Big Tech Bets

A very brief history of AWS and Android

Big Tech Bets

Let me preface this by saying that Bill Gurley’s main point was different than what I’m nitpicking at here. I’m referencing a small comment he made at the end of his thread. So don’t take this as an attack on his thread, but merely a good opportunity to delve into some fun tech history.


The other day Bill Gurley published this tweet thread:

In this thread he says:

“If this plays out like I think it might, it could go down in history with what I believe are the two most impressive internal strategic launches during my career. Amazon’s AWS launch, & Google’s Android launch. It could potentially be that important.”

That quote felt like a bit of hyperbole. I mean, really? Shopify launching a streamlined checkout system is a strategic launch on par with AWS or Android? It certainly doesn’t feel that way.

I could be missing something blindingly obvious here, but a streamlined checkout and payment system for a platform as large as Shopify is pretty much a no-brainer. This is what every digital wallet since Paypal has been trying to do. It doesn’t take a lot of strategic thinking these days to know that making it easier for merchants to checkout customers, and getting the platform a cut, is a smart move. That is why they launched Shopify Payments back in 2013 purely for helping merchants take payments, and then in 2017 upgraded it to a Shopify-wide accelerated checkout. It was recently rebranded it to Shop Pay, and they have also refined the experience over the year.

Shop Pay is successful in part because the experience is great, but the lion’s share of that success is due to the fact that Shopify has grown absolutely enormous. According to Statista, Shopify accounts for almost 30% of all U.S. ecommerce websites. That figure is pretty unbelievable. They deserve much praise for this success, but Shop Pay definitely doesn’t seem to me like a strategic play on par with AWS or Android. Not by a long shot.

That got me thinking, looking at today’s world, it feels inevitable that AWS and Android would become enormous successes. At the time of their launch, and for quite a while afterwards, almost everyone thought these products were doomed to failure. But each of them wound up dominating their market.

Their stories are all very different, and these are just the highlights, but I hope you enjoy this trip down memory lane.

Amazon Web Services – AWS

It was the early 2000’s and Amazon’s revenue was a comparatively paltry 3 billion dollars a year (It was 386 billion in 2020). Amazon was growing by leaps and bounds and it was also an age where other retailers were just starting to realize that maybe this whole ecommerce thing wasn’t a passing fad and they needed to get their inventory up on the tubes.

While Amazon’s revenue was growing like crazy, their investors were still terrified because they were burning an enormous amount of cash, and had several billion dollars in debt. Keep in mind, Amazon had launched in 1995 and wouldn’t turn any profit until late 2001. Investors were nervous and its stock was suffering.

But Amazon had done one thing really well, it had figured out how to run a large scale ecommerce operation, and other companies needed those skills. Because of this Amazon had embarked on building out a product called Merchant.com, which was Amazon’s ecommerce platform as a service, allowing other retailers to build out their own ecommerce stores.

It turned out though, that building out these platforms for other large retailers was much more painful than they had originally thought, and a big portion of that came from having to set up all of the same basic infrastructure for every project. The infrastructure was all having to be purchased, racked, and set up in a very inefficient and manual way; taking many months for every project.

At the same time, as Amazon scaled they were starting to move towards a more service oriented architecture internally, and as they were spinning up more teams and trying to launch more services, they were feeling this same infrastructure pain. On top of all of this, Amazon’s growing ecommerce business was starting to experience extreme seasonality in their sales as they started branching out from books into products that were bought more around the holidays. Keeping a huge amount of extra infrastructure around in order to deal with seasonal spikes just felt wasteful (and if Amazon was anything, it was frugal).

And so the idea of building a service with the basic building blocks they needed, of storage and compute, came to be. It was far from being a sure thing, “utility computing” as it was often referred to at the time, had been predicted for many decades but had failed to materialize in any meaningful way.

Amazon Web Services launched its first product, Simple Queue Service (SQS) in late 2004, and then two years later it launched Elastic Compute Cloud (EC2) and Simple Storage Service (S3) to an often lukewarm response. The early days of AWS were rough, the service was incredibly limited and somewhat unreliable, but it was showing enough growth that Amazon continued to invest. To highlight some of this early uncertainty, it wasn’t until 2010 that Amazon would announce that it had moved much of Amazon.com over to run on AWS.

It is hard to say how fast AWS grew in these early years, because it wasn’t until 2012 that Amazon even split AWS revenue out from Amazon.com. But by 2012 AWS was generating about 1.5 billion in revenue and growing quickly. Microsoft launched Azure in 2010, four years after the initial launch of EC2. Google launched App Engine in preview in 2008, but it wasn’t until 2010 that it launched Cloud Storage and 2012 that it launched Google Compute Engine to directly compete.

AWS was just one of many huge bets Amazon took in those days, but it is almost certainly one of their most successful.

Android

Android Inc. was founded in 2003 and actually started out as an operating system for digital cameras, but really struggled to find anyone who was interested. One of the founders, Andy Rubin, had previously founded Danger, the company who famously developed the T-Mobile Sidekick. Because of this background he was convinced by friends and investors to pivot into a mobile phone OS.

The idea was simple. Create a free mobile phone OS that they could give away to carriers, and then they would sell services to the carriers. For example, instead of making the carriers operate an online contact sharing service, Android Inc. would operate that service and charge the carriers for it. At the time, all of the other mobile phone operating systems were either proprietary (Symbian/Blackberry OS) or very expensive to license (Windows Mobile).

However, even after this pivot, they still struggled mightily to find investment and by 2005 were getting desperate for funds. At the time, the idea of giving their core product away for free was still seen as too risky for many investors.

In early 2005 they finally received a few term sheets, but at the same time Google had caught wind of their project and had brought them in for a few meetings. Apparently Larry Page had been a big fan of the Sidekick and he was interested in producing the best mobile phone out there. They ended up selling to Google for reportedly about $50 million after Google convinced them that they wouldn’t need to worry about figuring out how to monetize the product, they just wanted to build the best phone out there.

Google allowed them to hire all of the people they needed to build the mobile OS, with the team reaching about 100 people by the time the first version was finished. The team worked on building out the mobile operating system secretly inside of Google for years, and in 2007 had an internal working prototype that very much resembled the Blackberry. Then in early 2007 Steve Jobs unveiled the iPhone and changed everything.

Android launched in late 2008, and it too received a tepid reception. But a certain portion of the population, mainly technical folks, were excited. It was an open source OS on a phone! But the actual device was pretty mediocre, especially compared to the recently released iPhone 3G.

It took a while before Android finally got some real traction. The project was always a long shot, many of the engineers on the team were coming from Be and Palmsource and so they knew how hard it was to make an OS a commercial success. It is summed up well by Evan Millar, an early member of the Android team, in the book Androids:

We didn’t know if it was going to totally flop or if it was going to work. When it worked, I think people were as surprised as they were excited.

Evan Millar

At this point Android is such a huge success that it is hard to imagine that at one point it was a huge bet, but we have to remember that even well established platforms like Symbian and Windows Mobile are now gone, while Android is on top of the heap.

Who else deserves to be here?

After reviewing the history of AWS and Android, do you think Shop Pay is on par with either of these? I still don’t think so, but to be fair, Bill probably sees something I don’t.

As a software engineer though, there is something incredibly romantic about these kinds of long-shot products. Products that start off with everyone wondering what they are thinking, but end up changing the world and dominating their industries or creating entirely new ones.

What other products of the last two decades deserve to be in this list?

Loved the article? Hated it? Didn’t even read it?

We’d love to hear from you.

Reach Out

Leave a comment

Leave a Reply

Your email address will not be published.

More Insights

View All