Who would be crazy enough to compete head-on with AWS?
The question was almost as obvious seven years ago than it is today. Yet in just a few years since its founding, Digital Ocean, a cloud infrastructure startup based in New York with data centers around the world, has managed to build a very impressive and fast-growing business, successfully competing with the giants of cloud computing.
Ben Uretsky, co-founder of the company (with his brother Moisey and 3 others) and its CEO from 2011 to 2018, stopped by for a chat at Data Driven NYC to tell the story of the company and share some lessons learned.
Here’s the video, and below are my notes from our great chat.
- Ben and Moisey had started in 2003 ServerStack, a managed hosting business (dedicated servers + managed support service)
- ServerStack was a bootstrapped, profitable business, had 1,000 servers around the world
- In part because they were losing too much business to RackSpace, they decided to transition to cloud computing; started writing code towards a new platform in 2011, and formally incorporated Digital Ocean (as a brand new, separate entity) in March 2012.
- $200M in revenues (actual revenues, not ARR) in 2018
- 500 employees
- 1M total customers (all time); 500,000 currently active
Positioning and differentiation:
- “Simplest, easiest to use cloud out there”
- Focused on startups and SMBs that are looking for an alternative to the big cloud providers
- Has scaled with some of their customers, the largest of which spend over $1M a year on Digital Ocean
On the power of constraints:
- Positioning (around simplicity) has not changed since the beginning of the company
- Requires tremendous amounts of empathy to understand what the customer is trying to achieve
- It’s actually easier to build a complex solution… much harder when you take ownership over the complexity on behalf of your customer
On selling to developers:
- You have to pick a side – you are either on the developer side or the enterprise side
- We chose to help developers with their education and helping them solve problems… basically giving them something of value without charging for it like free content
- Built a community of content, with a couple thousand tutorials on how to leverage open source
- Get 3M visitors a month
- Have an NPS score of 70… since developers the product, they are likely to recommend it to a friend… referrals and organic acquisition are the largest source of sign ups
- Try to bootstrap and make it on your own as much as possible – debt, credit cards… and focus on the customer and the value first
- Two reasons to go to a VC: 1) you’ve proven out a business, have some traction and need to accelerate, and 2) you have a really audacious idea and there’s no other way of bringing it to life without venture capital
- A proven business model will also help accelerate the fundraising process… for each round we spoke to 15 firms… I have heard of startups that spoke to 150…
- Hacks for bootstrapping: figure out how to use other people’s money! We took up law firm Gunderson on their offer to delay legal fees… you can raise money from friends and family… leverage the credits offered by cloud providers like AWS and Digital Ocean through their startup/incubator programs for the first couple of years of your business
- We have raised more debt ($300M) than venture capital ($123M)… we use debt to buy all our storage and servers, much cheaper than financing through equity in terms of dilution
- Recruiting becomes your #1 job when you’re scaling… there were months when I spent more than 50% of my time recruiting
- In 2010, was hard to recruit talent in NYC… now there are more companies than ever in NYC
- Other key aspect of scaling is culture: when you have over 100 employees and you onboard more people, how do they know what to do and how to make a decision? That’s what culture is all about
- Culture hould be a key foundational element of any company… We messed up our culture initially… ended up having lots of issues… we hired a head of people and codified our culture but should have done it much earlier
On technology trends relevant to Digital Ocean:
- Virtual machines is starting to slowly phase out… containerized applications are becoming more dominant… they run within VM but that’s a legacy from the past and will evolve
- Serverless is accelerating
- Apps are going to become more stateless
- Those abstractions bring you more scale but they also bring more complexity
- Cloud providers are going to have a hard time figuring out the multi-tenancy part of the equation
- We need to deliver GPU for AI/ML and it’s on our roadmap
- We haven’t tackled edge computing yet, we’re a highly centralized data center as of now… we haven’t had much requests yet… we’re going but we’re not there yet.
On remote work:
- Even as a small company, we had remote employees… We thought remote was a differentiator for us… was harder to find people in NYC (especiallyin 2010) that knew about infrastructure and had seen hyper-growth
- Half of our team (250) is in NYC (half is remote).
- Our People Operation team almost places remote employees “almost ahead” of our employees in the office (how do they participate in events, etc.)
- Lots of asynchronous communication (Slack, email), but also plenty of opportunities to come together in town halls