Sep 6, 2022 · 2h 49m · acquired
Amazon Web Services · Acquired
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of Acquired, Ben Gilbert and David Rosenthal analyze the origins, technical architecture, and economic power of Amazon Web Services (AWS). They detail how AWS evolved from an internal infrastructure overhaul into a dominant global cloud utility that fundamentally reshaped modern enterprise computing and startup formation.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Ben and David hold 98.1% of the talking time here. How this is scored →
Ben and David as informed peer 6.2 Guest teaching 0.0 Guest disagreement 0.0 Ben and David pushing back 0.5
speaking balance: gold is Ben and David, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 5:00 Ben Miller criticizes traditional 20% VC carry
In the opening sponsor segment, Ben Miller aggressively challenges the venture capital establishment, arguing traditional VCs do only 2% of the work while capturing 20% of the upside.
Hardest push from Ben and David ▶ 2:38:00 Ben pushes back on evaluating Amazon solely on market capBen challenges the valuation assumptions of Amazon's $1.5T market cap, pressing on whether a perpetual reinvestment machine will ever return cumulative capital equivalent to its valuation.
Biggest teaching moment ▶ 15:40 Disproving excess capacity with technical architecture factsBen explains the architectural impossibility of sharing monolithic DEC/HP server infrastructure with external tenants during seasonal lulls.
Ben and David hold their own ▶ 1:38:00 Synthesizing why Microsoft and Google missed initial IaaSBen articulates in precise technical and strategic detail how Microsoft and Google over-indexed on future PaaS runtimes while completely missing the developer demand for basic primitives and lift-and-shift.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Ben and David as informed peer | Guest teaching | Guest disagreement | Ben and David pushing back | Why |
|---|---|---|---|---|---|---|
| Show Housekeeping and Episode Framing | 5 | 0 | 0 | 0 | Ben and David establish the episode thesis, framing AWS as more foundational and profitable than Amazon's retail business while introducing the four competing origin stories. | |
| Debunking Origin Story 1: The Excess Capacity Myth | 7 | 0 | 0 | 1 | Ben and David debunk the myth that AWS was built simply to sell off-peak Q4 excess server capacity, citing technical architecture constraints, Linux migrations, and Werner Vogels' Quora post. | |
| Origin Story 2: Tim O'Reilly, Web 2.0, and Associates APIs | 6 | 0 | 0 | 0 | The hosts detail Tim O'Reilly's visit to Bezos, explaining how Web 2.0 principles and Amazon Associates catalyzed the first programmatic XML/REST APIs and the initial 2002 developer conference. | |
| Origin Story 3: Monolithic Architecture Bottlenecks and Andy Jassy as TA | 7 | 0 | 0 | 1 | The co-hosts explore Amazon's early monolithic codebase bottleneck, code freeze restrictions, and Andy Jassy's appointment as Bezos's first technical assistant shadow. | |
| Bezos's Hardened API Mandate and Service-Oriented Architecture | 8 | 0 | 0 | 0 | David and Ben discuss Steve Yegge's famous rant detailing Bezos's strict hardened API mandate and the transition toward service-oriented architecture (SOA). | |
| Sponsor Segment: Pilot Financial Services | 1 | 0 | 0 | 0 | Standard sponsor read for Pilot Financial Services, highlighting back-office accounting integrations with API-first financial platforms. | |
| From Internal IT to Commercial Offering: Jassy's 2003 Proposal | 6 | 0 | 0 | 0 | The hosts analyze Jassy's pivotal 2003 six-page memo proposing cloud IT infrastructure and requesting a 57-person team to turn IT resources into APIs. | |
| Launching Cloud Primitives: S3, EC2, and Building the Internet OS | 7 | 0 | 0 | 1 | Ben and David highlight the launches of S3 in 2006 and EC2, contrasting the unopinionated cloud primitive strategy against traditional monolithic software stacks. | |
| Origin Story 4: Benjamin Black, Chris Pinkham, and Cape Town EC2 | 7 | 0 | 0 | 1 | The hosts examine the fourth origin story involving Benjamin Black, Chris Pinkham, and the decentralized Cape Town team developing EC2 in relative isolation. | |
| The Developer Revolution: Frictionless Self-Service and Startup Fuel | 6 | 0 | 0 | 0 | The hosts review how self-serve credit card billing revolutionized startup creation by replacing heavy CapEx procurement cycles with near-zero marginal experimentation costs. | |
| Cloud Adoption Waves: Startups, Enterprise Lift-and-Shift, and Netflix | 6 | 0 | 0 | 0 | The discussion covers cloud adoption stages from early-stage startups to enterprise lift-and-shift, anchored by Netflix's migration and public endorsement. | |
| Why Tech Giants Whiffed: IBM, Oracle, Microsoft, and Google | 8 | 0 | 0 | 1 | Ben and David dissect why incumbents failed to respond: Oracle and IBM protecting high-margin licensing, Microsoft defending Windows Server, and Google pushing an overly abstracted PaaS. | |
| Enterprise Sales Expansion, Database Dominance, and the $100B Backlog | 7 | 0 | 0 | 0 | The hosts analyze AWS's database expansion (RDS, Redshift, Aurora), data gravity dynamics illustrated by Snowball/Snowmobile, and the monumental $100B contractual revenue backlog. | |
| Financial Breakout, Custom Silicon, and Machine Learning Strategy | 7 | 0 | 0 | 0 | Ben and David cover the 2015 financial breakout, custom silicon acquisitions (Annapurna Labs, Trainium, Inferentia), and why co-locating ML workloads near existing stored data defends AWS. | |
| The Whiff on Data Warehouses: Snowflake vs. Redshift and Service Sprawl | 7 | 0 | 0 | 1 | The hosts critique AWS's major misstep with Redshift versus Snowflake, attributing Snowflake's victory to developer UX focus while AWS fought legacy Oracle paradigms. | |
| Sponsor Segment: NZS Capital on Margin of Safety | 2 | 0 | 0 | 0 | Sponsor segment discussing NZS Capital's white paper on redefining the margin of safety through management adaptability rather than static valuation metrics. | |
| Seven Powers Analysis and the Multi-Cloud Battleground | 8 | 0 | 0 | 1 | Applying Hamilton Helmer's 7 Powers framework to AWS, highlighting immense scale economies, switching costs, and counterpositioning against legacy vendors. | |
| Playbook Takeaways: Economies of Scale, Price Cuts, and Asymmetric Upside | 8 | 0 | 0 | 1 | Ben and David synthesize core business playbooks: proactive price cuts to fuel CapEx confidence, capital lease structures, and Bezos's long-tail asymmetric upside philosophy. | |
| Valuation, Final Grading, and Amazon's Day One Philosophy | 7 | 0 | 0 | 1 | The hosts award an A+ grade while debating long-term cash flow generation and the sustainability of Amazon's perpetual 'Day One' reinvestment strategy across a $1.5T valuation. | |
| Carve Outs: Moon Knight, Obi-Wan Kenobi, and John Carmack | 3 | 0 | 0 | 0 | Standard episode wrap-up and carve-outs discussing Marvel's Moon Knight, Obi-Wan Kenobi, and Lex Fridman's extensive interview with John Carmack. |
Statements from this episode (31)
Gilbert: AWS operating profit exceeds Amazon retail despite being 15% of revenue
“AWS's revenue is only about 15% the size of Amazon's massive retail business, but their profits, or the operating income, to be specific, from AWS Are in total the same, if not more than their e-commerce store. I think it's the case that every year since 2015,…”
Gilbert: Early Amazon enforced a two-month holiday code freeze
“So much so that for the first, at least five years of the business, there was a rule in November and December that you could not commit new code to production.”
Gilbert: Migrating to Linux and HP servers saved Amazon from bankruptcy
“I mean, in 2000, they almost went out of business because they were so tight on cash and they were spending so much on infrastructure that this sort of moved to the open source ecosystem and doing a massive rewrite of all of Amazon.com to run on Linux and run …”
Rosenthal: AWS was built to scale, not to monetize excess seasonal capacity
“The excess capacity story is a myth. It was never a matter of selling excess capacity. Actually, within two months after launch, AWS would have already burned through the excess amazon.com capacity. Amazon Web Services was always considered a business by itsel…”
Gilbert: Public APIs replace custom business development agreements
“Being a web two point O company means that you can do business with other companies without needing like a BD agreement in place. You don't actually need a partnership agreement. You basically can just publish your API. You can say pay as you go and here's how…”
Rosenthal: Only eight people attended Amazon's 2002 AWS launch conference
“They hold a conference for developers in 2002. A total of eight people attend the conference. They announced to the world the launch of this new division within Amazon. That is called Amazon Web Services.”
Rosenthal: Original 2002 AWS was an Associates catalog API, not cloud infrastructure
“This is called Amazon Web Services, but it is not. Cloud-based IT infrastructure. It's other developers using the amazon.com product catalog, and indeed it lives. Amazon Web Services lives within the Amazon Associates program, and that is run by a guy named Co…”
Rosenthal: 'Cloud infrastructure' originated at General Magic for mobile back-end services
“It started At General Magic. How crazy is that? The Apple spin out that invented the iPhone, 20 years before the iPhone, as part of what they were doing, they also wanted to have, you know, the internet sort of barely existed, so I don't think they thought of …”
Rosenthal: Bezos created the shadow role to save Andy Jassy from layoffs
“And then, 2002 1001 dot com crash. Amazon acts as the whole marketing department. Cause you know, we're not doing ads anymore. We gotta get profitable. And Jassy was gonna get laid off with the whole department, but Jeff liked him. And so Jeff said, I'm gonna …”
Gilbert: Microsoft treats product management as marketing and program management as engineering
“Microsoft considers product management a marketing function. So it's owned in the marketing org and is much more go to market oriented, whereas Microsoft's program manager is in the engineering org. It's on the same comp ladder and same promotion ladder as eng…”
Rosenthal: Andy Jassy's 2003 AWS vision document requested hiring 57 people
“In the document, there is an ask, a proposal to hire 57 new people to go pursue this initiative.”
Rosenthal: Building with web primitives makes the internet the operating system
“Andy would later say, this is a great quote, and this is absolutely true. He says, if you believe developers will build applications from scratch using web services as primitive building blocks, Then the operating system becomes the internet. And that's so tru…”
Rosenthal: Amazon's engineering team in Cape Town built the original EC2
“That becomes EC two. It's that team in South Africa that builds EC two.”
Gilbert: AWS succeeded by launching unopinionated infrastructure primitives
“They truly were unopinionated about this as a platform. They said, we're going to launch with primitives. It's the most basic story. That's the most basic compute. It's the most basic way to host your databases. It's the most basic CDN. And we can't wait to se…”
Gilbert: Targeting startups as a beachhead was key to AWS's success
“This is probably one of the biggest keys to success or sort of playbook Themes for why AWS became successful. They realized how perfect this was for startups. They realized how hard it would be for large enterprises to just wholesale move over. They realized t…”
Gilbert: Microsoft and Google initially missed the cloud wave via PaaS
“And the unfortunate thing for Microsoft and for Google, who really started at this platform as a service layer, was you basically didn't get the startups because you didn't have a mature platform yet that people were excited to build on and understood how to b…”
Rosenthal: Amazon originally targeted 20% to 40% margins for AWS
“Amazon targeted gross margins and operating margins for AWS in the 20 to 40% range.”
Gilbert: Amazon, Microsoft, and Google hold 40%, 22%, and 10% cloud share
“Amazon's got 35, maybe 40%. Microsoft's got 20 to 22%. Google's somewhere around 10%.”
Rosenthal: Amazon did not finish migrating off Oracle until 2019
“Amazon.com did not finish their migration off of Oracle databases. And onto AWS products. Until 2019.”
Rosenthal: As of 2022, AWS was on an $80B annualized revenue run rate
“AWS today is on an. Eighty billion dollar revenue run rate eight zero.”
Rosenthal: AWS has over $100 billion in contracted revenue backlog
“That backlog of committed contractual signed revenue is over one hundred billion dollars.”
Gilbert: AWS generates over 90% gross margins on bandwidth charges
“AWS makes 90 plus percent gross margins on their bandwidth charges.”
Gilbert: Data gravity gives AWS an ML moat over Google's superior models
“What's becoming clearer to me is the machine learning capabilities that Amazon has need to be good, but they actually don't need to be as good as Google's. Because here's sort of the strategy with machine learning. You're gonna use whatever ML is available wit…”
Gilbert: Missing the data warehouse market to Snowflake is AWS's biggest failure
“It's probably AWS's biggest failure.”
Gilbert: Two-pizza teams caused AWS to launch a fragmented alphabet soup
“The other thing where they're sort of a victim of their own success is the Amazon two pizza team thing led them to launch all these different services. Rather than having a cohesive product strategy, AWS has kind of been alphabet soup.”
Gilbert: AWS is probably the greatest scale economies business ever
“Scale economies, This is probably the single greatest scale economies business of all time.”
Gilbert: Massive AWS data center investments drove Amazon free cash flow negative
“They just invested so much in building out new data centers for AWS, they actually took Amazon cash flow negative. Their free cash flow is massively negative over the last 12 months because of these continued, unbelievably large investments.”
Gilbert: AWS cut prices 51 times by 2015 without facing competitive pressure
“By 2012, and keep in mind they had very little competition up to this point, they had already done 23 price reductions across the board for all of their services. By 2013, they had done 40. By 2015, they had done 51.”
Gilbert: Amazon's 100% ownership of AWS represents history's greatest venture return
“He just gets capital allocation, and that's why I think Amazon is effectively the highest performing venture returns in history. AWS is a venture bet in their portfolio that they own 100% of.”
Gilbert: Enterprise cloud providers must maintain failed services to protect customer trust
“When you launch a service for enterprise customers, it is really hard to kill it. You burn customer trust, and actually, if you think about what's the bigger risk, burning the trust and losing that customer in all their future revenue, or having to maintain ki…”