[00:00:00] Thanks for coming on. We're able to make this work.
Drew: It's great to be here
Logan: So we, we, uh, one of the topic is yours. I don't know. Uh, your perspective on this is artificial intelligence these days. And I think you sit in an interesting purview, both as a, uh, the CEO founder of Dropbox, but then also you fairly recently joined Facebook's board.
Drew: Yep. About five years ago, a little under five years
Logan: Okay. So, so not that reason, but you've probably [00:01:00] seen, I mean, this has been a, an interesting, was it Facebook at the time or
Drew: it was Facebook at the time. And then they, uh, changed the name a couple of years in.
Logan: So you've seen a lot of the, um, AI investment that that business has gone through. I'm curious, like, how has that seat, that purview informed your perspective on the AI opportunity for Dropbox?
Drew: Well, it's been great to watch. Um, and I won't speak for meta itself, but certainly like seeing the rise or seeing the power of open source and, and the impact it's had on democratizing AI has been really powerful. Um, you know, whenever there's like a new era of computing, there's a question of, you know, is it gonna be open?
Is it gonna be closed? Um, and so open source has been super powerful, clearly in terms of making AI more accessible and in terms of both being able to use it, build products with it, research. I think on balance it helps with things like safety because you can be, you know, you have a lot more people working on that problem versus a situation where, [00:02:00] you know, one, a couple companies are the gatekeepers to everything.
Um, I think there was definitely an alternate future where we'd all be kind of like paying rent to a couple of big companies, but, um, instead we've seen the price performance of these AI models, um, plummet, you know, like drop 10, a hundred times or a hundred, 10 or a hundred times better price performance every year.
Um, and that's for a product developer like me or like Dropbox. That means like what we can build is a lot more exciting and more affordable and things that would have been kind of priced out as being too. expensive to do that much inference at our scale suddenly become viable. So it's been awesome to have that vantage point
Logan: Did you have a moment in time that you can remember that was like, okay, this is gonna, I clearly need to focus on this for, for Dropbox business going forward?
Drew: For me personally, it was, it started to heat up. I mean, I started playing with more classical machine learning several years, you know, 20 mid 2010s. Um, cause I didn't, I wanted to go to, or at [00:03:00] least thought about going to grad school, did my undergrad in computer science, but never did like the grad level courses.
So I try to self teach classical machine learning to see what that was about. Uh, and then certainly as, um, deep learning, large language models came on the scene, then there was like that thunderclap, you know, iPhone launch moment of chat GPT. And when you, when they, uh, had these instruction tuned models and like GPT three, then it was a little embarrassing.
I was like on my honeymoon, like coding and playing with these things, um, where I was like, oh my God, it's happening. Like all this stuff I wanted to build. And, you know, back when I first started studying this stuff is like, I hit all these roadblocks because like computers basically couldn't understand text, but then the large language model is like, not only does it understand text, it can write text, it can write you JavaScript and then write a sonnet about the JavaScript.
I'm like, wow, a lot just happened in like one, one minute. Oh, it seemed
Logan: That would be a very unusual person that could do all of those things, which is, which is powerful to the AI
Drew: So it was a big wake up call.
Logan: Is there something that you look at [00:04:00] now, uh, both from your, your seat as, uh, CEO of Dropbox, as well as, you know, getting this purview into how Meta is making these investments that you sort of think is an inevitability that maybe we're not thinking about?
thinking about if you pick a point, I don't know, 10 years in the future, that sort of feels to you like innately, obviously this is gonna, this is gonna happen, but might not be, you know.
Drew: Yeah, I mean, the thing I'm most excited about is we sort of have this, we have our human intelligence or our human brain and then there's this new silicon intelligence or silicon brain. Um, and just as you saw in computing, we're adding a GPU to the CPU suddenly made all these new things possible like AI.
Um, in our working lives, it's, we're gonna have this really kind of wild partnership with our, like, as we plug in the silicon piece of our brain. Um, And so I think a lot of work will be reimagined, um, where we'll be able to offload a lot of our busy work and be freed up, um, for more like creative or [00:05:00] relational tasks.
You know, I don't think anybody can really predict exactly what this, the precise route will be through this, through the fog here, but I think we'll be able to have, or I think we're opening the door to being, you know, where anybody can be a 10 X person and people that really figure out how to master These tools can be a hundred X people.
Um, so I think it will be certainly the most transformative change in our
Logan: Have you played with any of the physical devices like the, the pendants or any of those, those things? Do you think?
Drew: not a lot. I mean, I'm very excited about voice or the ability to have sort of like an infinite memory and all these different contexts as you're using your computer or
Logan: Yeah, it needs to become more socially acceptable. I think a lot of those things are very, it's very cool, but, uh, there's, there's some social, uh, element of adoption that exists in the real
Drew: Yeah.
And I mean, even there's some, Basic, like, legal restrictions. Like, you're not allowed to just go around, like, taping everybody.
Logan: party consent and all that stuff. Yeah. No, no. Dropbox has some AI principles.
Drew: We [00:06:00] do.
Logan: And what are those?
Drew: So we established, um, pretty basic principles around, uh, transparency around how we use AI. Things like privacy, safety, um, kind of all the, all the, uh, uh, setting forth all the things that Dropbox will and won't do.
Um, because I mean for a number of reasons, but one is, you know, customers, and I'd say all of us are really excited about the good parts of AI, pretty concerned about the things that can go wrong. Um, and then when you intersect that with like your stuff or your like most important information, your personal information, your company's information, you care a lot about making sure that you can.
Trust your counterparties or the services that you're using. Um, and we had some experiences around this with Dropbox 1. 0 because people for the first time, you know, or a lot of people for the first time were putting their most important information into the cloud. Um, and this happens a lot, you know, whether putting your money in a bank instead of under your mattress or putting your credit card over the internet or [00:07:00] putting your files in the cloud.
People start out with a lot of the same apprehension that's very justifiable. And so this was a way for us just to be out in front of that, say, here's how we'll use your data, here's how we won't, and reinforce one of Dropbox's big advantages, which is that our incentives are much more fundamentally aligned with our customers from a trust perspective, because Um, I think a lot of people worry about, when you think about how these models are trained, it's like, oh man, is this service going to like take all my stuff and sort of grind it into little pellets to either like sell me ads or to train their next foundation model or, or have others some kind of, you know, undisclosed use of my information.
We wanted to draw a bright line and say, no, we're not, you know, you could, I think everybody's looking for providers they can trust. And so setting these principles affirmatively. was a step we took pretty early on. And a lot of companies are, have also published similar principles.
Logan: so for folks that don't know, what, what AI products have you released, uh, to date and what, what are some of the things you're excited about for the [00:08:00] potential going forward?
Drew: Yeah. The most notable one is we have a new products called a new product called Dropbox Dash, um, which is really centered around universal search. So the basic first basic problem we're solving is, you know, I can't, I know that thing exists, can't find it. Um, and this question of like, why do we live in a world where it's easier to search all of human knowledge with a Google search and then when I go to search for my company stuff or search my company stuff, search my own information, you know, I got 10 search boxes.
Um, and it's a much worse experience. And then weirdly, despite all this technological progress, this problem is like much worse today than it was 20 years ago. Um, 20 years ago, if you wanted to find your stuff, you'd just like search your hard drive, right, or your email. So maybe one or two search boxes, but now we've had this, it's kind of gone wild.
So Dash is really about fixing that, giving you one search box that can search everything. So it'll search your Google Docs, your Slack, your email, your files, uh, but really anything. You don't even need files in Dropbox. Um, so this is completely, it can [00:09:00] be a completely standalone thing or integrate with your Dropbox.
Um, But we see that as like a very fundamental challenge that, that a lot of knowledge workers have. And especially as we've moved into these like distributed ways of working, like where's the information I need to do my job. Um, and all the paper cuts that come along with that.
Logan: How, by the way, how is that powered or enhanced by AI today? Enterprise search is a market that's been around for a long time and it sounds like this is both consumer and B2B elements of it,
Drew: We're focused on use case work use cases. Yeah. Um, so it's, on the one hand, enterprise search has been around for a long time. On the other hand, sort of similar to, Cloud storage. Like when we started, there were a lot of other things that sort of claimed to do it, but you know, in our view, none of them had really done it right.
And I think there's a similar dynamic that we've seen with how, how enterprise search has evolved where, yeah, it exists, but no one really, when you ask the average person, like, do you use any of these enterprise search products? They're like, no. [00:10:00] And you know, have you used them? Yeah. It didn't really work very well.
Um, and so certainly the, um, generative AI allows you to not just do search, but really get answers. Um, and so DASH is also, we also view DASH as for a lot of the questions that ChatGPT can answer, because it's not connected to your stuff. So if you ask, you know, if you or I ask ChatGPT a question, we'll largely get the same answer.
But then if I want to ask, like, when does my lease expire, when does Where's the slide from last year's product launch, things like that. Um, you need a product like dash, which is connected to all your stuff and can, is grounded in all of your content.
Logan: And so, so is the search opportunity then, uh, and forgive me if this is a little nerdy, but it more enabled in the natural language of how you can ask and then how you can, I guess, is there anything on the back end that's more intelligently done through large language models or whatever that actually surfaces this stuff up?
Drew: absolutely. So with search, um, what was a couple aspects I'd say when you sort of like double click on this [00:11:00] specific problem, there's. People often have like retrieval use cases where it's like, I know there's a thing or I'm looking for a specific asset and I actually just want the thing. I actually don't want to have a question.
I want to answer, answered or I want an answer to a question like help. I'm really at the end of the day looking for, um, okay, what is, What's DASH's roadmap, or when's the release date, or what are the key features, things like that. So it's a question. And those, the blue links kind of format is actually pretty good for retrieval.
You just want to get the thing quickly. But then when you're discovering things, or you don't know the answer, don't know where it might be, then a natural language fits in a lot better. And then there are also, like some of the underpinning technologies of large language models, Um, these embeddings are a lot of semantic search, neural search, fuzzy search, vector search, a lot of names for it, same thing.
But basically, you, one of the unlocks was you can now search without having to match the exact keyword. So if you say like, oh, what's the 2025 [00:12:00] strategy or the 2025 plan, these, uh, vector search can help turn up the right results for both
Logan: The things within the proximity that are close to it rather than it could be, you could say 2025 plan and if it was only language based, it might show nothing if it was actually called whatever company estimates through 2030. Uh, but it can get directionally there now.
Drew: And you know, when you implement these things properly, they, these can also be learning systems where they're all, they're always getting smarter based on how you're becoming personalized to you as you use the service and then just smarter overall as the world is using the surface.
Logan: I, I guess at a personal level, are there things you're using AI for to make yourself more productive or regular behaviors that you've kind of adjusted?
Drew: Yeah. Um, Lots. I mean, you know, my first love is the engineering. Um, and, uh, and I grew up as a little kid coding. And so when the large language model. came along, you know, around the chat GPT timeframe, I was suddenly like [00:13:00] coding like an 18 year old again. And I'd say even before that, when I was like Trenton learning, classical machine learning or the kind of the pre deep learning techniques, um, there was lots of stuff that I wanted to automate.
And I, you know, even my career, I started out as an engineer. I knew I wanted to be a startup founders. Wasn't actually sure I wanted to be a CEO, sort of backed into it. Um, but as I really inhabited that role and this, the company really was really scaling. I was like, man, there's a lot of tedious stuff that managers or executives do.
Um, so yeah, I would have all these little tools to, you know, audit my calendar or figure out which of these emails needs a response or, um, actually the motivation for Dash from my perspective was like, I can't find my stuff. I need to, we need a better, I need a better way of managing this personally from, forget a product.
And then I wrote it and I was like, oh, and. There's these embedding based, or you know, these vector search is a new thing, and I built this little prototype of a personal search engine. And I'm like, oh my god, it works. It's super scalable. It's super [00:14:00] fast. Everybody's going to be using something like this in a few years.
This is maybe 2018, 2019. And I'm like, we should be all over this. Um, so actually this is a lot of, I still write many thousands of lines of code. Um, a year. Uh, and that's like one of the key ways that I really get more of a tactile feel for like, here's what the technology is, here's what it can do, here's what it can't, and this stuff's fully baked, this stuff isn't.
And, um, yeah, I mean I think, it wasn't just my prototyping that led us to Dash, but um, that's really what caused me to have a lot of conviction pretty early and put a lot of chips on the table.
Logan: AI, removing AI from the equation. Is there anything you've done from a productivity, either calendar tracking or email triage or just anything you would recommend that people that are busy and managing, you know, different constituents and all that, anything you've done that you reflect on even outside of AI?
Drew: Yeah, there's, there's a lot of good stuff and I'd say the, the best stuff is really timeless. Um, and, uh, often the principles are [00:15:00] pretty straightforward to describe. And the ones I lean on are, some of my favorite books are, um, High Open Management by Andy Grove. I guess one of the best books on management ever written.
Um, another is The Effective Executive by Peter Drucker. Thanks for watching! Um, I could give a lot more, but, but just on the sort of productivity and effectiveness, uh, angle, I think those two probably get you 95 percent of what you need to know from a theory perspective. But what's really hard about it is applying it, right?
Because it's sort of easy to read the book and then, you know, it's probably a super disappointing fraction of people actually apply the lessons. Um, and it's just kind of this constant fight. So, um, I think, A lot of my use of AI was really to automate some of these things. So for example, um, one thing that every, every busy person struggles with, like, is how am I using my time?
And, uh, the first chapter after the introduction of effective executive, the title is no that time. So what you do is you basically do a time [00:16:00] audit and you're like, okay, here's, um, here's all the stuff in my calendar and I'm going to categorize it, put in these buckets and, and then you line it up to your priorities.
And so. Um, you know, I read the book a long time ago, but 15 years ago, I probably did my first time audit with my admin at the time. And I was like, Oh, I probably spend most of my time on recruiting, uh, and on products. I don't know. And then after the audit came back, we tallied it up and we're like, actually, I spend my time on just about everything except for recruiting and product.
So it sort of highlights this. Um, it's one of these areas where, you Um, it's not just that we don't, we, it's actually worse than not knowing where your time goes. You actually, where it actually goes and where you think it goes are often like wildly disconnected and that's like one of the first findings in the book is like people are usually shocked.
By a time audit. And so, you know, it's something you have to do periodically, but it's a really manual process. And so it's something that lends itself well to automation, but yeah, all kinds of things like that. Um, how am I using my time? Am I focusing? Um, how [00:17:00] am I, how am I keeping on top of my communication?
Um, personal search. And yeah, there's a lot of, there's a very, there's a lot of little code repositories in my, my little bag of tricks.
Logan: You've, I've heard you've moved to like a memo first culture internally. I guess, what was the shift to doing that? Was that something you guys did in the early
Drew: Yeah, it was something we did in the early days. Um, and, but I would say it was really reinforced after we started or after COVID. And particularly after we shifted to, you being about 90 percent remote and really kind of rethinking our working model of being a distributed company and sort of navigating the new world.
Um, so I think it became much more important than in the origin or what really emphasize it is. Um, so Jeff Bezos said of all the decisions he made when scaling Amazon, you need to be like the single best decision. actually had nothing to do with like his products or his markets or any of the other [00:18:00] things that where they've been successful.
That's what I would have thought he would have said something about that. But he's like, no, the best decision I ever made was to ban PowerPoint and the company and switch to this narrative based culture. Um, for like a bunch of reasons. I mean, one is, um, uh, well, the reason why it was is as helpful as because like we've all been in those, those meetings where someone's presenting.
Um, you've already read the slide, but you have to wait for the person, you know, verbally to catch up. Um, or, you know, someone interrupts and asks a question, but the question is answered a couple slides later. Um, and, and one way or another, the presentation format kind of limits how many people you can have in a room to have an effective conversation.
It's, um, the, how compelled you are by the, you know, Presentation is less about the merits of their arguments and more about their skill in presenting and thinking on their feet. So there's a bit of a distortion or bias that comes in that way. And then conversely, um, narratives [00:19:00] are, they work really well because they, they have a much higher bit rate.
So most people can read a lot faster than they can hear. And then when you not just write the narrative, but spend the first, you know, 10, 20 minutes of a meeting reading it, which is part of Amazon's practice, something that we also do. It's like, then everybody has, instead of like some vague, like very vague and disjoint sets of context, suddenly everybody has the same kind of high def 4k view of like, what is, like, what are we talking about?
What is the plan? Um, and you know, everybody has that kind of fact based and it's super, it's super efficient to download that to. You know, you can have a like 15 or 20 people in a room and have it still be efficient. I mean, the challenge with it is that there's a couple issues. One is, um, it's very time consuming for the writer.
Um, but it, or it is a trade off. It's very hard for the writer, easier for the reader, but that's generally a good trade off, especially in a bigger company. 'cause there's a lot more readers and writers and often we're communicating something. It's a lot more efficient to just. Give people the source of [00:20:00] truth rather than verbally sort of game of telephone, broadcasting it over and over again.
And then it can also become sort of a proxying that can happen where people sort of forget that the point of the narrative is help get the work done. The point of the, the point is not to write the narrative itself. So if you have all these, like, and I think you've seen this with some of what Amazon's reported recent recently, like, yeah, people are having the meeting, the pre meeting for the meeting.
And, you know, and there's a lot of bureaucracy around the document. Then you can lose the plot that way.
Logan: Once the, once the people have done the readout, then is it, is it open Q and A for people to ask? Or is there generally, do you follow the, the points through for discussion at each point along the way? Cause that's one of the things I've seen is you read it and then it's like any questions and then it ends up being a little too open ended.
Drew: Well, I think there's different ways to do it. Um, we certainly start by doing the pre read. And then while we're doing the pre read, Um, I mean, we do it in Dropbox Paper, which is sort of our document tool. Um, [00:21:00] where, But you can do this in Google Docs or Notion or anywhere else. People leave comments in line.
Um, and so, um, you both want sort of the fine grained feedback on like, Okay, this specific little thing. But then you also want people to zoom out and be like, Alright, what's your overall reaction to this? And, um, And we've even had things where like you have a little form where at the end of the pre read you put your overall reactions and like a rating or things like that, which some other companies have done.
So I think all those things can work, but I think the biggest win is that, you know, you can get, uh, that 10 or 20 minutes, um, you can get a much more, much higher fidelity. Download to everybody in the room. Then you could get in an hour of just talking about it, interrupting each other. And then as far as the, how the, you know, do to start at a high level or low level, do you, you know, I think Amazon does care a lot about, they're like, they ask everybody in the room to go around and give their reactions.
And Jeff goes last. I think some things like that can be helpful to not bias people. [00:22:00] Um, So I think that, but I think there's like multiple ways to do the review. I think that just the key thing is like. Sort of doing your meetings through, through memos or narratives versus just conversation or presentations.
Logan: you reflect 17 years into the journey now, uh, I guess it's still not at the point of more time with Dropbox than not, but you know, at some point that I'll cross in the next couple of years. Right. Um, I'm curious if you could go back and tell drew from 2007 or 2010. Uh, something about, I don't know, leading a business or growing, or is there something you reflect on and you wish you could impart to someone going through this journey that you've now been on?
Drew: Totally. I think the first thing would be something like, it's a journey and just like hold on.
Logan: Um,
Drew: In that, and then more specifically, Becoming a CEO or becoming a leader is something that is more learned than innate. Cause I'd had sort of a picture in [00:23:00] my head, um, of like, oh, CEOs are like this, or they, you know, come out of the womb in an Armani suit with perfect hair or something.
And I was like an engineer and I hadn't really ever been a business person. All I could see is like, well, what those CEOs look like and the things they know. I don't know a lot of those things. I don't have that, seem to have that skill set. So. Um, you know, as a first time founder, technical founder, I was like, probably like a lot of others, like insecure about, you know, am I going to be able to figure this out?
Because always the more I know, the more I realize I don't know, so it's intimidating. Um, so I think the advice would be, you know, just understanding that's part of the, it's part of the game. All the best, you know, a lot of the best tech founders started out as engineers and learned the business part on the job.
Um, very few people went the other way. So it's a learnable thing. Um, but it doesn't mean it happens by itself. Like you really have to be systematic. The advice part that I would send myself now is like, you have to be systematic about training yourself because as CEO, you know, to be effective, you really do need to, so it's very helpful if you know a [00:24:00] lot about a lot, um, and like going, and there's an 80, 20 in each discipline, but there's like a lot of disciplines and a lot of skills you need.
Um, so yeah, So the way I've always broken that down is, um, I've got to keep my personal growth curve ahead of the company's growth curve. And in practice, what that means is think about, well, what is really thinking about, what do I need map out? Like, what do I need to know, you know, one year from now, two years from now, five years from now, and what will I wish I had been learning today to kind of get ready for that?
Because, and so, you know, 2007, the company, We didn't really have, we didn't really have much of a company. There's sort of like, there's just like coding, uh, or my to do list would have been like, let's build a prototype. Yeah. We're going to need to hire a couple of people. Okay. If we hire people, we're going to need some venture funding, et cetera.
Then like two years out, it might be more like, well, we're going to have a product, but we're going to need to distribute, we're going to need, um, distribute to get users. So we're going to need to do marketing or figure out distribution. We're gonna have to have a business model. [00:25:00] And then five years would be like, all right, how do I scale an organization?
How do I be a leader manager? How do I tangle with all these big platform companies who are all going to come after us? So first thing you know is like, those are all pretty different things. And so you have to work back to like, all right, well, what, you know, what do I need to start learning today? Um, And some of those things you can learn pretty quickly, like, all right, what are the mechanics of raising venture funding?
It's, there's a lot of technical details, but conceptually it's not that complicated. Um, but then something like, how do I become a great leader or a great manager, manager like that's a much longer, I have to get comfortable with public speaking like any of these things. Like it just takes a lot longer.
And I think people psych themselves out because they're like, Oh, I, you know, I didn't, I didn't. I'm not good at this. I'm not good at public speaking, so I'm not even going to try. Or it's intimidating to be a CEO. And so I'm like, I'm not going to try, but I think people sort of, yeah, you, in five weeks, you're probably not going to learn to be a great CEO and any more than you'd be learned to be a great guitar player or a great doctor or something.
But in five years you can go a long way. And so I think that [00:26:00] mindset of like everything's like learnable, um, And the best founders, like, and CEOs, like, end up learning on the job. You know, no one's born a CEO. I think that would have given me some comfort and some, you know, pointed me in the right
Logan: Were there, were there good resources? I know you're a big book reader. And so how much did that help you versus just simply doing and making mistakes versus turning to people for advice that have been down, which was the most useful for
Drew: Yeah, by far reading is the single, for me personally, I mean, people, different people learn different ways. But I, but I think reading is by far the most efficient. Cause like most of the lessons about. Leadership or any of the, or, you know, any of the disciplines, you know, it's a lot more efficient to like learn from other people's stories and mistakes and try to do them everything by trial and error.
Um, but reading for sure. I mean, even before Dropbox, like my first company was like an online SAT prep company, but like I would, I was in college and I would just like search for like sales on Amazon, buy a bunch of sales books, [00:27:00] search marketing on Amazon, buy a bunch of marketing books. And then like that was how I just sort of started getting the initial scaffolding.
Logan: By the way, is it true? You and your wife both got perfect SATs on both 1600 Is that true? I, I, I had heard maybe two of the only people in the U S uh, but you guys didn't meet over
Drew: We did not meet over SAT prep.
Logan: but probably bonded over it.
Drew: uh, but after our first date, I Googled my wife and I was like, no way.
Logan: There's an article from some like Texas periodical. Yeah.
Drew: found it. Um, so mine, hers was a little, Cause like the, she was like right on the bubble of the test went from like 1600 points to 2, 400
Logan: probably the same age. That was me. Uh, I was like, right, right in that.
Drew: And so she had to take both.
Um, she got perfect score both
Logan: I did not.
Drew: Mine was a little more, uh, business minded, right? When I was applying to college, I did, I actually wrote all this software to like flash, like to drill the [00:28:00] vocab. And then, um, that, that helped, to, to, I mean, I got a perfect score on, and when I was in high school and then to sort of, uh, to get ready for the SAT prep company, which was actually, I started because the test was changing and made all those like 800 page books obsolete.
And at the time there weren't like great, like adaptive online learning, online courses. So that was the opportunity we were going after. So I went, you know, as a 21 year old, To Brookline High
Logan: Went back and took
Drew: in my car and Starbucks and
Logan: Madison
Drew: Yeah, it, it was a hundred percent.
Yeah. Um,
Logan: at Brookline were like, what is going on here? If only they knew three years later, you would have started Dropbox. Um, that, that, that journey and all, all those lessons, um, about like, uh, I'm sure everyone has this element of imposter syndrome, I guess, as you think back on the, the 17 years that you've been at it, do you, do you reflect on.
Each moment with similar. [00:29:00] Nostalgia of like, this was, this was a fun part of the ride. Are there particular points that really stand out either to the positive or to the, to the negative on the journey?
Drew: Yeah. Um, you know, I started Dropbox so soon after. College that it was sort of that all of that was just this kind of continuous blur where, um, uh, I mean, so I loved all the stages of Dropbox. Um, I don't know if I loved all of them. I think they were all certainly the early years were magical in that, you know, you move out to California, a couple of, you know, kids fresh out of the dorms or whatever.
Um, and you know, you don't even know what you don't know. And you're just trying to build something cool to then you catch lightning in a bottle and have that kind of hyper growth or. You know, dot com experience, um, that I thought I had missed because I, I, I was still in high school during the [00:30:00] dot com boom.
And I was like, oh man, maybe, maybe I just
Logan: Too late.
Drew: it. Um, fortunately didn't. Um, and, and then that, that hyper growth period where the thing is just taking off and just no matter what you're doing, it's just the numbers just going up into the right. Um, it was a really cool. or sort of like, you know, being a novice surfer and then suddenly being on a wave that's like 100 feet tall.
You're like, okay. Um, I mean, it's, it's awkward and scary too. Cause you're like, you know, there's not a lot of style points. You're not doing things kind of technically. Right. But you're just trying to hold on and stay on the board. Um, then it gets tough when sort of that wave flattens out or you're. Um, or like they're like the natural competitive responses to your thing being successful.
It's like all of your, all the big tech companies are going to come after you. Um, you have to scale an organization. So there are a lot of, and then we, we launched a lot of new products that we had to shut down, which is a super painful experience. And then there are ups and downs. Then we had another kind of up period as we went [00:31:00] public, a down period during COVID AI happens back up again.
So, um, Yeah, I feel like either, like, all of the periods either were like, fun to go through in some way, or they weren't fun to go through, but they like, were pretty necessary, or like, you know, maybe I did need kind of a, just to get my teeth knocked in a little bit.
Logan: Um, what I'm curious actually, as you reflect, like the, the reason the product took off to the extent it did in the early days, obviously product was elegant, very well done, and, you know, market right timing for it. Was there anything I, I've actually not heard in the prep that you did that was like actually a particularly good viral distribution hack or some reason that it really shared?
Drew: Yeah, um, Well, I think there are a few things, and, I think part of why Dropbox was able to thread the needle is we sort of got a bunch of things right in disciplines that usually don't go together. Meaning, [00:32:00] Uh, we needed to, there were a lot of other like storage companies or sort of online storage was like the startup cliche the way that, you know, photo sharing or, um, social apps or, you know, there's these like fad things like storage was one of those.
It's hard to remember now, but that, but there are a million of these things. Problem is none of them worked really well. Um, and to get it right, you need to get a few orthogonal things right. So you needed to get the engineering right, because part of why I built Dropbox instead of just like using one of the other services, like if one of those services had worked, I wouldn't have started the company.
Um, but then I would go into like the support forums of each of these other tools and they'd be like, Hey, this thing like ate my wedding photos or I put in my tax returns, not coming out like help. Um, and so that was pretty terrifying to me. Or you put in like a file with like a non English character and the thing would crash.
We put in a big file and it would never finish sending all that kind of stuff. So I was like, I don't trust any of these services. I really just want to get it right. And I'd say there's even within the engineering, there's a part that's like a more [00:33:00] mathy algorithmic, like the sync protocol, correctness, rigor.
That kind of thing. And then there's just like grungy, like operating system, like being in the guts of the kernel and, and, you know, it's like all this and just the plumbing where you had to really, we had to like basically cloud enable other people's operating systems without the source code. Um, and even things like putting the, the sync icons on the files.
It's actually getting it to like, work. That way most people do not. Yeah. Twiddle with all these things. Whereas we wanted to build something simple. Um, but that was even just like, I was one of the, I was, I was that kid in the neighborhood who would just get called to fix everybody's computers. And then you learn a lot about[00:34:00]
Logan: like,
Drew: okay, here's, here's what easy, here's what easy looks like.
Um, so there's a design piece. And then as you're alluding to that, they're getting distribution, like how do you get distribution as a startup? Um, we, that was a huge, um, um, unsolved problem when we got started. Like, all right, is people aren't looking, you don't know you need something like Dropbox till you have it.
Um, for the most part. And, you know, some things like that, like TiVo, like, Oh, you didn't know you needed TiVo till you had it, but you weren't really looking for it. Um, But I, I had a bunch of questions. Like, how are people going to find something like that? I mean, maybe it's like antivirus. People don't, weren't like shopping for antivirus really is sort of.
pre installed on your computer. So like maybe we should go partner with a bunch of big, big, big companies. So, and then, or maybe we should try conventional PR marketing. Those, like none of those things worked. So we were like, Oh God, how are we going to get, how are we going to acquire customers? What really did work was borrowing from a lot of the, um, consumer internet playbook for how you scale [00:35:00] these products.
And particularly like the social platforms. Um, they were the first to really apply, a lot of the viral math, which actually came from, like, epidemiology. Um, so it's things like R naught and spreading and, you know, different factors. Like,
Logan: ai
Drew: yeah, they'd sort of transplant, yeah, exactly. Um, so, uh, but you could really, there was a lot of, um, good stuff about how the social platforms were able to go from, like, zero to millions of users, um, overnight.
That's it. And things like Facebook platform are all just flying. It's like taking off in 2007 when we were starting the company. So we were sort of in the, it was, it was all around us, like that sort of virality and great, you know, the, the birth of like growth hacking. Um, but we basically applied a lot of the same engineering, um, discipline to customer acquisition and that ended up being the thing that really worked.
Where, um, there are a couple of vectors. So one was, well, first we tried a bunch of stuff and most of the stuff didn't work, but the things that worked really worked. [00:36:00] So one was. Um, this referral program that we created, where if I tell you about Dropbox, um, you know, send you a referral link, you get some extra free space, I get some free, uh, extra free space.
So we sort of, sort of double sided incentive, kind of gamifying the referral process, process. And that drove probably triple digit million signups.
Logan: Signups and
Drew: we put little other bells and whistles on it, like we would do it within schools and have schools race against, the students in the schools race against each other, like which college.
Can have the most like free space. Um, so that, you know, I still have people who sometimes say, Oh yeah, I got Dropbox from that. Um, however, many years later, and then the other piece was just sharing, right? And this more classically viral stuff. If I share a folder with you, you become a Dropbox user. Um, or you hear about it and then, you know, and then there was all this crossover of like, Oh, I started using Dropbox at home.
I brought it into work. I started using it with my coworkers. And then that little team becomes a department, becomes a wall to wall deployment. And that's a [00:37:00] game of inches because you sort of have to really think about the funnel. Starts with a sign up, then how do you onboard people successfully? And then how do you get as many people sharing?
As possible. And then if you're sharing like Mac, have as many recipients as possible and then make the recipient experience really good. Like make sure that as many of those invitations are accepted or, you know, shared things are accepted. And so we were, we put unbelievable amounts of effort and like sanding down all these rough edges in onboarding experience.
So for example, when we started maybe like 20 percent of people would figure out how to download the Dropbox app and then put a file in. And that was like Unfortunately that was like a big kind of limiter on the whole system. But then we took, we brought five people in off of, literally off of Craigslist.
We're like, here's 40 bucks. Brought 'em into our office above the Walgreens on Kearney Street in the city, and we're like, all right, go from this email to sharing a file with this email address. And they all, and then we were like sort of [00:38:00] simulcasting it in the other conference room in the office where literally the whole team, we were all watching.
Live as people were like fumbling their way through this. None of them succeeded. Zero of five. Um, but we made this list of like, here are like 85 things here, 85 little rough edges, some big, some small, we send them all down. 25 percent went to 65%. Um, And then we did the same things with like viral, make sure that email deliverability is good, make sure the copy's clear, make sure there's a few steps as possible, contact list, importers, you know, all this stuff.
So, but we then, once we got that working suddenly, then the thing just, it was like an anti gravity machine. It was just like a million users, 2 million users, 10 million users, 20, a hundred, 500, you know,
Logan: How many rounds of Craigslist people did you have?
Drew: we just said that, you know, we did that. That first one was like humbling enough. we, we wrote down so many things that we were able to kind of build
Logan: You didn't need to go
Drew: I think we probably did. I'm of course we did more user
Logan: Yeah, sure. But not, not
Drew: but, but I think that was just like such a searing moment in all of our memories. [00:39:00] Um, and even that was born from like in the early days we would get like whenever you emailed like press at draw or actually get draw a box.
com in the beginning. Uh, but like press at, um, support at, like it would actually go to the whole, whole company, you know, all 10 people in the company. And so everybody could see all of the customer feedback continuously. Um, but I got these, I remember like reading through the support emails one day and I was, the first email was like, it's the most beautiful, elegant, wonderful product saves my life.
Love you. And I was like, this is great. And the next one I was like, this is the most like, impenetrable, incomprehensible garbage I've ever used. Like you MIT nerds, like can't design a product to save your life. And I was like, first of all, they like personalized it to my college. And I'm like,
Logan: You're like, thanks
Drew: okay, yeah, exactly.
So I'm like, but I'm like more than any of that. I was like, how can I literally be getting these two emails? Like one after the other. Okay. We should probably do some music testing. But it ended up being one of the most important things that we did.
Logan: [00:40:00] what about? Um, so, so there was the period of time of just like viral growth and it was a rocket ship ascension through that. And I'm sure there's a bunch of things that you look back on finally. And also you're just hanging on as that wave is, you know, Taking off, I guess, as you reflect on that, uh, I'm sure people would be very fortunate to find themselves in a position, but is there something that you look back on during that period of time that you wished you had done differently?
Or was it just kind of survive in advance as much as you could?
Drew: I mean, obviously what sort of is not. shown is like, there were like a hundred other things that we tried with like partnerships with PC manufacturers, any virus referral programs, like
Logan: all the survivorship
Drew: other stuff that like had no, no returns. But, um, no, I mean, I, I was super proud of what we, and there were other things, not just the virality, but also just monetization, like the freemium model, how we acquire business customers.
Like we were pretty,
Logan: pretty pioneering
Drew: serve, [00:41:00] viral, product led, you know, kind of transplanting this, like, consumer playbook to business software. Um, I mean, the industry's taken it many steps further, um, since then. But, um, but yeah, no, I'm like super happy with how
Logan: You, you, you referenced like having to shut down products, uh, which I'm sure was a hard decision. Uh, there was, there was mail and carousel and probably a few others. Um, but you did it fairly, you didn't, um, beat around the bush with it. At least as I reflect on how long it took you to do that. Um, one in making a decision like that to, to ruthlessly prioritize on something.
What. Was that an instinctual thing or was the data just telling you like this this isn't where you need to go?
Drew: I think it happened gradually then suddenly. Um, uh, I'm It was rough. I mean, in 2014, we were spreading our wings as a company. [00:42:00] Um, and we knew we had a, I think early on, even like well before then, I think even in 2007, 2008, we're like, we're just gonna get crushed by the big companies. Um, and we're just sort of waiting under the shadow of the Google Drive hammer that was one day just going to come down and just
Logan: And you were
Drew: turn us
Logan: I mean to your credit you were pretty Public about when people would say that you're like, yeah, probably like that's probably what's gonna happen. But if not, what could this be? And what could people singularly focused on this problem actually be able to accomplish, which I think is a level of self awareness.
And also to anyone that's fundraising, at least on our side, it's like, Yeah, no, I mean, statistically, most things will get beaten by some competitor. And there's a level of, you know, Self awareness and humbleness that I think in saying that is effective.
Drew: Well, and then I think when you just sort of start from the user experience for like two investors, they're like, yeah, this is a graveyard. No, none of these companies [00:43:00] have succeeded. It's a, it's a commodity. Um, but mainly, mainly it's just like too much competition. And I was like, okay, but do you use any of these products?
And they'd be like, no, I'd be like, isn't that interesting? Step one in our adventure was not like conquer earth. It was like. Not carry a thumb drive.
Logan: Yeah.
Drew: Um, so, and I, you know, it's also pretty clear, like, okay, even if we did get totally just like mogged by one of these big companies, like all the other companies would need some solution and they, you know, we could sell the company and go do another kind of crazy side project or, or whatever.
But, um, but that like didn't happen. Well, it was interesting. Like then all these products launched like iCloud, Google drive, one drive. Um, Amazon had. Things in the space, like everyone had things in the space. And so, um, but it was sort of like you sort of saw the mushroom cloud in the distance, but you didn't hear or feel anything.
So it's just a very delayed reaction where, I mean, the press likes [00:44:00] to write like, Oh, you know, Google launch launches competing product to start up just like dead tomorrow. Um, and it doesn't really, Comcast is more like a boa constrictor than a shotgun, um, in that. You know, we could not, you could not look at our analytics and point to a date, uh, you know, our engagement numbers, our revenue numbers, anything and be like, Oh, that's when I cloud launch.
Oh, that's when one drive launched. What it was, what it was more like was sort of a taking the oxygen, like slowing our future growth or taking our future customers. And then I think what made that, you know, So, I would have expected that, oh, maybe the 2011, 2012, that whole period when these things launch was going to be the painful period.
Actually, that was fine. Um, if anything, it just sort of reinvigorated everyone. They're like, okay, well, all the convenient things have launched, we can't really tell that there's a problem here. And yet, um, it was pretty clear, like, we were on borrowed time and, like, [00:45:00] strategically effed. As one of my, uh, Mentors might have put it.
Uh, where, you know, because what was challenging was like, we were spreading it, Dropbox was still taking off, but I reckon that it was pretty obvious that, um, and problematic that Dropbox was not a great, not going to be a great experience for all the different use cases. Um, so, We experienced the benefit of Dropbox being such a horizontal general product in that, you know, who's Dropbox for, what's Dropbox for, it's like, well, what's a computer for, what's a phone for.
So this meant we had like an infinity, kind of like literally anyone with an internet connection kind of TAM from one perspective. But then when you think of what people actually do with the product, tons of difference. And it's like, all right, well, you know, people use Dropbox to back up their phone and they use it to share photos.
They use it to like do replace their file server or they're just. Run their companies in the cloud, but like the ideal products for each of those use cases is like [00:46:00] quite different.
Logan: And so what you do from a complimentary standpoint, what the next act is, if you sample a hundred users, you might get 500 different
Drew: Yeah. And then, you know, some kind of products kind of die this way or don't reach a lot of products kind of don't reach their potential because they get kind of picked apart from unbundling like Craigslist in the beginning was like. You know, Airbnb plus Upwork, plus like all these companies, but like all the, you know, the list goes on all kinds of things.
Um, but the most valuable verticals got kind of picked off by more purpose built, like, Oh yeah, Airbnb is a much better experience for like short term housing or, you know, getting a job. There's a lot of better alternatives. Um, so we launched things like carousel as a photo sharing service, like recognizing that, no, we shouldn't, you know, you shouldn't have a bunch of Photosharing should not be a bunch of JPEG files starting with like DSC004, you know, right?
So, we built a more purpose built thing for that, and then we're like, well, we should also branch out [00:47:00] from storage and like, get into these more collaboration use cases. Mobile email, or like, Mailbox, we bought this company that was like the first good mobile email
Logan: What was the name of the company? It was
Drew: Yeah. And these guys had like a million,
Logan: It was a great
Drew: yeah, they had a beta waiting list with a million people on it.
And so we're like, we haven't seen this since like our own thing and, and, um, bought the company and founders are amazing. We made a little bit of a mistake where we put the founders in charge of all this other stuff at Dropbox and sort of the product itself kind of stagnated. But again, the bigger issue is like, we're just going to get like, everything we build is going to get copy, bundle, kill.
Logan: I mean, maybe it was just inevitability, but as you look on that one in making the decision, not what you should have done, but I'm curious to that in a second, but like in making that decision. Yeah. Yeah. Did you, did you take too many inputs and opinions from people? Do you wish you were more declarative about, no, this is what the company's going to be, or is this just an inevitable when you have a lot of different use cases, it's hard to know what signal and what's noise.
Drew: Yeah, [00:48:00] I think we're sort of, so during that. You know, 2011 to 2014 period, I think there were a lot of problems or kind of a lot of debt that was going to, that's maybe strategic debt that was going to come due. It kind of all came due at once in like 2015. Um, so what happened was, so, you know, 2014 we'll launch things like carousel, buy mailbox, we're improving these things.
Um, but it was clear along the way that like, all right, all these use cases of Dropbox. You're backing up your phone. Okay. So in that market we're like competing against the iPhone to back up your iPhone. Like how is that going to go? Then in photo sharing, it's like, all right, well, now we're competing.
It's like not just Google and Apple, but also Facebook and snap and Instagram and all these guys. And it's like, okay, to go? Then it was like, all right, well, there's like this work use case and collaboration. Unfortunately that was pretty durable. Um, but even then we're fighting against Microsoft and Google.
So, um, But I think the question is sort of like the face that faces you or that faced [00:49:00] me. It was like, well, how are we going to get out of this? This is, yeah, maybe it's a little bit time delayed, but you know, I, I studied companies like Netscape versus Internet Explorer, uh, or MySpace or Friendster versus Facebook.
Um, there are a lot of dead pioneers. Um, and I was like, I see this happening to us. Um, Where, and especially if you start losing, like no one wants to work at the, you know, the MySpace when the Facebook is kicking your ass. Um, so, uh, and then as you, as even back then, like Netscape didn't get killed by Internet Explorer 1.
0 or 2. 0 or even 3. 0. It was like, you know, IE3, IE4, IE5, then that's when things really started, the bottom started falling out. And so, we could start to see that happen 2014, 2015. Um, but I'd say like one of them, there were a bunch of notable things. Like one was, um, Google photos launched and just gave away like unlimited free storage to everyone for life.
Just like not [00:50:00] even trying to make money. Um, and then bundling it with their phone and just doing all the things that, you know, an incumbent would do. And I'm like, wow, do we look stupid or more? Wow. Do I look stupid? Um, but then like everybody, you know, all of the industry, all of our employees are like, what the hell are we doing?
Um, Like that looked really bad. And so I'd say there was sort of a groundswell internally of like, how are we going to solve it? Like, what are we going to do? Um, none of these things seem viable. Um, and I reread, uh, this book, Only the Paranoid Survive, um, because we were not the first company to be in this kind of situation.
Um, Intel actually had this pretty early in their lifetime. I mean, we all know Intel is like this microprocessor company, but actually they were founded as a, a memory company, um, making like RAM for your computer. Um, but then they, you know, long story short, they had all this, this sort of onslaught of like Japanese competitors that were like subsidized by the government or other one way or another.
There's like [00:51:00] totally unlevel playing field. Like they weren't dying yet, but they were probably going to. And Andy walks through that whole, you know, sort of a memoir of that whole experience. And he's like, if we were advising ourselves, Uh, or this little vignette where they're in and the two of the founders are in Andy's office and he's CEO at the time.
And he's like, you know, if we are advisors to ourselves or consultants to ourselves, what would, what would we advise? What advice would we give or what would we do? And they're like, Oh, we'd get out of the memory business and go all in on microprocessors. And sort of sounds easy, but it's sort of like Google saying like, yeah, let's get us search.
And then Andy talks convincingly about, yeah, there are these strategic inflection points and markets and, You know, when you're in one, you really, like, CEOs normally want to hedge and sort of keep all these options open. That's the opposite of what you should do. You should put all your eggs in one basket, watch that basket, and the Mark Twain quote.
Um, so I'm like, yeah, that's what we got to do. We have to sort of stop focusing on anything else, but this, like, collaboration opportunity. Stop, stop thinking about photo sharing, stop thinking about consumer [00:52:00] generally. And it was tough, but I'm like we, but I would never forgive myself if like we were fighting ten wars, lost all ten, when we could have won one of them.
And so that's, that's what we did and I'm like super glad that we did that. And so it was pretty, I mean it was a slow build, but you know, I remember going home, kind of licking my wounds in New Hampshire with my family, like rereading that book. I'm like, oh god, I gotta do this. And so I went back, shut everything down.
Um, I wouldn't say that, like, solved all our problems. I think that sent, like, the narrative of the company into a tailspin. Um, certainly sent morale and other things, all these other problems. But, um, kind of triaging the issues. You know, we, we went from hemorrhaging cash because we were sort of in land grab mode to being, you know, That was the start of being like pretty financially disciplined.
We turned cash flow positive maybe seven eight months after that in 2016 But yeah, that was sort of like the initiate that like the hazing ritual into the big leagues like but it was really it was really tough
Logan: Was there [00:53:00] something after that journey that you thought about reinvesting in? I mean, you subsequently lost your off box paper. Now we have Dash on the search side. And so at some point it, you've picked your head up and, and launched other products. Uh, so what was that decision to go after and do other stuff
Drew: Yeah Well, I'd say there was a pretty long period of malaise both like with the company and then me personally as CEO I was sort of like, okay, so we shut down all that stuff You People are like, got it. I understand what we're not doing, but what, what are we doing? And you know, the truth is like, if I knew the answer to that question, like we
Logan: We're doubling down. They're like, well, what is my
Drew: I'm like, uh, yeah.
So, um, and then part of it was also just for me personally, like stepping back, like, like, I don't know, like this is a pretty rough, like we built this awesome product. It wasn't that like these products were like bad or didn't work. Actually the product problem was like, they were really good and like rapidly copied, but I'm [00:54:00] like,
Logan: What was the scale of the business? How many people at this point?
Drew: About 500
Logan: 500 people doing a couple hundred million?
Drew: Doing, yeah, probably a few hundred million. That's right, yep.
Logan: And about to be break even. So these are caviar, caviar scale, uh, but you know, some very real problems
Drew: Totally, but you know, again, there were Blackberry, Nokia, like a lot of these companies that got way further than we did, and you know, um, So, it was pretty terrifying, and it wasn't even like we had that, it was like, okay, we're gonna focus on collaboration, but we didn't really have an answer of like, Um, what that would look like or why we'd be better.
And then, you know, for me personally, I'm like, I also am not, we just got totally like crushed by an incumbent, like in one area. So, you know, it's not even just a question of like, what, what do we do? It's, or what can we build? It's like, what, like we can have a hundred ideas on what to build and probably 99 of them would face the exact same fate of [00:55:00] being like, Oh, cool idea.
Dropbox. Thanks for being free R and D for us. Um, so that was really the problem.
Logan: Decision paralysis of like the
Drew: Or just like what, you know, yeah, you're like, okay, I'm, I'm trying to play chess as, as well as I can, but like they got eight queens in the back and this is, this game isn't that fun anymore. And so then it sort of drew me to like, well, why does Dropbox need to exist?
Um, if we're just like inventing stuff 10 minutes before, you know, the big, the big tech company is like, That's cool, I guess, but it's like really, it's not, it's not very fun.
Logan: Was that the closest you thought of getting off the ride? Like selling the company at that point?
Drew: Yeah, totally.
Logan: sounds like a point in time. Most founders would just
Drew: yeah. So yeah, it felt sort of like, I don't know, like being in, like in the basement and fight club, but not in a cool way. It's like, like in the, yes, I'm bleeding and my teeth are on the ground. But like, I don't know how I'm going to get up from this. Um, but it sort of, [00:56:00] uh, like a lot of these things were all coming together where it's sort of on top of that, like this company is like super successful, but I'd sort of run out of merit badges to get.
I'm like, we'd raised the money at the valuations we wanted. And there's like, is this it? It was just big numbers, bigger. Um, and, um, You know, yeah, maybe it's time for my like tech bro ascendancy, like maybe I should be flying car, working on flying cars or space or climate or, you know, but not files like we kind of did the file syncing thing.
Like maybe it's done. That would be okay. Um, I think it'd be okay. And then I was even just looking at my own experience and I was like, okay, this is cool. Um, You know, the company is like so much more, my 18 year old self would be like baffled why I'm upset at all. They'd be like, dude, you did it. Um, but you know, as I'm just going through running this company, I'm like, Oh man, like I'm just in meetings all day, emails all night, sleep, repeat, you know, [00:57:00] uh, first world problem.
But,
Logan: But still, I mean, you, you recalibrate to the new normal, uh, of all that stuff and like getting your, it doesn't matter what your paper bank account
Drew: Sure. Yeah. No, I was
Logan: your days are bad
Drew: bad. My days are bad. Um, and, well, because I was like trying to really do a retro of like, alright, God, I could have totally predicted that these companies were going to respond this way. Like, that was stupid. Why did I go after these things that were so easily countered?
And I was like, oh, I
Logan: But you knew that you knew that way back in 2000,
Drew: 2008? Yeah. You
Logan: you sort of, that, that was always the
Drew: But when I was like doing the root, sort of trying to do some kind of like root cause, and I'm like Oh, like I really have like no time to think. Um, this is, and I'm like, well, this is really weird. Um, because I, my like subjective experience of work is like, I'm just like re busy, but I'm not really productive or I don't know if I'm being productive.
And like, my brain just feels stuck in first gear [00:58:00] and like, yeah, I would have probably had like a better solution to some of these problems if I just had like some space to think. And then I just like, I'm like, wait, why does knowledge work? Like work this way? Isn't the whole premise, like we hire these people for their minds, but then why did we not give them like the space and time to think?
And so I just like went down this rabbit hole of like, what is going on with productivity? Um, and. I'm like, something, just, something seems to have, Something went, like, very wrong with, like, the tools we're using. Where, you know, imagine, I would think, like, Imagine if Einstein were alive today, You know, even back then, certainly still now, Um, but like, yeah, even back then, he'd wake up, Um, he'd check his email, He'd like delete a bunch of LinkedIn invitations, right?
Then he'd like, okay, start, sit down and start working or writing equations. But then like, someone would Slack him being like, Hey, do you have, you know, a PDF of that, of [00:59:00] your paper? Or, you know, and then, okay, yeah, here, think back to work. Then the phone would buzz with some like tweet. And I'm like, would we still understand relativity?
And then I would talk to people who were actually working on these other moon shots. Like I interviewed an engineering director from SpaceX and uh, I was like, I was kind of jealous. I'm like, wow, like it's just so cool. You guys are like actually going to Mars. Like how's that work? Like how do you guys work together?
Um, he's like, what do you mean? I'm like, I don't know. Like how do you collaborate? What tools do you use? Like what's it like to work there? He's like, and basically like, what's it gonna take to put someone on Mars? And the answer was basically, a lot of emails and a lot of files. And I thought about that, and I'm like, man, OK, yes, on the one hand, um, emails and files are probably better than what preceded them, but, like, We have to think of technology not just as a force multiplier, but also as like the limiting factor on progress and somehow, and we can only go as fast as our tools let us [01:00:00] go.
And I felt like, the, um, No, all of these companies and even the people working on these other moonshots were like all constrained by like, I can't think. And the, in our environment at work had become so kind of distracted and overwhelmed or at distracting and overwhelming and full of all this sort of busy work.
And these things that were, we were hailing as like the frontier of productivity, like Slack, we're actually these like, Distraction engines that are introducing like all these empty carbs, you know, into the, into our experience. There's like, or bombing us or peppering us with all these interruptions. And, and it was bizarre cause I'm like on the, you know, in brain science or even just like common knowledge, like, yeah, people do their best work when they can focus and there's some kind of flow state, uh, not being in or when they're not interrupted, blah, blah, blah.
Um, but then we go to work and. There's a bunch of like cultural problems, but even just like technically, like the technical environment, [01:01:00] the screen we look at, it's like if you wanted to design an environment that made it impossible to ever get into a flow state, like that's kind of what we have. And I'm like, okay, maybe, maybe the real moonshot here, um, there's a moonshot that powers all the other moonshots.
It's like helping people get back to being able to use their brains. At work or like being able to run their brain and something more than first gear or just like not be distracted and overwhelmed all the time. So we, that was sort of the first answer for like, okay, what, what, what problems are not solving themselves?
Because like, you know, a photo gallery in your pocket that was synced to the cloud. Like we were the first to ship that at scale, but we, that probably would have like happened. Um, and you know, Apple would have figured out the whole premise of iCloud was like, yeah, your phone should be back to the cloud by default.
Uh, But, um, so like a lot of things we were building, which probably have been figured out by other people, but I'm like, here is a problem that it's like, not even, not only is it not solved, like no one's even [01:02:00] framing it properly. Um, and when you think about like the, the ultimate non renewable resource that we have, like that I have as a person or my company has, or our economy and society have, it's like really your brain power, right.
Or your time. And I'm like, We don't even, like, we're really good at managing money as a species, but like when it comes to our like time or our attention, like we are completely blind. Um, so we established this new mission for Dropbox to design a more enlightened way of working on the premise that, yeah, the way we're working isn't functioning, like it's unenlightened, unexamined.
It'll look really medieval in 10 years. Um, and I think, yeah, well civilizationally we'll look back on this and be like, yeah, remember that we had that weird detour. Sometime in the 2010s where we basically just like stopped using our brains at work and like put ourselves in these like super cognitively polluted environments.
That was nuts. And so that's really what we're, that's
Logan: What year was
Drew: [01:03:00] umbrella 2017.
Logan: and that's led to what products were, have been sort of the offshoots from, from that for you.
Drew: So the biggest one is Dash. Um, I mean, we've had other products like Dropbox paper. We've done some acquisitions, but I'd say to really move in the direction of our mission, Dash is the most important first step. And it also requires a couple of things. So first. Um, COVID was another big turning point where, um, you know, suddenly all of work is digitized and the world recognized, Hey, we can work out of screens instead of offices.
Um, and that shift had been happening for a while, but like COVID, like completely permanently finished the swing. But then that created all these new problems where like, now we're using our screens a lot more, putting a lot more stress on the system. And when you're distributed or remote or hybrid or whatever, whatever flavor same problem, like I can't find the basic information I need to do my job.
Um, And that's a problem because like, knowledge work without the knowledge is pretty hard to do. Um, so Dash is like a very direct, [01:04:00] um, it's both like a direct response to that problem, like how do you make it easier, make a better environment that sort of knows what you're working on and can surface this to you.
Um, but search is just one piece of it. It's really, Dash is like the 2024 version of the problems we started solving in 2007. Problems I started solving in 2007 was like, yeah, I kept, I didn't like carrying my thumb drive, didn't like emailing myself files. But really the bigger picture was like, I can't find my stuff.
I can't organize it. I can't share it. I can't keep it safe. Because it's like all my stuff is scattered in all these different devices and these different operating systems. And you know back in the day the solution looked like file syncing to the cloud and across everything, but today we're sort of back to the same problem.
And maybe a hundred files on our desktop are now a hundred tabs in our browser. And we've talked about search, but there's also problems like, how do I organize my stuff? How do I share my stuff? Um, and they're as basic as just like, there's no persistence in the modern world, right? So when [01:05:00] you go to bed and wake up, your physical papers are still on your desk.
You reboot your computer, your files are still there. But when you're in your browser, Your browser kind of like clears itself out either because you like close it by declaring tab bankruptcy or your operating system like updates in the wrong way or one way or another, when you close your browser, it takes all of your state with it.
And so if you don't even have like a durable place, what does it even mean to organize anything? What does it mean to organize things across multiple tools? It's just like a broken system and you know, files and folders had their problems too, but it's sort of weird that it's like, well, files have folders, songs have playlists, links have, right?
There's like no collection concept, right? So there's, there's a lot of like UX problems hidden in plain sight or things that actually were worse, that were better, were better 20 years ago and worse today. So just concept of organizing your stuff and then sharing, right? If whether you're remodeling your house or getting ready for a board meeting.
What do you do if you have like a Google Doc? Plus an air table, plus a 10 gig 4k video. There's [01:06:00] no common container that holds mixed format things. Um, and so dash is really designed to solve all these problems, like both starting with search, but then giving you a start page that kind of organizes your stuff and what you're working on and the system kind of like organize your stuff the way that, you know, YouTube or Netflix or Spotify, organize your stuff, organize your content for you in a consumer context.
Um, and then sharing, like we, we have this, uh, we have stacks, which are basically these smart collections. That can hold any kind of content. Um, and, you know, our aspiration is for this to be like the default way to share mixed format stuff on the internet. Um, so, but these are really kind of solving the same genre of problems that we were successful with in, with Dropbox 1.
0. But with Dash, it's really about how do we solve these in more of like a cloud native, browser native, modern context. I mean, we also support your files too, but it's, we're revolving from, storing and syncing your files to organizing all of your cloud content. And then that, that sort of step opens up a lot of new stuff on the tech tree after that.
Logan: As you think [01:07:00] about galvanizing behind a mission like that, it obviously started with coming up with the new way of working and operating, which I totally resonates. It's amazing how by interruption we operate. Uh, even today, it just gets, it gets worse every day. Um, As you think about like setting the mission, uh, or the vision around something like dash, um, how do you go about identifying that, that opportunity?
I assume there's been a bunch of different iterations on it. Then when you find it, do you, how do you get the company behind it, especially as a public business that, you know, you, you have to share things with shareholders and all that stuff.
Drew: Um, well I think you triangulate on these opportunities from a bunch of different directions. Um, and you know, the ones that pass all these filters are the ones you really double down on. So, You know, first as a mission, I talked about, you know, okay, how do we, I think a lot of people agree. Okay. [01:08:00] Yeah, sure.
Work is, work sucks. I know. Okay. But like, what do you do about it? Right? So where specifically should we try to fix the environment? Um, and then one of the things that, what, what plays to Dropbox's strength. So there's sort of a strategy piece of it where it's like, well, you know, one of Dropbox's advantages is that we're platform agnostic.
And a lot of the other companies are going to try to keep you within their garden. Um, yeah. Dropbox is not going to try to do that. And then, you know, universal search, enterprise search, you kind of, it has to be kind of platform agnostic by default. Cause we just don't live in a monoculture world anymore.
So like everybody has this problem. So like, okay, that's a good fit for Dropbox. And it also requires a lot of customer trust. Also requires a lot of technical, like solving new technical problems. Also, you know, it's also helps if, you know, if step two is organize all your cloud stuff, step one, get hundreds of millions of people to trust you with your files is like a good stepping stone to that.
Um, So even if we had gone directly at Dash, we would have built all these kind of prereqs along the way. Um, then it's, uh, you know, I, I like to experience [01:09:00] things myself, and that, you know, just coding the first, sort of the proto version of Dash, or coding a personal search engine myself, and being like, whoa, okay.
There's a big why now here, because like there's all these new search technologies that make a completely better experience. Um, but weirdly, you know, enterprise search was kind of a graveyard. It was sort of, yeah, a lot of people had swung at this, this is an idea. Yeah, enterprise search, plus like ideas like the intranet, like, had been around for a long time.
You know, Google desktop search, 2004, like, So there were also precursor things that were good. So, like, Google desktop search is an example. Like, that was a great product. Um, but Google kind of killed it. You know, and then plus, well, we ceded the Dash effort by buying a company called Command E that had been working on this.
Um, so, kind of a lot of things pointed in the same direction. Direction. I don't think it was any one like, you know, at home. Well, I think that if there wasn't a home moment for me, it was like doing that search myself and be like, Oh my [01:10:00] God, this is so much better than what I'm doing today. Like we have to get after this.
So that's probably the singular most important thing to me. But then it passed all the other filters from like a product distribution strategy standpoint.
Logan: It sounds like there's a lot of enthusiasm right now about all of this stuff. And AI has been something of a, of a jolt for you in the, is that business? Is that fair
characterization?
Drew: I mean, I was pretty rough in the. late innings of last season where it's sort of like, yeah, we're, we're sort of checkmated. Like, well, first of all, you know, a lot of the workflow files will be around forever and they're still a critical part of, um, the kind of work experience.
And they're especially critical with our kind of key customer segments, like people that create media or people that do a lot of external sharing. Like they rely on Dropbox.
Logan: for, for this
Drew: yeah, right. So, but it's sort of like. I don't know. Were we in a situation, like creating more demand for it is like, is that like blockbusters?
Like, how would you create more demand for walking into a store and renting a DVD? It's just sort of, [01:11:00] there, maybe it's just sort of had its time or it's peaked. And so like, what do you do about that? And then, yeah, with AI or any sort of new era in computing and like the ones I've been sort of was, uh, Grown up through where like, I saw the internet era, I saw the cloud and mobile, they made Dropbox possible.
And then now AI was the next big one. I'm like, okay, yeah, the concrete kind of unfreezes. And for a few years, there's just like this chaos, um, sort of like raining on the racetrack. Suddenly all these people can make all these moves and really change position in ways and create new franchises, uh, in ways that are just not possible when things are more kind of more hardened.
So yeah, I got a massive jolt of enthusiasm, both because of what it meant to the market, but then more importantly, what it meant for the technology and the products we can build. Suddenly there's like all these new colors we can paint with. Yeah,[01:12:00]
Logan: Along the journey, I'm sure different motivations have come and gone. You don't need to be doing this, doing this next act in Commandee. I'm curious, like, why, what keeps you motivated to get up every day and sort of fight this knife fight at different points in time when they had eight queens or whatever
Drew: Yeah, yeah, yeah.
Logan: of the board.
Drew: Um, well, I think it was in the early years, it really was kind of the novelty of. Okay, I want to start a company and that's something I dreamed of and and then that that process can be pretty linear and kind of merit badgy. It's like, okay, let's get into Y Combinator. Okay, let's raise money from, you know, Venture.
Let's
Logan: sure the first five to seven years were pretty,
Drew: Yeah. And it's like, okay, we raised money at a six million valuation. Now we're going to raise money at 27 million valuation. Oh, Now we're going to raise it to 4. 2 billion valuation a few years later. Like it was sort of, yeah, there was always like these next hurdle. I know 10 million users, a hundred million, 500, whatever.
Um, so, [01:13:00] and, and I think that was part of what made the competitive stuff for that sort of reckoning pretty difficult to sort of like, you know, not only are we in this like really bad competitive position, but now I'm just kind of rudderless, like, what do I even want to do? And so I think,
Logan: like identity becomes tied up
Drew: yeah, like what do I want to do in my life?
Logan: and the, you know,
Drew: Well, and, and that sort of stopped being motivational to me. Cause not in, you know, I, I, I certainly like enjoy the process of making a commercially successful company. It's fun to make money. Um, the most fun is like having access to all these interesting people and, and just, um, so there's a lot of good stuff there, but it wasn't like a two X more of that money or valuation was really gonna make it that much of a difference.
Um, Um, and so, but like going through that fire, like did get me through the other side with like a new purpose of like, okay, yeah, this meta moonshot of like helping people use their brains at work. That's something that's like [01:14:00] unsolved really matters to me. Something I feel like I can put a dent in. Um, so personal sense of mission and purpose, but then also just like, all right, why do I want to be a CEO?
Um, and yeah, I'm not, I'm not really, I've sort of cleared what I want to clear as far as, you know, resume. Um, and so I decided like pretty explicitly, I was like, all right, I'm gonna do this, um, keep it, or I really want to get great at the craft of being a CEO, because that's something you can, um, like it's totally within your control, it's a, it's something you don't really even master.
Uh, you know, in a lifetime, you know, Steve jobs was probably like just hitting his stride at 55, uh, you know, 55, 56 when he died. Um, so I was like, I feel pretty lucky to be in a line of work where you don't like peak in your twenties or thirties. Cause I mean, you know, if you're in sports or like math or different things, um, you often have like a sort of artificial ceiling on, or, you know, [01:15:00] it's sort of an externally imposed expiration date.
Um, but like business and being a CEO isn't like that. Um, and it's painful. And I, even when I was like 20 in my early twenties, like, I was like, I want a job that gives me, that has like a high learning curve. And I was like, Oh God, well, I certainly got that. Um, and then, you know, I just never get tired of, you know, hearing people's, of hearing about how Dropbox has changed their experience.
And it's like really gratifying when people say like, Oh, I love Dropbox. You know, I will always look over someone's shoulder in a Starbucks to see if the little blue, you know, little icon is there. Um, and that feeling of like, Oh, we actually like changed how people do something. Um, you know, or it's certainly at the peak when like people were using Dropbox as a verb and it was like, you know, on movies and stuff like that.
Just that feeling of like leaving, um, just really making something better at scale. And that kind of impact is like super. And, um, and I [01:16:00] think like tech was really rallied around that concept for a while. And then sort of tech had a pretty necessary correction, you know, the narrative where, um, you know, stop taking so much credit for all the good things without taking any responsibility for the bad things.
But it's like always been true that technology is both like the source of all our problems and it's still the solution to all our problems. And there's really, you know, pick your, pick your, favorite problem of the world, like we're gonna need to invent our way out of that. So it's really around like craft and invention and impact that still keeps me going.
And just like I love that the game is hard. And you know, I think one of the things you have to learn too beyond reading or, you know, the more technical aspects of the job are some of the things that are more around character or mindset. You know, if you, you should, you never want to feel like a victim.
Practically speaking. Um, and you want to focus on the things in your control and like, you know, you can't control what happens to you per se, but you can control how you respond. So, [01:17:00] you know, sure. You know, I'm human. Like I certainly, I'm not, not bitter about some of the competitive stuff, but at the same time, like, well, uh, There's definitely things I could have done to like navigate us through that a little bit better or stuff that I've learned now.
And even when I talked to some of the execs who worked at like Netscape, um, I was like, man, that must have been, like, that was, that was BS. Like they shouldn't have like copied, you know, that you blah, blah, blah, Microsoft. And the guy kind of snorted and they got Bill Campbell was the name. He's like snorted.
Laughed. He's like, no, trust me. Our, our wounds were entirely self inflicted and I, that just stuck with me as like, oh man. Um, Uh, that's like really honest. And I think it's like, mostly true. Like most of our, you know, if I'm really honest with myself, we've, we've had like a long period of stagnation. Um, but that's okay.
Like companies do. And, you know, fortunately, tech's a sport where you can hit like a 50 run home run and, um, and no really, nobody [01:18:00] really cares about what happened last season.
Logan: I guess you've gotten exposed Bill Campbell, did you get Andy Grove too, a little bit? You sort of, you sort of span both generations of, of tech in that way, I guess, right? Uh, coming up in 07 through this period of time, like the old guard was still
Drew: Yeah, totally.
Logan: some ways. And those two passed away around in
Drew: They did, and actually that was, I left that out. I think, um, you know, as I sort of, all this stuff was swirling, and I'm like, what do I want to do with my life? And like, what do I, um, uh, you know, I was like, at least I was able to read Only the Paranoid Survive, and, you know, Andy's books. Um, And so I'm like, Bill Campbell was like kind enough to spend time with me.
Like we, we didn't, we weren't that close, but we went out to dinner a bunch of times. And I was like, you know, again, like nursing my wounds. And he's like, he's like, shut, shut up and get your ass back out there. You'll be fine. You'll figure [01:19:00] it out. And he's just like, I really needed that. And I felt like it
Logan: that don't know, Campbell was coached, uh, jobs and Sergei and Larry and, uh, I mean, on, on, on the
Drew: Yeah. And so he's sort of like made you believe in yourself more than you actually believed in yourself. Um, which meant a lot to him because he wasn't an advisor. He didn't have any equity. He got sort of nothing sort of tangible from that. And I was like, that's just like so cool that someone would do that.
And I really needed it. Um, and then Andy, you know, his whole, one of his, um, big inventions was like this concept of like managerial leverage. And he really. You know, and what's nuts about it. I mean, Andy had a crazy life, like immigrating from Hungary and all these different things, but like, just think about, um, you know, he, he was running Intel, which was kind of the Google of its time.
And like in whatever spare time he did he he had he would just like write these books to like light the path a Little bit for those who may follow so like, you know, only the paranoid [01:20:00] survive high output management Um, like those had a big impact on me I'm like i'm like at least i'm having this set of problems because if I didn't read that stuff or I weren't exposed to things like this Then I wouldn't you know, my life would be very different um and so Uh through both and then knowing when they both died.
I was like man You You know, maybe there's something I should be doing kind of in this vein, like make it, it should be easier to scale human organization. Like we should be man, we should be treating our, um, you know, limited brainpower with a lot more reverence. Like we should be redesigning our environment to like, like to not have these, have these problems.
And I'm like in this privileged position where I can actually like have, I have tons of like capital and resources and people to go do something about it, which is like super cool. Um, And, but like, you know, and when I have like so many more, so much more leverage to solve a problem like this, then either Bill or Andy did like Andy his best, like mechanism to make things better was like write a book.[01:21:00]
But the book, you know, hyper management was. Like publish in like 1983, like of the people who like need to read it, maybe 0. 1 percent have ever heard of it. And you go all the way down the funnel. You're like, how many people like read it and then take no action whatsoever as a result. But you know, our customers would tell us like Dropbox is not just a place where I like sync files.
This is like my, this is where I like go to work. Like this is where my dreams come true. This is, it doesn't keep files in sync. It keeps like, My team in sync keeps people in sync. And we were just like, people were using all these metaphors, even 10 years ago, of like Dropbox being this workplace or this like working environment.
And I was like, we had just like, never thought of it that way. And I was like, man, if we did think of it that way, we would have done things like so differently. And so we get to do a lot of that now.
Logan: I guess, uh, one that I think you might be appropriate to weigh in on, but, uh, you, you had a number of, um, You, you had a number of successful fundraisers. Uh, I'm curious as [01:22:00] you reflect, do you think fundraise and valuation, like it was, is that just a rounding error in the history of the company and hey, raise as much money as you can.
And at the end of the day, you'll survive or not on your own merit. Or do you reflect on high valuations or things and say some way that might have, um, exacerbated some of
Drew: Oh yeah. Yeah.
Logan: Because I can't say that it sounds very self serving when it comes from me.
Uh, when it comes from you, it's more credible.
Drew: well, I think the way I think about it is more, um, I mean, especially with some hindsight, like I've been overvalued, I've been undervalued. They can both be quite painful. Actually being overvalued, it can be more painful in some regards. So, uh, yeah, for sure. I mean, we, we went from like a raising, uh, raising a 4 billion valuation in 2011.
[01:23:00] To raising a 10 billion valuation in 2014. I wish I could say that our thought process or my thought process, I was just like, 10 sounds like a good number and investors are willing to give it to us. So we took it. And then after, I just remember afterwards, like looking at the market caps of all these public companies and like being like, wow, Dropbox is like more valuable than all like the, you know, all these like blue chip companies that I really admired.
And I'm like, either I'm a genius or I'm an idiot. It was like, wow. And I knew very little about corporate finance at that time. And, you know, I don't, I don't think it was like categorically that I think we did, we overshot by a lot, um, which then made a lot of the, made a lot easier to paint this like pretty rapid reversal or reversal in the narrative from like, Oh, these companies now.
Going from the little engine that could to then the first dead deck of corn and all
Logan: There's only two narratives, right? It's, it's either up or down. And so you gotta know which side of
Drew: I think if we raised it six, probably [01:24:00] much of that would have been like easier, but you know, at the same time, I'm like optimize dilution a lot. Um, so it's just sort of, you know, you're, you're trying to keep a lot of these things in balance.
Like you, and importantly, I think whenever you're the way I think about a lesson, like overvalued or undervalued is good or bad, I think it's more like. Well, just like try to play it, um, uh, like play the game forward a few steps and, and certainly work back from like, okay, if we're a public company, um, you know, by and large our valuation will not be vibes based at all.
It will be like a pretty basic function of like revenue growth, profitability. Some plus or minus based on like
Logan: Yeah.
Drew: But like if you can't draw, you know, if you draw the dot where you are today to the dot, you were where you would be as a public company. Um, and it's like where you're raised, where you're raising is like way off of that trend line.
Um, then you're in for it, then you're either leaving a lot [01:25:00] on the table or you're setting yourself up for, you're like opening 10 years of Christmas presents.
Logan: Most founders don't generally think they're going to end up on the wrong side of, of this stuff when they're raising. I've, I've found like, I'm sure Zuck doesn't, uh, It doesn't feel like he should have raised it at a lower price from Yuri Milner way back when, right? It's like, it's, uh,
Drew: not really practical, but like, I think it's more of the things where there's all, yeah, it's like, you know, you do have, you know, at some point the market will set the price, but then you do have some latitude to be like, are we really going to like juice it
Logan: wiggle around,
Drew: know? Um, and I think, you know, this, it was compounded by me sort of like elbowing it up because I thought like, why
Logan: Ten sounds good.
Drew: Yeah.
Logan: It does sound good.
Drew: And it went, you know, by, it actually wasn't that nuts, you know, in, by modern
Logan: Yeah,
Drew: But, you know, we were 250 million going to 400, so, you know, 40x revenue multiple going to 25. We're still very rich, but like, you know, still 60 [01:26:00] percent
Logan: exactly. There's, there will be way worse,
Drew: Yeah, but like, we spent the next 10 years digging out of that.
Logan: What, um, the other one
Drew: still below that market cap today.
Logan: just a nerdy, a nerdy one to end on, um, data centers.
Drew: Yeah, sure.
Logan: When did you decide to do all that stuff in house
Drew: Yeah, it was, um, the mid, somewhere kind of the mid teens, like probably like 2015, 2016. We, we'd been like sort of experimenting our way into it for a while. Yeah. We were like the one car, we were like the one car going the other direction on the highway, cause everybody else is migrating to the public cloud.
We had actually started on AWS and went off of it.
Logan: I actually don't even recall. What was the, was it purely costs at the time? Yeah.
Drew: was part of it. Um, I mean, it's kind of funny where, uh, I mean, there, there are actually a few issues. One was. Um, I mean, this one will sound kind of crazy, but like,
Logan: the, I
Drew: remember, like the public cloud itself was like pretty nascent and we were like users of S3. [01:27:00] Um, uh, Amazon S3 is one of the AWS services.
Um, the storage service. So we're big users of that. And then we were growing so fast. Actually, our biggest concern was like, are they going to keep, be able to keep up with us? And like, I don't know what our recourse is if like suddenly they can't keep up. Like. Like having a hotline and being like, Hey, can you guys please get our company back up?
Please, please, please. You know, I think we were in the, certainly in the top kind of echelon. I don't know if we were like number one, number one, but we were, you know, we didn't, we didn't want to be number one. Um, so there was a first question of like, is the public cloud even going to scale, um, as fast as we need it to?
And if something goes wrong, do we have, you know, what are our tools to actually make it better? If we don't control it, um, that ended up not being a big deal. Um, actually Amazon was a great partner. Um, and they scaled just fine, but like that was a concern. And then they did [01:28:00] get a little trickier. Cause like all these guys, the public cloud providers were also competing with us, like including Amazon.
So, you know, Google had their thing, Microsoft had their thing. Um, or, you know, OneDrive and Google drive, Amazon had like a storage product. And so. We were like pretty uncomfortable with like, all right, we're going to, so let me get this right. We're going to like put all of our keys in someone else's hands.
They're also competing with us. What could possibly go wrong? So it's more like an, a regret scenario. I felt like we'd look really dumb. Um, or just, you know, I felt like the responsible thing to do for our customers, make sure we were able to like, you know, protect their destiny as much as possible. Um, and then, but then third, yeah.
Cost performance, like the. The public cloud pricing like stopped kind of following the hardware cost curves down and like, yeah, they needed to take margin. And, but like, yeah, there's absolutely no guarantee. And probably it was unlikely that we would keep getting cost savings as the, um, or [01:29:00] we basically be relying on the kindness of strangers to like lower prices on the, um, on the, we, and we were only really using the commodity pieces of just like raw compute, EC2 and S3, like raw compute, raw hard drive space.
Um, and so. We were able to really build a tuned system for, um, just for our workload. And then you get not just costs, but a lot of performance advantages. You can like more vertically integrate across the stack. Um, and then that also is helpful for things, things like, as you have more compute that needs to go to the storage or machine learning, then there are like advantages like that, that you can translate into the world today.
Um, I mean, you have to be careful not to, we were very surgical in what we do in our kind of private cloud. Um, Um, because you can sort of end up in a situation where you're sort of re implementing the whole public cloud badly. Um, but it's worked super well for us, but we're, we're, we're a pretty unique case in that it's
Logan: So you, you're, you're glad you did it.
Drew: Super glad, super glad we didn't both, I'm super glad we started on the public cloud. It's really worked [01:30:00] out to be on our own infrastructure. We still get massive. Cost and performance, or cost savings. Um, and we were able to, you know, we go down to the, we co design our servers, you know, we go down to like the hard drive and the rack, like as far as, um, how all that infrastructure works.
And our teams have done like phenomenal things to just keep writing that down. So, therefore we still have like software margins, even though, you know, you'd think a lot of the storage piece is kind of a commodity.
Logan: I guess one last one before we hop anything. That you would say if you're a founder listening, um, we've touched on a bunch of different things that you've learned or thought of over the course of, you know, 17 years of, of doing this, but any other, uh, lessons or particular insights that you, you reflect on this, this journey you've been on that, um, might be useful to leave people with.[01:31:00]
Drew: I think like scaling, you know, the company was a lot harder than I feel like it should have been. Um, and we've talked about some of the books which are largely, some of them around management and things like that, but I think, um, you know, building on what we were saying about narrative culture and Amazon, I'm a big fan of Working Backwards, which is, um, a book that really kind of chronicles the development of all their scaling mechanisms.
And I think they had really good insight into, like, what goes wrong as you scale a bunch of people up. Um, and, uh, and narrative culture was just one part of it. So there's, you know, Like, how do you maintain a high talent density? And like, we'll keep raising the bar as you hire people. How do you do business reviews?
How do you, they basically have this operating system that's really powerful. And I was like, during some of the more difficult years, I was like, sick, I would give anything to like, sort of uninstall the operating system we have, uh, and install someone [01:32:00] else's. Um, and I wouldn't, Amazon, Google, whatever. Um, and it was kind of nuts to me that like, I, there was like no way to do that.
And even like implementing all these things. Uh, or these mechanisms are great around, um, the kinds of things we've been talking about, but they're very, it's still very like manual and human mediated and the only real technology is like still like a memo or an email or like a spreadsheet. So I think there's like a huge opportunity to, you know, in a world where now.
All of work is like fully digitized. Um, meaning like, you know, your AI agents or whatever can sort of see and hear and read and write, um, just like a human can. Um, that was, that's a very big change in the world from pre COVID where, you know, even no matter how smart your AI system is, it's still, if it doesn't have sort of like missing cell coverage, like if it doesn't know what happens in the real world or, or if the real action is happening there and the system is blind to it, then that'd be a problem.
But after COVID that's kind of fixed. Um, [01:33:00] and then with the large language model where it's like, you know, and then large reasoning models, multimodal models, whatever, but just like these large models, um, you basically have sort of cognition in a bottle, um, or sand that can think. And so I think that it'll be so much, I'm really excited about the kinds of organizations you can build.
And, you know, have, uh, I think we'll see like many 10x things in terms of like a team of 10 people can do what used to take a team of 100. Um, and either make like small teams perform big or make, you know, big teams feel small. Um, I think the whole practice of management is going to be completely different where, um, a lot of the repetitive aspects are going to be automated.
And, um, and so. I'm super excited about like opening my laptop in 2030 and, uh, having a much more sane and calm and focused experienced. And, and hopefully we're going to have a lot of stuff well before we basically turned Dropbox [01:34:00] into this lab for distributed work after COVID.
Logan: Cool. Drew, thanks for doing this.
Thank you for joining this episode of the Logan Bartlett show with co founder and CEO of Dropbox, Drew Houston. If you enjoyed this discussion, we really appreciate it. If you subscribed on whatever podcast platform you're listening to us on as well as shared with anyone else that you think might find it interesting.
We're continuing to try to grow the show. And so we really appreciate anyone's support and continuing to help us do so. Thanks everyone for joining and we'll see you next week with another great guest in the Logan Bartlett show. Have a great weekend.