When PLG Meets Enterprise: Drew Houston on Building Dropbox from Viral Growth to $2.5B in Revenue

This post is part of our ERC 2025 Recap series. Read our full recap post here.

Drew Houston, Founder and CEO of Dropbox, sat down with Michael Grinich, Founder and CEO of WorkOS, at ERC 2025 to tell the story of how Dropbox came to be and how it came to scale to serve 18M paying customers.

The Accidental Enterprise Company

Dropbox didn't set out to become enterprise software. The company famously started when Houston forgot his thumb drive on a bus to New York and decided to solve his own problem.

What emerged was a consumer product that spread through word-of-mouth and clever growth hacking—most notably a referral program that rewarded users with free storage for inviting friends.

But enterprise adoption happened organically from day one.

Houston recalls getting support emails from unlikely users: farmers coordinating tractors in their fields, anti-poaching teams in Africa tracking operations with intermittent Wi-Fi, astronomers on different continents finding new planets together. The common thread? People were bringing Dropbox into their work environments because it solved real problems, regardless of IT policies.

From Shadow IT to Strategic Tool

The first enterprise "feature" Dropbox built wasn't really a feature at all—it was shared billing. Teams were hacking together multiple individual accounts, each requiring a separate credit card. "Can you just let us pay for everyone at once?" they asked.

This humble beginning revealed the fundamental inversion of traditional enterprise software: Dropbox was already deployed before anyone purchased it. The value was proven, the users were already onboarded, and IT departments just wanted to manage what was already there.

This created an unusual leverage point. There were no steak dinners, no golf outings, no six-month procurement cycles. The software was already indispensable. IT teams weren't being sold—they were being given the tools to embrace what their employees had already chosen.

The Holy War Inside Dropbox

Not everyone at Dropbox believed in the enterprise opportunity. Houston describes it as a "holy war" over the company's identity.

On one side stood the consumer advocates. The founding team was young, consumer-minded, and had little context for what enterprise software even meant. They wanted to build features for photo sharing and personal productivity. They saw traditional enterprise vendors as purveyors of "dark patterns"—opaque contracts, vaporware sales pitches, and user experiences that prioritized procurement over people.

On the other side was an emerging recognition: their happiest, most engaged users were at work. Dropbox had become mission-critical infrastructure for companies large and small. And those IT admins asking for visibility and control weren't being unreasonable—they were being responsible stewards of their organization's data.

Houston's argument for commitment won out: "We're not actually providing a good product experience if people are blocked from using Dropbox by their IT admin. We have a real opportunity to do this differently. We can land in companies really cheaply and figure out how to hold a really high bar for UX while also reaching business audiences."

The Mechanics of Viral Enterprise Growth

The path to 700 million registered users wasn't linear, and it certainly wasn't obvious in the early days.

Houston and his co-founder initially tried everything: traditional PR and marketing (didn't move the needle), antivirus-style distribution partnerships (wasted time with no leverage), and promotional videos on Hacker News and Reddit (worked great until they didn't scale).

What finally worked was viral growth mechanics borrowed from an unlikely source: epidemiology.

The team studied how infections spread—the R-naught of disease transmission mapped perfectly to software adoption.

They built viral loops where sharing a folder or link exposed new users to Dropbox, who then entered the top of the funnel. The key was optimizing conversion rates at every stage: sign-up, activation, usage, and sharing.

The 80 Problems Between Sign-Up and Success

Early on, only 25% of sign-ups actually downloaded Dropbox, installed the app, and put a file in. That meant 75% of potential users were falling out somewhere in the funnel.

To figure out why, Houston did what any good engineer would do: he went on Craigslist, paid people $50 to come to the office, and watched them try to use Dropbox. The team identified 80 friction points—places where users got confused, failed to complete installation, or didn't understand Dropbox's mental model.

After fixing those 80 issues, the onboarding success rate jumped from 25% to 65%.

Similar optimizations to sharing flows created compounding effects. The result? Linear growth became exponential. Dropbox went from one million to 10 million to 100 million registered users, 10x-ing annually.

The Self-Serve Ceiling

Eventually, this viral motion hit a ceiling. Once Dropbox reached a certain level of adoption within an organization, IT would get involved with one of two messages: "Give us the version we can buy" or "We're kicking you out because this is shadow IT."

This was the moment many product-led companies struggle with—the transition from self-serve to sales-assisted, from individual users to organizational buyers. Dropbox's answer was to build the enterprise features IT needed while maintaining the frictionless experience users loved.

That meant investing in audit logs, SSO, authentication infrastructure, and the entire list of enterprise requirements that SaaS companies need but users never ask for. Dropbox had to build all of it from scratch—the kind of infrastructure that tools like WorkOS now provide out of the box.

The AI Pivot: Back to First Principles

Eighteen years into running Dropbox, Houston is doing something most CEOs don't: debugging CUDA kernels and teaching himself machine learning.

Why? Because he's frustrated again.

"I started Dropbox because I kept forgetting my thumb drive," Houston explains. "But really, the bigger question is: why is it so difficult to find, organize, and share my most important information?"

In 2007, that information was scattered across devices and operating systems. The solution was syncing files to the cloud. In 2025, the problem looks different but feels the same.

The 100-Tab Problem

What used to be 100 files on your desktop is now 100 tabs in your browser. Work tools have proliferated—everyone has Google Docs, Dropbox, Airtable, Slack, email, and a dozen other purpose-built applications. Each gives us new superpowers, but the overall system has regressed.

At home, you can Google anything and search all of human knowledge from one box. At work, you have 10 search boxes that each search 10% of your stuff. You run a couple of searches, give up, and Slack someone to find it for you.

"We're the frogs in boiling water," Houston says. "You don't even notice it anymore."

The Context Gap

The AI revolution hasn't solved these problems—it's added to them. Every chat conversation starts from scratch. You're manually hauling in documents one at a time, cutting and pasting context like it's 2005. Meanwhile, the industry is spending trillions of dollars making models win math olympiads when there's an obvious bottleneck: the models are only as smart as the context you provide them.

This insight led to Dropbox Dash, an AI assistant and universal search engine that connects to all your work apps—Google Drive, email, Dropbox, Slack—and surfaces everything in one place. The pitch is simple: close the context gap to make AI actually useful at work.

Houston's personal experiments validated the approach. He built a toy search engine using semantic search and vector databases that indexed all his documents. "It returns results faster than I can type, and it's not that hard to build something that gets way better results than keyword search," he says. "I'm like, oh my God, normal people are going to have something like this someday. But right now, zero people have this. Big opportunity for Dropbox."

The CEO Who Still Codes

It's rare to find a CEO who's been at the helm for 18 years and still talks about GPU tensor cores and kernel debugging. Houston's hands-on approach is deliberate.

"I think it's really hard to do your job if you don't understand technology," he explains. "There's so much noise in AI and hype. You can't just be right about what the end state is going to be—you also have to be right about the timing."

His method for staying sharp? Keep coding. Dive deep into algorithms. Understand everything from the GPU level up. Not just because he enjoys it (though he does), but because it informs where Dropbox should place its bets.

The Personal Growth Imperative

Houston frames founder development around a simple principle: keep your personal growth curve ahead of the company's growth curve.

As a founder-CEO, your job changes every six to 12 months. In 2007, he and his co-founder Arash just coded because there was nothing else to do. Then they raised money and hired engineers. Then they needed users, so they learned marketing. Then they needed a business model and started charging customers.

"Your plate keeps expanding every day," Houston says. "You have to do everything you were doing yesterday even better while also spinning a bunch of new plates."

His toolkit for navigating this? Reading voraciously (he recommends "High Output Management" by Andy Grove and "The Effective Executive" by Peter Drucker), learning from founder communities at different stages, and always asking: "A year from now, five years from now, what will I wish I had been learning today?"

The Questions Nobody Asks

When pressed about what interviewers don't ask him enough, Houston didn't hesitate: fundamentals.

"How far can you scale a company that is losing $10 for every dollar of revenue?" he asks. "At some point, gravity kicks in."

His advice for founders in the current AI frenzy? Study history. Every era of computing has followed similar patterns of hype and disappointment, followed by sustainable growth. Understanding these cycles helps demystify what's happening now.

He also pushes founders to evaluate their business the way Warren Buffett would—to understand company valuation beyond the current moment's exuberance. "It was way easier to raise money at a $10 billion valuation without having any idea what that meant," Houston admits. "We were humbled not long afterwards."

The key lesson? Don't get over-extended. YC's Paul Graham has a maxim about being "default alive"—not relying on the kindness of strangers to keep your company afloat. "Going off a ski jump at 60 or 80 miles an hour can be fun," Houston says. "Going off at 500 miles an hour causes you to vaporize."

The Mission That Emerged

For years, Dropbox didn't have a formal mission beyond "keep your files in sync." Eventually, Houston realized: we did it. The files are in sync. Now what?

The answer came from his own frustrations as a CEO. His brain felt stuck in first gear—meetings all day, emails all night, repeat. He couldn't remember the last time he'd done something creative. And he noticed the same pattern across his organization: simultaneously overworked but underutilized.

The cognitive experience at work is fundamentally broken. Our tools are distracting and overwhelming. Organizations scale dysfunctionally. These aren't just Dropbox problems—they're universal constraints on what humans can accomplish.

Houston realized this was the moonshot every other moonshot depends on. Even putting someone on Mars, as a SpaceX engineering director told him, requires "a lot of emails and a lot of files."

So Dropbox established a new mission: design a more enlightened way of working.

"Technology is this force multiplier, this enabler, which it is," Houston says. "But if you look at it from the other perspective, technology is also the limiting factor. We can only be as good as our tools."

After 18 years, what keeps him motivated is painting on a bigger canvas—solving problems at planetary scale, looking over someone's shoulder in a Starbucks and seeing that little Dropbox logo on their laptop.

That's the thing about building products people love: the mission never gets old. Even when you're debugging CUDA at 3 AM.