And so the, what we- we- we compared it against, and this came later into the deck, were product needs based off of, uh, size of company. And so the- the bias that everyone in the industry has was you need to acquire customers as soon as they're incorporated, and if you don't get them, you'll never get them again. And so it behaved a lot like consumer cards.

   4. HSHarry Stebbings11:00

      Mm-hmm.

   5. EGEric Glyman11:01

      Um, you know, it was who could offer the most points? Who could offer the biggest sign-in bonuses? And, uh, who could get you into a lounge? Uh, and when you had, like, 10 people at a company, that was what people wanted. Founders wanted that. But when you got to, like, 15, 20, 25, more than that, the needs changed. People said, "My business is working. I just feel like I- I can't close my books on time. Suddenly people are telling me to contact sales, and I'm getting charged different rates. Uh, and people are looking more- for more predictability." And- and more critically, um, the problem was to buy one thing, uh, you needed two to three sets of products. Um, you needed an Amex. You needed a Concur, Expensify. You needed bill.com. And so what we said was, "We actually care a lot less, um, about being someone's first card. Uh, we're much more interested in how do you be their last? How do you help them spend less?" Uh, and so not only did it differentiate our positioning, but our- our business model was we made a percentage of purchase volume. And when you went from, like, 10 comp- c- employees to 20 employees, businesses didn't spend twice as much. They would spend like five, 10, 50 times as much in- in some cases. And so it allowed us to be much more surgical, just breaking or distilling what the market is telling us and mapping our strategy back to that, and just focusing the whole team on that.

7. 12:13 – 15:53

   Essential Advice for Founders on Speed Execution

   1. HSHarry Stebbings12:13

      My word, Bill and Concur are beautiful products, aren't they? Uh, I mean, the thing that always strikes me with your business is bluntly the speed of execution. And you have the day count-

   2. EGEric Glyman12:23

      Yeah.

   3. HSHarry Stebbings12:23

      ... which is so different, I find. When you think about advising founders, say, on having the same speed of execution that you have-

   4. EGEric Glyman12:31

      So-

   5. HSHarry Stebbings12:31

      ... what are one to two of the biggest piece of advice that you have to them on driving that execution speed?

   6. EGEric Glyman12:36

      I, I love that you started with the day count. It, it was act- so today is day 1,866, um, we do know it by the day, but it's not what you think. It's, it's not like a, "It's day 1,866, YOLO, carpe diem, like, go, let- let- let- let's, let's get it." It, it's much more focused on thinking about the passage of time. So the first time after, um, I don't know if I, if I ever told you this, w- when, when we met for the first board meeting, it was day 133. And part of the context we wanted to drive is, you know, it's, it's just four months old, here's what we've done. But in that October meeting, the second meeting Keith was talking about, we kind of mindlessly went to go update it, um, from 133 to 199, and it forced us, we were like, "Wait a minute, 66 days have passed. Do we get the same amount done with the same set of people? Do we get more done or did we get less?" Um, and as far as I know, no one has more than 24 hours in a day, um, but there are certain hours that certainly counted for a lot more, and it, it set off this, this thought exercise, um, for us, or, or more formally maybe called a calendar audit-

   7. HSHarry Stebbings13:38

      Okay. (laughs)

   8. EGEric Glyman13:39

      ... which we learned from, from Keith, which is, um, if you measure time in this way, it creates this space for you to say no to the things that are less impactful and to say yes to the areas of your business that create a lot more leverage. And for a small start- young startup just getting going, um, you know, uh, your most limited asset is time, um, and there are certain parts of your week that simply count for more. So, it was an obsession with measuring this and focusing your time on that. So the number one piece of advice I, I genuinely would give to, to early stage companies once you're past your business equation, you're spending time with customers, but, like, count the days, uh, and think about, uh, what gives your business leverage and what gives you a chance to out-compete, uh, others would probably be my, my top, I don't know how Keith would buy into this.

   9. KRKeith Rabois14:22

      And it was very stark too. So there's a lot of complexity in setting up issuing cards, just the very basic fundamentals usually take an infinite amount of time, usually 6 to 12 months is pretty good, as, as someone who's been involved in companies like this for, like, 20 years. I think Eric did it, like, 50 days, 40 days, yeah, there's different- Y- you need so many different layers of things you have to negotiate, different partners you need, and it was, like, 30 days, 40 days, 50 days, 60 days.

   10. EGEric Glyman14:45

       Exactly.

   11. KRKeith Rabois14:45

       It was, like, unprecedented.

   12. HSHarry Stebbings14:46

       Yeah. Can I ask, what did you say no to that you wish you'd said yes to?

   13. EGEric Glyman14:50

       Interesting. I, I have few regrets. I, I would say, in general, we try to say move pretty quickly, say yes, launch and test things because I know I'm gonna be wrong, um, about a lot of things I just wanna find out very quickly, um, uh, would be the general framework. I mean, I, there's different moments where certainly we could have capitalized, we certainly could have predicted, could not have predict... So something we don't do, uh, you know, historically, we don't store funds, um, you know, in the wave of bank failures a, a year ago, um, we, we could have been very well positioned to capture a lot of things. So there's things that, uh, we certainly would have benefited from. I would say, though, you know, we've tried to stay fairly consistent in having all of our products be focused on how do you help people save time, save money, you know, batch run that, and just be very focused on this workflow. Uh, and I think actually there's advantages to not changing directions too often. And so, though we may have missed some opportunities, I think the regular repeated motion and focus on the same mission, same goals creates a lot of advantages. And so I, I could list things we, we missed, but I, I don't have too many

8. 15:53 – 24:08

   Top Challenges Ahead in Growth

   1. EGEric Glyman15:53

      regrets.

   2. HSHarry Stebbings15:54

      Can I ask, when you look forward today at, like, the next, I don't know, 6 to 12 months, and you think about the hurdles that one has to overcome and the challenges that you face in this next stage of growth, both Keith and you, how do you think about what those one or two biggest challenges are that you have to see, identify, and then overcome?

   3. KRKeith Rabois16:12

      So typically, in organizations this size with this growth potential, it usually becomes organizational design, and just, like, how you simplify decision-making, how do you- where there's seams in the organization, like, how do you, how do you adjust that? That is very, very complicated because what worked historically sometimes often doesn't work when you multiply the number of people by three, four, five, and it definitely doesn't work when you multiply the number of people by 10. So even though the organization has go- led you to your success, the way you divide responsibilities, the way you make decisions is why you're here. At some point, that will break, and so you have to really think about that. The people, the people side gets increasingly more difficult, actually.

   4. HSHarry Stebbings16:50

      Eric, how do you think about that?

   5. EGEric Glyman16:53

      I mean, it's, it's... I think Keith is exactly right. There's other things, like we're, uh, we're thrilled, we're, call it 1% of how, um, you know, um, businesses spend money on, on cards in America, but there's 99% to go. How do you build a nationally known beloved brand? There's a lot of questions around that. Um, scaling the full platform, it's, you know, the majority of how we're, we're moving funds is, is, uh, uh, you know, it's, it's gone beyond card, um, and so just thinking through just the constantly reinventing or other questions.

   6. HSHarry Stebbings17:22

      On, on the org design side, I am fascinated. You- I think it's Aaron Levie says, "You need five turnovers before you go to a public company-"

   7. KRKeith Rabois17:29

      Yeah.

   8. HSHarry Stebbings17:29

      "... in the trajectory of company scaling." Do you agree with that? And how do you think about keeping-

   9. KRKeith Rabois17:34

      That's a lot. Uh, uh, I think you need some. Um, you know, Facebook re- reimagined the whole executive team. Aaron definitely turned over a fair amount. So some, there's some difference. It's like baseball. They, uh, Jeff Weiner likes to use the metaphor of, like, baseball, you have a starting pitcher who gets you a lead, that doesn't mean you want your starting pitcher to finish the game. And sometimes having a relief pitcher has different skills, different energy, is fresh, you know, et cetera, might be a good way to continue to have success. So you have to be, like, very judicious, but I don't, I wouldn't recommend, like, five (laughs) turnovers at all. That, that sounds actually scary. (laughs)

   10. HSHarry Stebbings18:07

       (laughs)

   11. EGEric Glyman18:08

       I, I, we- we're actually very unusual. I, I don't... (laughs) You know, we, we, we try, like, certainly we can make the, the dream hires. I, I do think there is a tendency, and I, I found this is particularly true from what I hear, it, it, it, you know, inv- investors who are not Keith is, "Go hire, you know, this executive from here and replace," and, and the supposition is that there are people outside the company who are inherently better, um-

   12. HSHarry Stebbings18:32

       Mm-hmm.

   13. EGEric Glyman18:32

       ... sometimes you need outside expertise, but I find there's nothing more motivating than taking someone who's doing well and stretching and giving more. And so, um, we actually try to-... you know, hire, um, somewhere in the org and stretch and stretch and stretch. And so most of the team you'll see today running Ramp is the same team as it was two years ago, three, a lot of the same people. And, and the depth of that and people just knowing each other's patterns, their strength, their weaknesses, um, are good, so.

   14. HSHarry Stebbings18:59

       I, I always remember, I think it was Keith that told me about like the smoothie test, where you, like, give someone a small task and then you expand it and expand it and expand it.

   15. KRKeith Rabois19:05

       Yeah, keep expanding. And this is definitely the PayPal model. We basically had no external hires that thrived, and so everybody was internally groomed from within. That's challenging. Usually you're dealing with a ratio of-

   16. HSHarry Stebbings19:15

       Yeah.

   17. KRKeith Rabois19:15

       ... you know, 70% internal promotions and 30% you're selectively adding new capabilities, so you're always playing around with some ratio. I don't think 100% works, but I also-

   18. HSHarry Stebbings19:25

       Yeah.

   19. KRKeith Rabois19:25

       ... think if you're below 50/50, then you're irresponsible in a different way.

   20. HSHarry Stebbings19:29

       Yeah. How quickly do you know when an internal hire doesn't stretch? When does the stretch not stretch?

   21. KRKeith Rabois19:34

       So I'll g- I'll give you a couple of tricks I've learned from actually just watching other CEOs. So Brian Chesky, for example, taught me the six months ahead test, which is can you think, like, an executive you have in place, leader in place, will deliver results, or you wouldn't have them there, like, right? By definition, sort of. They're not delivering the outputs that you expect, that business needs, you're going to need to make a change. But there's people who are delivering just in time and there's people thinking three steps ahead. And Brian sort of taught me six months is about where you want someone thinking ahead, because so- not every lever can be tuned in a week. Most can't. Not every lever can be even tuned in a month. So if you think six months ahead though, if you're really consistently six months ahead, most levers can be manipulated within six months so you land where you're supposed to land or bet- or beat expectations. So that's wha- like, one framework. Second framework that Jack Dorsey taught me is, is editing-writing meta- metaphor. And you know when you're editing or when you're actually writing. And you want to be editing people's work as the CEO. You don't want to be writing. And if you feel like you're using a lot of red lines consistently in the same area of a company, that's a really bad signal. The reason why I like this metaphor is people have a feel with what, what a red line means. Like, you know when you're taking someone else's work product and you're red lining. And if you're simplifying, that could be a good thing. Like, you know, like editors cross things out and they try to simplify, simplify, simplify. But if you're constantly asking clarifying questions and not really getting, you know, to the root, that's a problem, if you're s- systematically doing it in the same places in the organization. So those are the two, like, signals that I've learned to look for.

   22. HSHarry Stebbings21:04

       I constantly ask myself the question of, do we agree with like, hire great people and get out the way or promote and get out the way? Or do the best micromanage to an extent and are in- incredibly important ...

   23. KRKeith Rabois21:15

       So I'll give you a real answer. It's actually in my YC lecture that, uh-

   24. HSHarry Stebbings21:18

       Yeah.

   25. KRKeith Rabois21:18

       ... Eric alluded to. It depends. You want to be actually... They have a leadership training, um, uh, sort of deck I kind of built for Square, which is you actually want to be inconsistent on this, whether you're, like, mi- quote unquote "micromanaging" or you're giving a lot of rope, because it depends on the task-relevant maturity of the person and the project and the consequences to the business. So Andy Grove had this concept of task-relevant maturity from high output management, which is basically, has someone dealt with similar problems, you know, in their life? And there, there's a low, you know, never seen anything like this before to high, like, I've done this exact thing before. And depending upon that task-relevant maturity, you may sample, like using Eric's, you know, sort of metaphor, sample at different rates. Like, so if it's low relevant maturity, you might sample, like, every day or every week. It also depends on the consequences, th- how catastrophic or asymmetric is the result- are the results to the company. And so this metaphor allows you to figure out what to do. But ultimately, the CEO's ultimately responsible for everything. There's no excuses. So you can't ever abdicate. There's no such thing as, like, I let this person do X but I'm not responsible for the results when you're a CEO. So then you have to decide and use frameworks for how much rope do you give someone, which is a, really a function of your conviction level and the consequence level. So you have, like, almost like a two by two grid in a matrix. It's in my YC lecture of your level of conviction and the consequences, and whether to intervene or delegate or what to do, you know, in the other cases.

   26. HSHarry Stebbings22:42

       Eric, can you talk to me about when you've had the highest level of conviction and you've been wrong?

   27. EGEric Glyman22:46

       What I would definitely say for, like, any- anyone listening is like, uh, I would listen to what Keith said twice. I, I think it's really, really good. And I, I think that, um, people don't always in- internalize the importance of, uh, as a founder, as a CEO at the end of the day. Like, um, everything that's going well, for a good thing, like credit to your team. Everything that's going bad, like, is your fault and you need to fix. And, like, that is the right mentality. Like, for me the question is not will people be wrong, it's a question of how does your company deal with it culturally? And, and I, I think that, uh, two things. When good companies talk about wins, great companies talk about misses. The way that we start, uh, every board deck is what is going well, what is going less well than we hoped, uh, and why. And so we, you, you know, it's, it's not letting there be a culture of sweeping things u- under the rug, uh, and not crucifying people when things do, do go wrong. You know, and next, um, you know, Karim is, you know, oversees, um, you know, half of the org, a lot of the most important in- investments. He has every right to, to tell me, like, on marketing, you know, this is off or you're missing this and here's what we're seeing. 'Cause a lot of what we're doing, it's a, it's a pursuit of the truth. It, it's not about, um, you know, was someone right and, and, and kind of acing your own scorecard as an executive, is what moves the company forward. And we've certainly made m- made mistakes. Um, we're gonna make more. Um, it's just a question of, of can we create the right system to bring

9. 24:08 – 29:17

   What Defines a Great CEO

   1. EGEric Glyman24:08

      that up?

   2. HSHarry Stebbings24:08

      People say that it's a matter of one to two big decisions a year is what defines a great CEO, and then others say it's an- about number of at-bats. It's about just getting a lot of decisions constantly right.

   3. EGEric Glyman24:20

      Yeah.

   4. HSHarry Stebbings24:20

      Again, two different schools of thought and I'm like, w-I don't know. What do we think?

   5. EGEric Glyman24:26

      I definitely lean the latter. For most people, um, uh, I really do think that activity, um, and hours, uh, o- of yourself and of others being spent in the right place, uh, is leverage. And so certainly it's some extreme, uh, I think it's about one to two, but for the vast majority of people, I- I think it's the number of at-bats and having the system to understand your own view of- of reality and spend your time in the right places, um. A- and- and- and I- I just think that what I wouldn't underestimate for- for most people is most of the world is fairly linear, like an hour in in any way is an hour out. But unfortunately, in- in company building, there are some things that just matter a hell of a lot more than others, um, and, um, creating the space to, you know, where you take an at-bat so you can find those things and just double down on it is a lot of the way we think about things.

   6. KRKeith Rabois25:17

      Yeah. A lot of company building is you want to tap into a vein. Like inertia is not your friend. When you start a company, everything, time's not your friend, the world doesn't care about you. You have invert, literally invert inertia. So you have to create momentum, like literally the physical definition of momentum, and then you want to figure out, "How do I amplify that momentum?" Or sus- minimum sustain it and then, as long as possible, and then ideally amplify it. That's usually tapping into like a new idea. And so you want to have s- a system that allows you to surface new concepts and ideas and look for that, "Does this have that explosive potential?" And then double, triple, quadruple down.

   7. HSHarry Stebbings25:50

      What are the lowest hanging fruit ways of creating momentum? And is it possible to manufacture it artificially?

   8. KRKeith Rabois25:56

      I think you can manufacture it. I mean, YC teaches that, you know, proverbial, do things that don't scale. I think there's some truth to that. That you get... The- the body in rest stays in rest kind of thing from physics again. And so you've got to push somehow, and then when you get the body moving, like fric- you know, you want to reduce the friction and accelerate the speed. It's really is physics actually, (laughs) it's just applied to startups. So I think you can sort of pseudo-hack your way into momentum. You can't scale it forever if you're really hacking it. You're gonna have to make the train rocks- train tracks real, and you're gonna have to make the engine that propels things pretty real at some point, and the sooner the better. But like, starting somewhere. Like I've seen examples where people start with PR. Doesn't usually work, but once in a while you can get a spark. And then if you see the spark, then you can figure out how to do it more scalably. Or once in a while you start for friends and family. And sometimes that works, often it doesn't. Like all these are custom, there's no formula for building a startup.

   9. HSHarry Stebbings26:53

      Keith, when have you had the most conviction and it's turned out to be wrong, and what did you learn from that?

   10. KRKeith Rabois26:59

       As an investor or exec or...

   11. HSHarry Stebbings27:01

       Either. Which one comes most strikingly to mind?

   12. KRKeith Rabois27:04

       Um, well, you know, in- the investor thing is weird. E- We've talked about this a few times. On- on almost all the really good investments that I've touched, I kinda knew in the first three minutes. It's very bizarre, um, a- and you'd say that that sounds crazy.

   13. HSHarry Stebbings27:20

       But have you missed, Keith? Like I love you dearly, like it doesn't seem that you have really. (laughs)

   14. KRKeith Rabois27:25

       Um, not someone I met in person.

   15. HSHarry Stebbings27:27

       Mm-hmm.

   16. KRKeith Rabois27:28

       I've, as I talked about, I've screened out people I definitely should have met, and absolutely would have if- Um, but like in person, I've always made an offer to someone who's done super well. You know, we talked about the biggest mistake ever was not raising the valuation on Rippling. Um, you know, Parker was nice enough to Tweet our episode afterwards, which was really helpful in my sleep. But... (laughs)

   17. HSHarry Stebbings27:48

       (laughs)

   18. KRKeith Rabois27:49

       But like you're gonna miss. As an investor, like, look, let me put it this way. I'm playing baseball. You hit 400, you're Ted Williams to go to the Hall of Fame. Early stage investing, you're gonna miss more than you hit, like period. Like and you just admit that. As executive, you cannot be right 40% of the time and miss 60%. Imagine hiring, for example. Imagine 40% of your hires were good and 60% were terrible. Like you'd fire that person as a leader, like you'd have to. Like you can't have an organization with that ratio.

   19. HSHarry Stebbings28:14

       Would you though? 'Cause actually I have so many people on the show who say, "At best, you're 50% right when it comes to talent and assessment."

   20. KRKeith Rabois28:20

       No, you can't have s- you can't have 50% fires. You can have 50% like A level and 50% like competent, but good luck running an organization where maybe a sales team you could have something like 20-30% wrong.

   21. HSHarry Stebbings28:34

       Sh- should you have competent people in your company? I know that sounds strange but like competent is like B minus.

   22. KRKeith Rabois28:40

       I don't know. It- it... There's a lot of dimensions to this. It'd be like a topic in and of itself. But actually, so whereas most wrong as an executive would be on hires.

   23. HSHarry Stebbings28:47

       Mm-hmm.

   24. KRKeith Rabois28:47

       Like, you know, one bad hire, just one, can really set back an organization and undermine the momentum pretty severely.

   25. HSHarry Stebbings28:55

       Mm-hmm.

   26. KRKeith Rabois28:55

       And like everybody who's ever hired like hundreds to thousands of people, you're gonna make some mistakes. You don't even want zero defect hiring. That's actually a flaw that has even worse consequences probably. But there's no way you build organizations over the years and don't make mistakes, and some of those mistakes definitely cause a company to lose a year of momentum and you may never get it back. Hopefully you

10. 29:17 – 31:52

    Direction for Today's Entrepreneurs

    1. KRKeith Rabois29:17

       can reignite it.

    2. HSHarry Stebbings29:18

       I- I- I- I totally get you and agree. Eric, you've mentioned before about the unique moment in history that we're in now for people to build in. I'm just intrigued to hear when you think about that then given the unique moment in time, for young people today, for founders listening, what do you, what do you think they should build? What should they not build?

    3. EGEric Glyman29:36

       I'll start with your question of what should people build or- or not build, and then I'll, we'll- we'll talk about the unique, um, point in- in history that we're at. Uh, and in many ways they're at odds. What I would say is that, uh, great businesses, 99 out of 100 times, uh, start with being really curious about people's problems, uh, and how can you solve them, uh, in a better and deeper and more true way. Uh, not with what are the new breakthrough capabilities of technology and how do I apply that. Um, uh, when I think about, um, what is timeless about Ramp's business...Um, we started with the question of, you know, most business people, finance teams, people building companies that I know, um, weren't interested in more points or more lounge access or anything. They were interested in how do you be more profitable? How do you spend less? How do you go home earlier, achieve more with a s- with a, with a smaller team? And what we found that was so timeless about, um, really the starting mission of, of, of, of Ramp was this enduring thing. I couldn't imagine in the future pe- people saying, you know, um, you know, to maybe, uh, paraphrase Jeff Bezos, I couldn't imagine people saying, "You know, I just wish I was less efficient with my spending. I wish it took longer to close my books." Um, you know, it, it's ludicrous. That would never happen. What they would (laughs) always tell you, you know, is that, um, "You know, if I could get the same things for less, I could reinvest more." And, and we started with this timeless belief, people want to spend less, and then came in search of this technology. Um, you know, w- w- we realized it was possible suddenly in 45 days if you were focused to get live on Visa within 60 days of a card in someone's hands, and then build lots of tooling and infrastructure so that you're not just making transactions, but you're collecting receipts, you're closing your books, and you are constantly showing new ways to spend less. And so I would start with, um, actually, uh, not trying to come up with ideas but being curious. Um, um, asking where is, where is pain? What are problems that you have? Savoring it and going a lot deeper. Um, and of course, it's in how you carry your business too, doing right by people, investing in relationships, um, making your service better every month, being diligent when people complain and trying, you know, uh, to actually solve their problem and listen, being there are all timeless truisms that I would focus

11. 31:52 – 34:07

    Identifying Market Potential

    1. EGEric Glyman31:52

       on.

    2. HSHarry Stebbings31:52

       How did you identify this as a good market, bluntly? When you talk about, like, solving a problem. You know, as you mentioned, Concur, Bill, many incumbents which oh, we agree are shit, but also then a lot of other startups at the time, and other kind of scale-ups doing the same. Uh, how did you identify good markets and how do you, and I guess Keith think about that?

    3. EGEric Glyman32:10

       In many ways, this has been like a, a decade-long project. It is year five of Ramp, but it... There are aspects of Ramp which were a continuation of Paribus which Harry, you, you first interviewed me about this in 2015, um, almost a decade ago when both of us were, were getting started. And we had this obsession-

    4. HSHarry Stebbings32:27

       (laughs)

    5. EGEric Glyman32:27

       ... you know, de- (laughs) How do you help people spend less in, in, in, in sop- in, in, in, you know, in, in... It was in, if you'd call it that in 2015, it was an AI agent, it lived in your inbox and it would file for refunds when you bought things and it dropped in price. We were bought by Capital One, and we were obsessed with this notion of what if there was, you know, um ... Over the years, we'd built this idea of good credit cards are really profitable but they're misaligned. Every card company is trying to get people to spend more money or earn more points, that kind of thing, and got obsessed with the idea of what if they help you spend less. It's evolved a lot since then. We still are very much focused on delivering a great card that helps people spend less. We do that, we believe that we are the best at that, um, still. But one of the things that we started noticing because of the system of getting input and, and, and tuning was in the second board deck, you'll see the first, it was the first time we ever talked about saving time. We missed it in the first time around. So we knew cards were profitable and misaligned, but there were some early customers who said, "I, I can't stand Expensify. I spend half a million dollars a month. No one will turn their stuff in. If you could just replace that, I would use you entirely." And we were open in listening enough to our customers to say, "You know, okay, this is very profitable. This would drive our business equation. People would want to use it more." And our mission is help people spend less money and time, and so it was both true and timeless to customers, and served a very large TAM. Um, and I think that's still continued to help us. We, um, you know, are, uh, uh, still about the same thing truly we were about five years ago. Every per amp product should be designed to save people time and money, um, but even five years later, at, you know, well into the nine figures in revenue, uh, we are less than 1% of the market in which we were largest. Um, and that helps us just continue to grow and build venturous scale, and, and we hope, um, societally shifting businesses.

12. 34:07 – 38:38

    Product to Platform Challenges

    1. EGEric Glyman34:07

    2. HSHarry Stebbings34:07

       Keith, what are the biggest challenges one faces when making the transition from product to platform?

    3. KRKeith Rabois34:12

       That's a great question. I think generally you start with product is a much better formula and then it evolves into the platform. One of the challenges is when. Like sort of, how do you ... You know, when do you know, like how do you know, like it's the time to do that? Second would be how do you not lose sight of what customers care about? Because it's very ... And it's related to the sequencing point but it, it's not so obvious. Like we have a very clear value proposition, right? Save time, save money. It's super clear, we can measure it, we can amplify it, can play it back to people, show them how much time, how much money we're saving them, et cetera. When you go p- uh, platform more broadly, you're somewhat sometimes losing sight of that unintentionally, but so you want to make sure that the thing you're building that's broader and horizontal will create even a bigger value proposition. And sometimes you don't know that right away. Um, sometimes you can measure it, test it, you know, get a feel for what that would be, but like now we know sort of, like for example, Ramp, when you have all of your data, when you run your entire finance organization, when every th- all the data seamlessly interacts from everything you do, there's value propositions that you could never ever imagine doing, and certainly not imagine doing except by hand, uh, unless you had one integrated system. So now it's very clear, like if you have all your data flowing through in real time, one system, what are the things you unlock? You can name them and now it's Eric's job (laughs) really to prioritize them 'cause there's like 10 or 12 really great value props.

    4. HSHarry Stebbings35:39

       When you think about that kind of platform expansion play, and the, the data benefits that one gets from being in that platform play, is this a monopoly market? You know, obviously Keith, you were at, you know, Founders Fund before where Peter talks a lot about monopoly markets and the benefits that come from them. Is this a monopoly market or is this a more even distribution of outcomes?

    5. KRKeith Rabois35:58

       Well, I think all CFOs of the future will want to run on Ramp, and they'll be reckless if they don't. They'll be more efficient, more successful, and more strategic. So for example, typically a finance organization spends about 17% of its time today doing strategic analysis, and 83% doing mundane tac- tactical tasks. So obviously, if CFOs want to be more successful, more impactful, more influential, they're gonna want to run on Ramp 'cause that's gonna enable them to show- showcase their skills.

    6. HSHarry Stebbings36:26

       I'm sorry, Eric, as your CMO that I've just appointed myself, your new billboard is the Keith Raboy picture with the quote, "Is your CFO reckless?"

    7. EGEric Glyman36:36

       (laughs)

    8. KRKeith Rabois36:36

       (laughs)

    9. HSHarry Stebbings36:36

       Come on, come on, that's good. I-

    10. KRKeith Rabois36:38

        I love that.

    11. EGEric Glyman36:39

        (laughs)

    12. HSHarry Stebbings36:39

        I'm here all day, baby. This creativity-

    13. EGEric Glyman36:41

        It's great.

    14. KRKeith Rabois36:41

        Fear, f- fear, fear does work. (laughs) Like you don't wanna be irresponsible. Actually, like when I led an investment in Stripe in 2013, I think, I actually said something like that to some journalist, that it would be irresponsible to build a business, except on Stripe.

    15. HSHarry Stebbings36:53

        Yeah.

    16. EGEric Glyman36:53

        (laughs) I love that.

    17. KRKeith Rabois36:55

        And she was like confused as hell. And turns out, it's irresponsible to build a business, except on Stripe. (laughs)

    18. EGEric Glyman37:00

        What Keith is saying, th- there's something really there though. So, so it's, you know, anyone who's ever hired a fire, finance function and, and looked at who comprises what, you know, what are the job titles if you're working in finance, I think statistically it's something like 5% on the order of that, uh, of finance jobs are categorized as strategic finance.

    19. KRKeith Rabois37:19

        Hm.

    20. EGEric Glyman37:20

        What's amazing to me, uh, about that statement is people aren't internalizing, "Are 95% of finance roles non-strategic?" Is, is that just something that people are used to? And I think the answer is unfortunately yes, because it is so hard historically to, you know, automate a lot of processes. There's so much pain in getting people to turn receipts in on time, to tag every transaction correctly in your general ledger, to pull, um, data from multiple systems down into a pivot table, cleanse it, put it into a new system to project forward.

    21. HSHarry Stebbings37:52

        Is there a data network effect that you can apply across different customers in Ramp?

    22. KRKeith Rabois37:56

        Well, absolutely should be. Like for example, Ramp will be your fiduciary agent, so it's gonna save you time. Make, I mean, save, save you money. Make sure that you're allocating your resources in the best, most prudent way. Secondly, it's gonna be your compliance agent, which is we have processes and procedures. We just wanna make sure that they're adhered to without throwing lots of humans at it and making it painful on the employees, they don't get the work done. So like you can only do things like being your fiduciary agent and your compliance agent and then be insightful, let's say, at the strategy level, unless everything, all the data from all sources, from all employees rolls together.

    23. HSHarry Stebbings38:28

        I totally get it. Um, fuck, Eric, I can't believe this. (laughs)

    24. EGEric Glyman38:35

        Eric, we have a long way to go.

    25. KRKeith Rabois38:36

        (laughs)

    26. EGEric Glyman38:36

        It's not too late.

13. 38:38 – 44:53

    Quick-Fire Round

    1. EGEric Glyman38:38

       Don't, yeah.

    2. HSHarry Stebbings38:39

       Okay, listen, we're gonna do a quick fire and that's gonna be part of it. Um-

    3. KRKeith Rabois38:42

       Yeah.

    4. HSHarry Stebbings38:42

       If we, if we take the flip side though, and I'm not being European and I think Keith's got views on Europe, which I do share.

    5. KRKeith Rabois38:48

       (laughs)

    6. HSHarry Stebbings38:48

       But if you were to do like a post-mortem of like what could go wrong, okay, what's the one thing that could harm the trajectory of the company?

    7. KRKeith Rabois38:56

       Well, we talked about hiring the wrong people, which fortunately I think the super IC model is working extremely well here and the company's extremely proficient at it. So I don't think we'll hire the wrong people. I think typically in the core business where we built, um, like the Ramp initial traction, getting risk wrong, um, on either side, either being too conservative or not really understanding risk. Anytime you're moving a lot of money around, it has to be in your DNA. And unfortunately, we've been extremely, uh, proficient and the performance is extraordinary. But like that is typically something that goes wrong in a company that looks a little bit like this. And then third is prioritization, I think.

    8. EGEric Glyman39:32

       I, I would double down on that. You know, in many way, I think that the, uh, it is much more in to build a compound startup. Um, you know, in the age of, of AI where there's, you know, um, ability to process a much vaster set of data more accurately, um, and automate workflows, uh, there are advantages, uh, to streaming multiple workflows to powering multiple parts of the platform. To your question earlier, Harry, of as it goes from singular value propositions to you can run all of your finances through Ramp, uh, as tens of thousands of businesses do, um, you could, the risk is to lose focus and to not make your singular products that are best in class and to lose that edge. Uh, so I, I think just being very aware of that and focused on excellence, um, is critical.

    9. HSHarry Stebbings40:17

       You mentioned time-saving as a core value prop. What is one thing that both of you do today that you wish was not on your calendar?

    10. EGEric Glyman40:26

        (laughs)

    11. KRKeith Rabois40:26

        (laughs)

    12. EGEric Glyman40:27

        It's, um, you know, I have, I have a lot of reports right now. I think it's the flip side of, of, uh, one-on-ones. I don't... I'm, I'm not quite with Jensen with, uh, 60 direct reports, but I do spend a lot of time in, in one-on-ones and, and there's probably ones that merit the full 30 others that, you know, you can skip. And so that's probably, uh, to your point, Harry, something I'm gonna change next week. (laughs)

    13. KRKeith Rabois40:47

        So I, I don't really have-

    14. HSHarry Stebbings40:48

        Keith?

    15. KRKeith Rabois40:48

        I, I, I like try to redo my calendar every week, so there shouldn't be anything on there (laughs) that I, I don't wanna be doing.

    16. HSHarry Stebbings40:54

        Keith, how many direct reports is too much?

    17. KRKeith Rabois40:57

        So the canonical view from Andy Grove was like five to seven. But obviously, Jensen's done extremely well with 60, so it's clearly not like (laughs) a, a ironclad rule. Um, I, I think the, the reason for the five to seven was you wanted to do your weekly one-on-one sort of thing at a certain pace. I think that's a little archaic. I used to get yelled at, I went pretty flat at Square when I was running hot at like 11, and like Vinod and Roelof gave me feedback that that was too much. There's some, it, it totally depends also on your level of depth actually as an executive. When you know certain content areas, subject matters and like you have some ability in that craft, it's a hell of a lot easier to have a direct report there 'cause you're diving in really quickly to what's key. When you have to manage functions you don't really know that well yourself, there's a lot more preparation drag coefficient. So I don't think there's a one-size-fits-all formula, actually.

    18. HSHarry Stebbings41:47

        Okay. Final one for you both. What was the last price on the Round? It was public, I think, so it was like 6.7 or whatever it was.

    19. EGEric Glyman41:55

        It was, uh, it was a seven point, uh, s-

    20. KRKeith Rabois41:58

        Six?

    21. EGEric Glyman41:59

        Yeah, $7.6 billion post.

    22. HSHarry Stebbings42:01

        Fuck, I lost like almost a billion dollars. (laughs) Sorry about that. That's 7.6. Okay, so let's outcome scenario plan this one. Like first Keith and then we'll finish with you, Eric. But Keith, when you underwrite this, what do you underwrite it to now? And like-

    23. KRKeith Rabois42:17

        I don't-

    24. EGEric Glyman42:17

        ... who's-

    25. KRKeith Rabois42:17

        ... kind of do that kind of analysis. I kinda think about like, again, the inputs and like how big could this be in terms of like-

    26. HSHarry Stebbings42:22

        Yeah.

    27. KRKeith Rabois42:22

        ... would every company in the planet use this and feel like they need to use this?If every company on the planet feels like they need to run their business on Ramp, it's (laughs) gonna be worth a lot more than 7.6 billion. So, I don't try to get, like, too nuanced about it. Um, I need to believe that there's a vision there of like, "Wow, this is actually very realistic and a certain amount of probability that this can happen." But I don't, I, I don't, like, build spreadsheets and, like, try to underwrite to 3X, 4X, 5X. N- n- not what I do, at least.

    28. HSHarry Stebbings42:52

        All right. If I, if I were to say to you, 10 years, like, where's Ramp then? It's 2034, which is almost the time that we first met. We met nine years ago.

    29. KRKeith Rabois43:01

        Yeah.

    30. HSHarry Stebbings43:01

        Christ. Take that time again. Where's Ramp then if everything goes well?

Episode duration: 44:53