What Building towards Billion-Dollar Products Really Teaches You - CEO Jeff Seibert

0:00One of my mindsets of like, where should leaders be involved? And so what I've told our team as we scale is like for everything we take on all across the company, it doesn't matter which department or thing, I want to be involved in two places.

0:16What would happen if we just told the truth? Welcome to TruthWorks, where we dig into the nitty gritty of leadership and work. And what needs to change. I'm Jessica Neal. And I'm Patti McCord. Our journey together started in HR, but trust us, it's evolved into something wild, honest, and well, a bit rebellious. So throw out the handbook. We're here to redefine rules to work for us, not against us. Let's dive into another episode.

0:54Hi, everyone. Welcome to another episode of TruthWorks. I'm Jessica Neal. Patti McCord is going to be rejoining soon. So keep tuning in if you want to hear her. She's been out because she's had a little recovery time. But I have a really cool guest today. You know, there's like a moment when every leader hits where the job stops being creative and starts being reactive and becomes a stream of updates and small decisions and fires. And the real leadership that you're supposed to be doing gets squeezed out. and the guest that I have here today Jeff Siebert has spent his entire career trying to solve that problem he is the creator of Crashlytics which is I guess I learned this in researching you I guess this is like a thing that we probably all use but we don't really even know we're using it which you can explain to us later and now he's a founder of this company called Digits which he's going to tell us about.

1:55But what I'm so excited to talk to you about today is just that, like focus and chaos. And we just live in this weird reactive state all the time. And I want you to help us get into the strategic state, right? And then not even strategic, but where should we be spending our time. Hi, Jeff. Sorry. Hi, Jessica. Thanks so much for having me. It's great to be here. Yes. Good. So I don't tell us, tell everybody who you are, because I kind of described you, but tell, tell everybody who you are and where you are today. And let's get into all this chaos stuff. Yeah. Well, thanks again for having me. I'm Jeff Seibert.

2:41I'm the founder and CEO of Digits. We are building AI native accounting software or really accounting software that works for you, that does all the tedium for you. And that's everybody just like was like accounting. I know, I know, but I know the dots will connect because I had no background in finance or accounting. My whole life has been coding. I taught myself C when I was 12. And I literally like what, what got me so excited about it back when I was 12 was I realized you could write the code once and the computer could do that a hundred times, a thousand times, a million times, a billion times, it didn't matter.

3:14And so the energy invested was like not correlated to the output you could get. And I just felt that that was so powerful. Like what other activity do you do where that's possible? And so that I just, I fell in love with anyway, moved out to Silicon Valley, went to college out here. Wait, where did you, where were you when you were 12? I grew up in Baltimore, Maryland. Okay. I can kind of hit that accent. Yeah, in the city, sort of northern edge of the city near Johns Hopkins. And it was a great place growing up. That was awesome. Not a lot of technology going on. I had sort of stumbled upon this because my mom actually gave me the book Mac Programming for Dummies for Christmas in the mid-90s.

3:56And I had no interest or experience with computers. And I just sort of read it. And honestly, I didn't understand it. And then I put it down. And six months later, that summer after school, I read it again and tried the Hello World example. And that's when the light bulb went off. And from that moment, I've just spent like most of my life coding. Was there a reason why she wanted to buy you that for Christmas? So I didn't know this or appreciate this at the time. She had actually minored in computer science in college. Interesting. And was doing punch card programming. She never ended up using those skills really, but like had kept it in the back of her mind.

4:32It was amazing. That is so cool. And ahead of her time as a woman. Yes. Yes. So, okay. So then you moved to Silicon Valley. You went to Stanford, right? As many of you founders do. And so I'm guessing that that was sort of your first kind of taste of like real like technology and this spirit of entrepreneurism And yeah, that was the main realization because I had been writing and selling software. I had my own sort of quote software company in high school. You did? I did. What kind of software were you selling? So many programmers get into it from gaming. And I had been playing this game, this Mac desktop game, and I got bored of the game.

5:21So I started, it was called Escape Velocity. One of the greatest games of the mid nineties. Okay. But I ended up getting bored of the game. And so I then started hacking the game and then I decided to write an editor to modify the game. So then I sold this editor to the whole sort of community around the game for people to modify the game. How much money were you making? Not a lot. I think I in total, I netted a couple thousand. What'd you buy? Oh, my. Another game? I can't even remember. Yeah. What was interesting? I mean, this was the mindset difference because in Baltimore in the mid 90s, like I was like, I have a software company, right?

5:58I write software. I sell software. I then came out to Silicon Valley. And at Stanford, no one talks about running a software company. It's a startup. It's high growth entrepreneurship. It's the idea of how do you get what you build out to the entire world? And so that was the total change in mindset. And I spent my entire time on campus going to entrepreneurship lecture series and trying to meet as many folks as possible. And this was now the early 2000s. Everyone was like, what's the next Google, right? Like who's going to be the next big company to come out of campus? And so that was the really like light bulb moment of coming from the East Coast to the West Coast.

6:38So then you're at Stanford and then what happens? So graduated 2008, right as the housing crisis was hitting. Yeah, it was a great time. Really good time. And a couple of friends and I just sort of failed to apply to jobs. We were working on some side projects and we had actually won a class competition to build a site. And so we just kept working on it, didn't apply to any jobs, ended up meeting a couple of investors. and we went into DFJ's offices on Sand Hill and met with an associate there and just wanted to show him what we had built. And we actually didn't think the meeting went all that well.

7:14But at the end of the meeting, he's like, let me just see who else is around. I'll be right back. And he comes back in the conference room with Tim Draper. And Tim goes, all right, you have three minutes. What did you build? And so we quickly show him our site and he goes, how much are you raising? And of course we weren't raising. We were just there to get advice. Yeah. And so we make up a number on the spot and say a half a million. And he literally says, I can do that. And he walks out. Great. So that's how you've got your first funding. That is literally how we got it. We later found out DFJ had what they called a silver bullet program where any of these, any partner could make a deal under a million, no questions asked.

7:54I like that rule. So that was sort of interesting, but yeah, so that was our now like kickstart into startup life. So we ended up closing the round from them, got a little office space in Mountain View and started building document collaboration software. Okay. That was sort of my first company. We ended up getting acquired by Box in 2009 as the first acquisition Aaron Levy ever made. And our tech powered all document preview, document display collaboration on Box for the next five years, which was pretty cool. So cool. And then you're like, I have to do it again. So the problem is what I've described is I think I'm just cursed because I love coding.

8:31And so if I have nothing to do, I code. And it always spirals into something. So at Box, I ended up, they sent me back East. I moved to Boston to open and run Box's East Coast office. And we ran our sync team there. So competing with Dropbox because Dropbox was crushing Box at the time with their technology. And sync algorithms are super complex and buggy. And it started crashing a lot. And so I started spending my nights and weekends, of course, coding, learning about and trying to figure out how crash analysis would work, how we could automatically determine what line number of code we needed to fix.

9:07Right. And because again, back to your theme. That's what I do on the weekends too. I mean, we're so similar. Back to your theme, that would save a ton of time. Yes. Right. And so I put this prototype together of what would become Crashlytics. And I actually called one of my friends from Stanford, who was Mikey Krieger, who had just started Instagram. and Instagram then was still very small, but I called him up and it was like, Mikey, I don't mean to be rude, but Instagram crashes. What do you do about it? And he said, I get more crash reports per minute than I can read. And that was the light bulb moment for that.

9:41And I was like, okay, this is a startup. Right. Yeah. And so then this Crashlytics has been implemented like everywhere. It is. We got so fortunate with market timing. So yeah, we launched basically at the end of 2011 as mobile was ramping and it took off. We were on 300 million phones within 12 months. We get acquired by Twitter. It keeps scaling. It then gets acquired from Twitter by Google in 2017. Today it's on 6 billion MAU. So effectively every smartphone on earth. Look at you now. And now. Just trying to save time. I know, which I love. Now you're doing it again. Now we're doing it again with digits and trying to fix accounting because accounting is so tedious.

10:27It's so tedious. And this actually was like, you lost everybody with accounting. Right now. But the accountants are like, yes, well, maybe not. Maybe that you're going to. No, they actually desperately want it because you're not getting them out of jobs. Yeah. You talk with hundreds of accountants. None of them enjoy the accounting, right? The actual accounting is very tedious. They love working with clients, advising the business, right? Like actually helping management strategy, et cetera. Yeah. And so if you can eliminate the accounting tedium, it's actually a huge unlock for the profession.

10:59Okay. Well, look at you. So, so now you're doing that, but then you're on the side sort of talking about this time efficiency chaos thing too. Yes. Yeah. It's just a broader passion of mine. I mean, going all the way back to middle school, I just felt like if you can help people save time by automating the parts they don't want to do, you just unlock so much more creativity, free time, other passion projects, whatever folks want to spend their time on. Yeah. And maybe you kind of already mentioned this, but I too, as we were talking before we started recording, I too have kind of found that that's my life's mission too.

11:40Like I never thought about myself as a chief human resources officer, because it sounds horrible. I was like, I was the efficiency officer. You know, how do I make people more efficient? Because it's all the things like miscommunication and, you know, nobody's got the context to, you know, people are spending the time on the wrong things or they've got too many direct reports or like whatever. But what, and you might've touched on this already and just talking about your history, but what was like, you had sort of this game and you were like making the game more efficient for people. But was there like a moment in your career where you saw this and it just fueled that passion even more so?

12:25So the funny thing is it was actually somewhat in reverse. So I remember I was like in love with coding. And so through high school, I wanted to spend all my time coding, but I had to do homework. and so i started actually i know i started thinking about it from the reverse is like what could i write to try to automate like homework and stuff so that i could spend more time coding yeah that's so funny did you were you able to automate so there was one example we had in science classes we had this whole year we had to build we had to deliver graphs as part of our homework assignments right like histograms bar charts and so on and the option was you could either drawn by hand, which most students did, or you could use Excel, which I always found way too complicated to figure out.

13:12And so I actually wrote a Mac desktop app again, back in the late nineties to draw graphs. And so it made it very easy to just type in or import the data and then boom, it printed out the graph and I would print it out and turn it in as my homework. Of course you did. I mean, you know, kids, if you're listening, you know, you can automate homework. There's a way to do it, especially now. I mean, back then it was harder. You can probably automate too much of the homework. It's a slippery slope. I know. My six-year-old doesn't have any yet, but I can't imagine by the time he does have homework, how easy it's going to do because it just is going to get better and better and better.

13:49Yeah. I mean, so on the side from digits, I'm actually chair of the technology committee for the high school I went to. So it's come full circle. And now we're advising the school on how to use AI, how to allow it for assignments, how not to allow it. It's actually a really interesting problem space. Well, yeah. And okay. Well, this kind of leads into the next question, but before I get to the next question, how should they use it in schools? I just want to know, like, what do you think? Yeah. It really comes back to the invention of the calculator, which blew up math programs. Right. And so in a good way, right.

14:23Yeah. We weren't allowed to use the calculator. Exactly. Exactly. Yeah. And so now, of course, everyone has a calculator in their, in the form of their phone, wherever they are 24 seven. And so there's no real need to learn long division and memorize all the different multiplication tables and so on. What's interesting with AI is there's, there really seems to be two ways students are using it. Obviously a lot of students are using it to try to cheat or get out of it and do the work for them. Of course, That would have been me. What's been interesting is the really smart students are using it to push them and help them understand better why something is true or not in their work and go a level deeper than they're able to from the problem set or the textbook or whatever it is.

15:08And so we're also seeing kids get way ahead of their normal grade level because they're using the AI to tutor them. And schools could train the AI to do that and not to do the inverse, correct? That would be the idea. I would hope so. Can you do that? I guess you could change your... I guess you could probably fine tune it. Yeah. Right. Say don't allow students to do X, Y, Z. Right. Right. But be their coach. Help them go deep. Like, right? I think... I don't know. You know more than I do. Well, we had a really funny session where we had the class president, the student president of the school come and meet with the board.

15:48And we asked about the honor committee and like what's happening because all of the honor violations are, of course, all AI related. Right. And folks having chat GPT write their essay and so on. Right. And we asked him, like, how do you know they cheated? Yeah. And he was like, oh, it's so easy because, you know, the student, you can see their normal work and what they write. And then you see this and it's like, come on, you obviously didn't write that. Yeah. No, that's the other thing, kids, if you're listening, don't got to like tailor it to be you a little bit. You can't just have an impeach chat GPT.

16:19be. Listen, I was, I hated homework, you know? And we also, with this memorization of like long division that we had to do, do you think as, cause you're a math guy at the core, right? That's what coding is all about. Like, is that a good thing or a bad thing? I don't, it's a real difference. I think between memorizing stuff versus understanding it. Like, I think you should, you should understand the premise of long division. Yeah. I don't know if that means you need to memorize the exact mechanics to do an arbitrarily hard division problem. Yeah. Okay. Here, I hear you. Okay. So that leads to my next question, which is like, when AI removes like the noise and small decisions, like what, what type of insecurities, if we think about the workplace, does that reveal like with leaders and, and just more broadly?

17:13that's a really interesting question and i think touches on sort of one of my mindsets of like where should leaders be involved and so what i've told our team as we scale is like for everything we take on all across the company it doesn't matter which department or thing i want to be involved in two places i want to be involved at the 10 of the project so i understand like what we're trying to do why we're doing it like what's what approach do we think we're going to take just so like we have directional alignment at the start. Yeah. And then I want to be involved at the 90 % phase so I can add the like remaining polish and like get in there and tweak the copy, literally like polish the pixels.

17:51Like are the corner radiuses correct? Like how, what is the button location and all that stuff? Because that adds that, like that just feeling of like, oh, this was really well thought through and well designed. Um, and that goes for marketing product sale, everything we do across the company, like the pitch decks we use, et cetera. Yeah. And I think, you know, like in thinking about these companies and insecurities, I think the, I don't know what you see, but this is what I see. The companies that are sort of really embracing this idea of AI and allowing it to fuel ideas, school, like mediation, to this point of like going deeper at schools, like getting kids to go deeper.

18:36They're the ones that I see getting ahead where the other ones that are more unsure or insecure about like, oh, well, if they're doing those types of tasks with AI, like, I don't know, like if that means they're working hard or not, right? Like, and then they're more scared of it. So what's your advice for leaders and companies on like, how should they, because I think everybody, nobody's really figured it out. Like everybody's trying to, we're all trying to figure it out. But what should their mission be or where should they start? Yeah, I mean, I think obviously you should use the AI to save as much time as possible.

19:15And that's why you're seeing business move at a faster and faster pace. But you also, as a company, need to stand for something. And so what's been really interesting is you can have ChatGPT write a really good marketing blog post, right? Like, and you give it a little prompt and it spits out a post and like, oh, that's pretty good. But then you need to realize, okay, well, it has just generated that from what it's read on the internet. So there's nothing probably uniquely special about it. Right. And so we use that as the first draft. We never use it as the final draft. It's like, okay, what is interesting about this?

19:47What do we like, dislike, remove, restate? What other ideas can we add to like push it another 20 % better? And so I think it's a really powerful brainstorming partner. But if you just take what it gives you, you're going to be mediocre at everything, right? Like it is by definition, just going to be a mediocre outcome. Because it's giving you that, like the average. Exactly. Yeah. Yeah. That's so interesting. And I don't know. I mean, I love, I love chat GTV. And as like an ideator, right? Like, because, and I sort of in the beginning was like, oh, like I should just automatically have these ideas.

20:23You know, like I kind of felt a little bad, like GTP was smarter than me. No, no. But, but I mean, of course it's smarter than me. It like knows the whole internet and, you know. But I think then when I started to get better at using it was really like going deep, like this idea of like going deeper and getting ideas. And it's like, I had an idea, but then I would ask it some questions that would make my idea a little bit better. And then it would like prompt my brain to go further. Yeah. Because it was like, oh, I didn't think about that. What about that? You know, and then all of a sudden you're on sort of a different, you know, plane than you were when you started.

21:03But I think some of this fear, though, that maybe I had and other people have in the beginning, I think like organizations, and you've probably seen this because they are scared of something or they don't know it fully. They're like, OK, we'll do it later. Like, let's not do that. But I feel like if you're not doing this as a company, I don't care what industry you're in, you're going to get left behind. This is without question true. And like I can speak from firsthand experience seeing the accounting profession because there are stereotypically, right? Everyone's like, oh, accountants are risk averse, old fashioned, move slowly, et cetera, et cetera.

21:40And there's definitely some truths to that in a general sense. But there are a ton of incredibly tech forward, very fast, innovating firms that are adopting this stuff very quickly. And then, of course, the firms that aren't are going to be completely left behind. And so all the hype is like, oh, AI is going to replace you. No, not at all. you're going to be replaced by the people who use the AI, correct? Right. Well, let's talk about that. So because, you know, I was talking to a company the other day and they've been really encouraging their customer service agents to utilize AI more. And, you know, a few of them are, but a lot of them aren't because they're afraid that, you know, it is going to replace them.

22:22So they're like, well, you know, I know I should, but I'm not going to because I'm always going to be better than the AI. Help us understand, like, because I haven't thought about it. I do think that there's a lot of work that will be repurposed. There's still a job for a human. So how do you think about that? Yes, there's a macroeconomic principle called the lump of labor fallacy. And so this actually came out from the late 1800s. If you look, if you think back, like electricity, people were terrified of. Yeah, because it was going to replace jobs. And obviously cars were going to replace jobs and all the people who kept them took care of their horses.

23:03Right. And so this principle lump of labor fallacy is that there is a lump of labor that the world needs to do. And as technology makes it more efficient, there'll just be less jobs because that labor will be automated. Right. Right. It's completely fake. Right. It's never been, has never happened. Correct. It has literally never happened. Right. It's like only created more jobs to my knowledge. And so it is true that the types of jobs do change. Like, yes, there are fewer people relying on horses every day, but it's not like the job just disappeared. They just migrated to new roles. And so there will, I think, be a few years of some structural friction in the job market for sure as things are reshaping.

23:47but imagine five years from now, like it'll be very clear what jobs are still valuable and there'll be tons of those jobs because everything else is now moving faster because the tedium is gone. Yeah, and what do you think about like all these companies say they're laying people off for AI? I just don't think it's true. I think they're using it as an excuse. I completely agree with you and there was data on this actually. Okay. I can't pull it up off hand but there was a graph I saw on Twitter of basically all of the layoffs were actually driven by COVID overhiring. And they are not correlated with AI use in that industry.

24:23Yeah. I mean, and also it's just, I don't know, you and I are sort of closer to this than other folks because, you know, I'm in VC now and we're, you know, we're investing in companies like this, but we're also, I'm a board member and advisor on the side and I'm talking to companies and CHROs and CEOs about how to think about this. But it's not to a point where lots of roles are going to be eliminated. We're just not there yet. And maybe they're a little bit here, a little bit there, but not 1 ,000. Like, we're going to lay off 14 ,000 people, 20 ,000 people. Like, I haven't seen that in a company yet.

25:00What will be interesting is the transition for the junior labor force. And so it is the case right now, if you are coming straight out of college as a computer science major and you are not very good, the job prospects are challenging for sure. My poor nephews. They're coming into the labor market soon. So this is what I get back to, right? But, you know, I'm their aunt, so I'm probably biased. If they're, what we're seeing though, is for folks who have been using the AI to really uplevel their programming skills, they're incredible. You can find quote junior talent that performs at the level of a senior engineer or above.

25:41If they've been using the AI to sort of do their homework and they're not very good engineers, that is challenging. And so then you start to draw other parallels and like what will happen. And you look at plumbing as an industry, right? You can't afford to hire a bad plumber. Well, how does the industry prevent that? They have trade schools. And so it'll be really interesting how the technology industry evolves because you need senior engineers without question. It's how do you bridge people into that effectively so that they can continue. And so you think it's going to be like a trade school type of situation?