This is a brief story, told by Dan Rose, of how betting on a risky choice and almost bankrupting Amazon in the process, 20 years ago, turned out to lead the way to one of the key global digital assets of our time: Amazon Web Services.

In 2000, the internet bubble popped. Capital markets dried up and Amazon was burning $1 billion a year, with expensive Sun servers at their data centers as their largest expense.

The company spent a year replacing Sun servers with HP/Linux machines, which formed the foundation for AWS. Linux kernel had only been released in 1994, the same year Amazon started, so it was a novel and risky approach at the time.

The transition made Amazon freeze all new features for over a year, causing a deceleration in revenue growth. Amazon came close to going bankrupt around that time. With the infrastructure completed, Amazon then decided to rent it out as the company only needed high capacities during peak retail seasons.

Amazon, founded in 1994, is the world’s largest online retailer of books, clothing, electronics, music and other goods and, in 2019, earned net sales revenue of over 280 billion U.S. dollars.
Amazon’s service offerings have increased over the years and now include a retail products branch, retail third-party seller services, retail subscription services, Amazon Web Services (AWS) as well as other related services.

Due to global interest in cloud services and the disruption they are causing across various sectors, AWS has played an increasingly important role in the cloud services industry and has become an important revenue earner for Amazon. In 2013, AWS earned revenues of just over 3 billion U.S. dollars, a number which has since ballooned and sits at over 35 billion U.S. dollars as of 2019.
(Source: Statista)

Dan Rose “spent 20 years in leadership roles at both Amazon and Facebook. Dan was at Amazon from 1999-2006, where he managed retail divisions and helped incubate the Kindle. As a VP at Facebook from 2006-2019, Dan helped grow the company from 130 employees to 35,000 and was responsible for early monetization strategy, business development, M&A and community operations.”
(Source: LinkedIn)

In a set of tweets, he recalls how it all happened.

I was at Amzn in 2000 when the internet bubble popped. Capital markets dried up & we were burning $1B/yr. Our biggest expense was datacenter -> expensive Sun servers. We spent a year ripping out Sun & replacing with HP/Linux, which formed the foundation for AWS. The backstory:

— Dan Rose (@DanRose999) January 8, 2021

My first week at Amzn in ’99 I saw McNealy in the elevator on his way to Bezos’ office. Sun Microsystems was one of the most valuable companies in the world at that time (peak market cap >$300B). In those days, buying Sun was like buying IBM: “nobody ever got fired for it”

— Dan Rose (@DanRose999) January 8, 2021

Our motto was “get big fast.” Site stability was critical – every second of downtime was lost sales – so we spent big $$ to keep the site up. Sun servers were the most reliable so all internet co’s used them back then, even though Sun’s proprietary stack was expensive & sticky.

— Dan Rose (@DanRose999) January 8, 2021

In 2000, brand new Sun servers started appearing on eBay for 10 cents on the dollar as VC-backed start-ups went out of business (this was pre-AWS when you had to roll your own datacenter). Amzn could have negotiated a better deal with Sun, but Jeff chose a more radical approach.

— Dan Rose (@DanRose999) January 8, 2021

Amazon’s CTO was Rick Dalzell – ex-Walmart, hard-charging operator. He pivoted the entire eng org to replace Sun with HP/Linux. Linux kernel was released in ’94, same year Jeff started Amzn. 6 years later we were betting the company on it, a novel and risky approach at the time.

— Dan Rose (@DanRose999) January 8, 2021

Product development ground to a halt during the transition, we froze all new features for over a year. We had a huge backlog but nothing could ship until we completed the shift to Linux. I remember an all-hands where one of our eng VPs flashed an image of a snake swallowing a rat

— Dan Rose (@DanRose999) January 8, 2021

This coincided with – and further contributed to – deceleration in revenue growth as we also had to raise prices to slow burn. It was a viscous cycle, and we were running out of time as we ran out of money. Amzn came within a few quarters of going bankrupt around this time.

— Dan Rose (@DanRose999) January 8, 2021

But once we started the transition to Linux, there was no going back. All hands on deck refactoring our code base, replacing servers, preparing for the cutover. If it worked, infra costs would go down by 80%+. If it failed, the website would fall over and the company would die.

— Dan Rose (@DanRose999) January 8, 2021

We finally completed the transition, just in time and without a hitch. It was a huge accomplishment for the entire engineering team. The site chugged on with no disruption. Capex was massively reduced overnight. And we suddenly had an infinitely scalable infrastructure.

— Dan Rose (@DanRose999) January 8, 2021

And this is where genius strikes!

Then something even more interesting happened. As a retailer we had always faced huge seasonality, with traffic and revenue surging every Nov/Dec. Jeff started to think – we have all this excess server capacity for 46 weeks/year, why not rent it out to other companies?

— Dan Rose (@DanRose999) January 8, 2021

Then something even more interesting happened. As a retailer we had always faced huge seasonality, with traffic and revenue surging every Nov/Dec. Jeff started to think – we have all this excess server capacity for 46 weeks/year, why not rent it out to other companies?

— Dan Rose (@DanRose999) January 8, 2021

Amazon AWS was born!

Then something even more interesting happened. As a retailer we had always faced huge seasonality, with traffic and revenue surging every Nov/Dec. Jeff started to think – we have all this excess server capacity for 46 weeks/year, why not rent it out to other companies?

— Dan Rose (@DanRose999) January 8, 2021

Then something even more interesting happened. As a retailer we had always faced huge seasonality, with traffic and revenue surging every Nov/Dec. Jeff started to think – we have all this excess server capacity for 46 weeks/year, why not rent it out to other companies?

— Dan Rose (@DanRose999) January 8, 2021

Amzn nearly died in 2000-2003. But without this crisis, it’s unlikely the company would have made the hard decision to shift to a completely new architecture. And without that shift, AWS may never have happened. Never let a good crisis go to waste!

— Dan Rose (@DanRose999) January 8, 2021

PS: Amzn recently spent years ripping out Oracle, something few have attempted. It takes muscle to do hard things, and muscle gets built by doing hard things. The best companies look at every challenge as an opportunity and engrave that mindset into their culture.

— Dan Rose (@DanRose999) January 8, 2021

Thank you Dan for sharing this story, truly inspirational!