A $3.6B founder's honest take on the software market right now

I do. Yes.

Not exactly. So, uh, it was, uh, we took a flyer on that one. They actually, well, the weirder story was they asked for both of us to do it. Somehow I had better judgment. So...

So, um, three of us went to middle school, um, and high school together, and then four of us went to high school together. And we had tried lots of different ideas, uh, throughout middle school and high school. And then finally, um, as we went to college, kind of people split off to different schools. And then this idea kind of emerged and we all kind of got back together on it and then dropped out of college in kind of two parts, uh, in 2005 and, um, 2006. And so we've just been working together for, I mean, honestly, like almost 30 years, uh, on different things, which is kind of crazy to think about at this point. Um, so right now we have, uh, Dylan is, you know, CFO. He runs a bunch of functions in the company. He was famous for Millionaire Matchmaker. Um, apparently, uh, definitely his main claim to fame at this point. And then Jeff and Sam. Jeff has a bit of a farm. Um, he's kind of getting into the farm world. And then Sam is at Anthropic and, uh, on Claude Code. And, um, he now, um, is a constant thorn in my side because every three days somebody says the CTO of Box left Box to go to Anthropic, but he actually retired from Box like six years ago. Um, and it causes like, uh, you know, sort of unending viral fodder, um, on that one.

Instagram. I mean, they have, to their credit, they've done obviously an insane job on recruiting. So, I think it might literally be a requirement to have been a CTO of a public company to work there at this point. But like, they have like this, like, there's like a list of like 10 of these people. You know, Mike's over there obviously now with Andre. Um, you know, uh, Sam, uh, on our end built, you know, some of the most important, uh, software and infrastructure that we run on to this day. So he's, uh, you know, he's obviously a huge asset for them. Um, but you know, they've done an incredible job at recruiting.

Yeah, I mean, it always sounds way more dramatic, you know, when we kind of compress it into the brief story, but it was a multi-month period of kind of like your classic wandering period that all startups kind of deal with, um, where you don't know if you're going to pivot, you don't know if the business model is going to work, you don't know if you're going to get your next round of funding. Like some things are clicking and working, some things aren't. And we had started, um, I would say not even as a consumer or enterprise. We kind of started as agnostic to who the user would be. We just said, hey, you know, there should be a secure way to access your files from anywhere. Like it was an obvious idea to us. And then what happened was it started growing, but consumers, we had this kind of, you know, very straightforward fork in the road. Consumers wanted to pay as little as possible and they wanted a certain set of features that you'd have to go and build. Enterprises wanted to pay a lot more but they would need like a hundred times more features. And as we kind of thought about like, well, who do you focus on? We did eventually kind of conclude, you know, something really obvious in retrospective. And I can't like unsee it as I look at other people's strategies over time. These were just like totally different markets. Like the enterprise needing to securely manage their most important data as an organization would require just a completely different set of functionality than what a consumer would need to back up their photos and access their music from anywhere. And those were just different markets. And there was different business models. One would pay like $5 a month, the other would pay maybe $5 million a year. And like completely different business models, different markets, different teams you'd have to build, different products you'd create. So we did eventually run into, you know, effectively a fork in the road. And after a few months of kind of debating it out and assessing the opportunities and, you know, people kind of having, you know, pretty different views on what to do, we eventually pivoted, um, you know, kind of very forcefully into the enterprise. We almost burned every boat other than a couple that we still wanted a premium model in the enterprise. We wanted you to be able to sign up as a kind of knowledge worker, but we very firmly wanted to make it an enterprise-only business model. And I was actually the most reluctant and the last one to be convinced to pivot. Um, uh, so kind of credit to the other founders and some early employees that I think had more conviction. But once we collectively had conviction, then it was very straightforward.

Well, it was definitely the right move for us and where we were. You know, the way we kind of did the math was Google had to own the consumer. You know, we saw the G Drive kind of writing on the wall. They would want to bundle it with Gmail and most consumers would sort of be satisfied with that. And then iCloud had added on top of that and then OneDrive. And so it like the consumer really looked like a total kind of death pit. And, um, uh, Dropbox, I would say, performed far better than I would have estimated if from just a pure like economic standpoint. I would have thought that the commoditization would have been much more impactful. So huge kudos to them on their execution and just, you know, obviously building a world-class product on that front. What we were very clear on was the only way that we would not go out of business was by being enterprise because we just were, we had, we were too convinced that over enough time that the consumer space would just be too competitive and too commoditized. So, uh, not only are we, you know, fantastically happy with the decision, but I think it was the only outcome that would have produced really any form of success. And then I think over the long term, like let's just say we add another 10 years to the timeline. I think the only way to build a very large business as an independent company in this category is by being enterprise-focused, just because like, you know, where most dollars going to go for managing data, securing data, uh, you know, kind of caring about how it's governed in a workflow, it's going to come from businesses. And so, and there's actually, you know, I think relevant lessons as we look at the AI space is I think most dollars in AI will eventually be enterprise dollars. There'll be some, you know, there'll be fantastic outcomes in consumer, no question, because there's, you know, some ways to build consumer businesses out of this. But by and large, where's intelligence valued? It's going to be in the enterprise. So just as where is software valued, it's in the enterprise. Um, and that's where most dollars of technology go. And other than, you know, three companies that make money on advertising in consumer tech.

Caller. Yes, that was the person to be in the valley. I mean, other than, you know, Zuck and...

And it's the person to be again 20 years later.

Uh, sure. Yeah. Well, when we first started, we dropped out of college. It was four of us living and working in Berkeley and we got a call from Yahoo. Um, and it was the corp dev team at Yahoo that had basically the team that had more or less just been responsible for buying Flickr. So there was this product that was like in the late 90s, early 2000s called Yahoo Briefcase. And it was sort of one of our predecessors. So it was an online storage, you know, kind of product, but it was like you could store maybe like 50 megabytes of data in Yahoo Briefcase. And for us, you know, we had finally achieved a gigabyte of storage that you could have online. And we were like the modern, simpler, faster, easier, you know, kind of more up-to-date version of Yahoo Briefcase. And we got called in by the Corp Dev team. And for us, we were like, 'Holy, you know, [__], this is the biggest moment of our lives.' And we were like debating like what acquisition price would we be willing to take. And I think like probably the most we could have ever imagined was like $5 or $10 million. And we were like, you know, that's our price. We would definitely take five million, you know.

I think we didn't even have enough. Uh, we probably felt we would jinx it if we even did that. So it was more like a very serious discussion. So we drove down to Yahoo corporate headquarters in a Nissan minivan that was like totally breaking, you know, falling apart. And we did this serious meeting. We presented our whole strategy and, you know, we went through the product. And I don't know, I don't remember, it's very hazy, but like I don't, somewhere on the order of probably two weeks later, we just got like an email saying it was really nice meeting you guys. Thank you for coming by. And we had done all this buildup in our heads of like what would the number be that we would sell the company for. And again, any of those numbers we would have been just ecstatic about taking. So it's one of these things where it's like, you know, we have turned down offers, but we've also been in situations where we totally would have taken, you know, that very early offer and just taken it off the table. And then, you know, later as we scaled, because we've had every problem thrown at us. We've had rounds that didn't happen and just like totally busted rounds. Uh, we've had to be bridge loaned by our investors twice. So there are definitely, you know, there's definitely parts of the journey where we would have, you know, if anybody had shown up with any offer, we would have accepted it probably. And then it's, you know, as these things go, like when people do actually show up for offers, you get, you like, your chemicals in your head are totally different and you're like, 'Oh my gosh, like we just got to keep doing this.' And so probably the most classic one that we faced was a very kind of serious interaction where we would have, you know, been, I think, quite happy about the outcome on any kind of financial measure. But we looked at the situation and we were maybe our mid-20s, early to mid-20s at the time. And I think this is now like well documented by a bunch of people, but I think it kind of just happens probably pretty uniformly, which is like you just, you like, if you really deeply process it and like in a very intellectual sense, and you're like, 'Okay, this much money, like this is super interesting, you know, etc.' And then you like start to play out like what am I going to do in two years from now or five years from now or 10 years from now. We basically just processed like we would probably be doing something just to get back to exactly where we are now. Like there's, it's unlikely that we're going to work at this new company for more than five years each. Like so that's not going to happen because every one of our friends that had gotten acquired had already left their acquired company. So like that was probably not going to happen. So then you just look at it, you're like, 'Okay, well, you're probably trying to do everything you can just to get back to this exact situation.' But of course, you have more cash. That's obviously positive, but now we're in this situation. We've already defied all the odds of getting here in the first place and all the things that kind of got in our way. Like why don't we just continue to double down on this given that we still believe the market is still a hundred times larger? So it became this very kind of calculated decision, which is what's the amount of, kind of, you know, how big is this market still ahead of us? We thought very large. Uh, do we like our kind of compounding approach where we think we're getting better every day, every week, every month at our product and our strategy? We know that there's going to be a lot of headwinds and a lot of severe competitive pressure that we're going to face. So it's not going to be easy. And then it basically just came down to like, you know, the kind of Bayesian regret minimization framework of like what thing are we going to regret more or less. And we at least convinced ourselves that we would more regret not continuing and just seeing the next set of cards and keeping on scaling more than we would regret, you know, sort of turning down this offer and having to start over. I don't know if that's actually true, like what would we have really regretted more, but that was the decision. It was gut-wrenching. Like we did an offsite with the four of us. Uh, and...

We were in our mid-20s, so probably two of us were like 25, 24, 23. And we don't really talk specifically about the offer, but, you know, call it like in the half a billion range.

Not safety net levels of secondary. So, um, this was a very different time period in the valley. So this was, you know, very early 2010s and secondary was not, neither in fashion as much as it is now nor were the amounts of capital the same level. So I think it was like, you know, people could, you know, feel better about the apartment they were renting. Um, as opposed to like we're like good on the decision.

I honestly, if there's like any analogy that works for my entrepreneurial life, it would be that. So, um, uh, but, um, we have had to grind through every worst practice that you could imagine. And, uh, I mean, we've lived to tell the tale. So...

That might be, there might be some hallucination on two of those. Um...

I think there's some good embedding space clustering of some of those brands. Um, unfortunately, so I met Dylan Field in the seed round and, um, uh, what a lovely character and kid. Um, uh...

I yahooed him. Uh, I think I was, I hope I was like 3% better at the follow-ups, but, um, I met him and I did not have the creative imagination for what he was talking about and I obviously should have because I believe in cloud-based software for everything. And he was like, designers are going to do real-time collaboration on stuff. And I was like, I don't know, man, like we kind of make images fine. So I was like very, like I was a huge laggard on the pitch. Um, and definitely to my detriment, but unfortunately that would be a hallucination. So I didn't get on Figma early enough, but a few of those. Yes.

Uh, I fortunately, at least for other factors, Box is still ahead. But, um, the big thing I should have done is hedged on all of our underlying suppliers. Um, we're like one of the biggest customers of like Seagate and Western Digital. And, uh, and so like we could see the stack that you would need for all of this. And, uh, and it's, I don't know if you guys have watched like the SanDisk stock, but like this is just the most insane, um, you know, it's just set...

SanDisk is up about what, 3,000% maybe in the past two years and stuff.

Uh, they made everything and I mean, we probably wouldn't exist without SanDisk. So USB thumb drives were like one of the catalysts for, okay, we should just like move that to the cloud. Um, so if you had just bought SanDisk stock like on just because you were really into USB thumb drives, you would be doing fantastically well right now.

I mean, way more than I've invested in. Uh, but you know, I think if you just looked at probably even our own tech stack over 20 years and you just bought the stocks of what our tech stack represented, like that portfolio alone would be, you know, you would have every index.

Yeah, it's interesting. The funny thing is like this data is basically out there for every investor. And, um, and I do think that it's probably not kind of leveraged enough, but yeah, I mean, most of the best practices are just well known by engineers very quickly.

You go from 100 billion to a trillion...

Yeah, you're like more likely and you get there faster. It's like, oh...

I, you know, some of these things though are a little bit tough simply because we are in a, you know, we're in a pretty kind of feverish environment. So it's always hard to like, how do you normalize for the particular multiple that we're seeing and is that like a sustainable investment strategy versus right now we're in a moment where that is kind of working when you look at it backwards. But yeah, that was definitely counterintuitive when I saw it.

You know, so much of the idea that jobs go away or that we do less, uh, you know, has to come from a place of effectively a short of human creativity and ingenuity and the idea that there's sort of, we don't have an insatiable appetite for more and new things. And I've just seen very limited evidence that suggests that we don't want to go and discover the next cure for the next, you know, niche problem that people have or the next new form of entertainment people want to experience or the next new consumer product that people want to go and sell or the next new podcast that wants to be created. And so like if you don't believe that that's going to happen, then sure, then you would basically believe we have that abundance sort of comes at the expense of jobs and of us, you know, doing things. And so that theory has to be that basically, you know, the agents are going to do all of effectively the useful work, which then frees us up, you know, so much that there's really not much else left. And I just think we will find a way to create a ton more work for ourselves for better or worse. Like it's not obvious that my view is particularly like utopian. Like I think to some extent, Elon's vision, to his credit, is actually a far more utopian one than the doomers. It's actually funny because they believe in the same underlying technology trend. So they basically both believe in, if you extrapolate out to AGI, then the doomers believe that destroys us and Elon believes that we get this utopia where everything is just done for us. And I kind of am more just until proven otherwise, I'm just in a third camp, which is like it's like the same progress of maybe both of those two, but with more of a pragmatic outcome, which is like we use that technology to just then create a new set of needs that we have to go and all kind of support and fulfill, like a very long list of things that the world is still going to effectively value. Like the world is going to still value in-person education for children. They're going to value child care. They're going to value going to a restaurant and having, you know, human interaction. They're going to value going to a show. They're going to value like talking to a financial adviser that appears to sort of, you know, have some sense of the market and your set of needs and also has 10 other clients that they can kind of like triangulate with. Or, you know, a tax professional that you can just like, you know, is like accountable for if they get the thing wrong, like their job's on the line versus like an agent that just can be just shut off and so thus you're not really sure what accountability they have. So like for all these reasons, like humans just stay in the loop and so I just think we end up having still a lot more work for everybody to do. I would say then the four-hour, the four-day work week thing also kind of presupposes something different, which is you basically have to believe that assuming that anybody in your sector decides not to do a four-day work week, then that company, you know, with the power of AI will get 20% more or 25% more output than you will. And so which market is going to basically have some kind of like collective agreement that says, no, our category, everybody in our industry is only going to work four days a week. So it's just like it requires such a collective sort of like agreement on the part of everybody that you wouldn't then just like have some actor in the system decide, no, I'm just going to like, I will just ship more software, I will sell to more customers than you do, which then gets everything back to five days a week. So that's why it's like just very implausible for that outcome to really exist.

They're way busier than we were before AI. And the reason for that is because AI is sort of like this deceptive technology because it lets you get started on so many things so easily, but then you still have to complete all the things you started. And so you think that I'm just going to deploy all these agents and then I'm going to go to the bar or go hang out. But when the agents are then done, somebody still has to be responsible for what do I do next with that information? What do I do next with that piece of software? What do I do next with that video clip that got created? All of that becomes human work again. So I think every single person that is the most AI-pilled right now is just drowning in work because we're kicking off way more work for ourselves and we can't ever get off that treadmill because of how easy it has become to just create this work. Like I don't know, like an hour before this call I kicked off two processes that I didn't even need to start, but I started them and now I'm going to absolutely add another hour to my day because I'm going to go and do whatever the agent produced, I'm going to go and follow up with all that work. And I didn't even have to, but it was so easy to kick it off that now I've created more work for myself. So we're just going to do that for everything.

Uh yeah, I mean if we want to run with that, we can name it.