☁️ Amazon: This Team is Cooking
☁️ Amazon: This Team is Cooking
AWS delivers but other segments come short
Welcome to the Free edition of How They Make Money.
Over 130,000 subscribers turn to us for business and investment insights.
In case you missed it:
Premium members love our latest Earnings Visuals report!
Digest the performance of hundreds of companies in seconds.
Many readers expense our newsletter out of their learning and development budget. Here’s an email you can send to your manager.
Amazon (AMZN) just unveiled its Q2 2024 results, and the initial market reaction was less than enthusiastic. The stock dipped 8% as the company missed revenue expectations and the outlook disappointed.
However, CEO Andy Jassy was quick to emphasize the work done on gen AI:
“During the past 18 months, AWS has launched more than twice as many machine learning and generative AI features into general availability than all the other major cloud providers combined. This team is cooking, but we're not close to being done adding capabilities for our customer's usage base.”
☁️ AWS hit a $105 billion run rate, and Jassy highlighted that AI already represents a “multibillion-dollar revenue run rate.“
Management is increasing investments in AI and infrastructure, expecting long-term tailwinds. If you read our coverage of Microsoft and Google’s earnings, this story is familiar to you.
The strong AWS execution was partially overshadowed by softness in other segments, from e-commerce to advertising. Let's dive into the numbers.
Today at a glance:
Amazon Q2 FY24.
AWS and the AI tech stack.
Key quotes from the earnings call.
The NBA deal and Amazon’s reach.
1. Amazon Q2 FY24
Income statement:
Revenue breakdown:
💻 Online store (37% of overall revenue): Amazon.com +5% Y/Y.
🏪 Physical store (4%): Primarily Whole Foods Market +4% Y/Y.
🧾 3rd party (24%): Commissions, fulfillment, shipping +11% Y/Y.
📢 Advertising (9%): Ad services to sellers, Twitch +20% Y/Y.
📱 Subscription (7%): Amazon Prime, Audible +10% Y/Y.
☁️ AWS (18%): Compute, storage, database, & other +19% Y/Y.
Other (1%): Various offerings, small individually -6% Y/Y.
Revenue grew +10% Y/Y to $148 billion ($0.7 billion miss).
Excluding AWS:
North America grew +9% Y/Y to $90 billion.
International grew +10% Y/Y fx neutral to $32 billion.
Gross margin was 50% (+2pp Y/Y).
Operating margin was 10% (+4pp Y/Y).
AWS had a 36% margin (+11pp Y/Y).
North America had a 6% margin (+2pp Y/Y).
International had a 1% margin (+4pp Y/Y).
EPS $1.26 ($0.23 beat).
Cash flow (trailing 12 months or TTM):
Operating cash flow TTM was $108 billion (+75% Y/Y).
Free cash flow TTM was $53 billion (vs. $8 billion a year ago), primarily driven by the operating cash flow improvement. Purchases of property and equipment only grew by 2% to $55 billion.
Balance sheet:
Cash, cash equivalent, and marketable securities: $87 billion.
Long-term debt: $58 billion.
Q3 FY24 Guidance:
Revenue growth between +8% and +11% Y/Y (a slight miss in the mid-range).
Operating margin between 7% and 9%.
So what to make of all this?
Amazon missed overall revenue expectations, but earnings were 22% above consensus (boosted by AWS).
AWS continued to rebound for a third quarter after seven consecutive quarters of decline caused by customers trying to cut costs in a tougher environment. The growth acceleration was two percentage points sequentially, better than Google Cloud (improving by 1 point to 29%). Conversely, Microsoft Azure decelerated by 1 point to 29%.
Advertising delivered +20% Y/Y growth, a slowdown from +24% Y/Y in Q1 and slightly missing expectations. It remained the fastest-growing segment, primarily driven by sponsored ads and Prime Video.
Operating margin improved by four percentage points. AWS alone was responsible for three percentage points of overall margin improvement. This was partially due to an accounting adjustment (the useful life of servers changed in Q1 FY24).
Capex increased by 58% to $16.4 billion, driven by investments in AI. Capex was $48 billion in FY23, and management expects a meaningful increase in FY24 to support the growth of AWS and gen AI efforts. Amazon is spending more than any other company.
Q3 Guidance wasn’t a home run. The mid-point of the revenue guidance was just shy of Wall Street’s expectations. The operating income outlook implies stable margins sequentially.
2. Recent business highlights
☁️ AWS market share and margins
Synergy Research Group estimates that cloud infrastructure spending grew +22% year over year to $79 billion in Q2 2024 (accelerating from +21% year over year in Q1).
Amazon maintained a 31%-34% market share in the past five years and captured an estimated 32% of the market in Q2. That compares to 23% for Microsoft Azure and 12% for Google Cloud, which have gained market share at the expense of others. The overall pie is growing for all participants.
In the June quarter, Azure and Google Cloud grew 29% on a GAAP basis, while AWS grew 19% year over year from a much higher base. AWS maintained an elevated operating margin of 36%.
🤖 An update on the three layers of the AI stack
CEO Andy Jassy always provides an update on Amazon’s position within the three layers of the AI tech stack: 1) Infrastructure, 2) Models, and 3) Apps.
So, let’s revisit them.
1) Infrastructure Layer (Trainium and Inferentia)
Jassy explained:
“We have a deep partnership with NVIDIA and the broadest selection of NVIDIA instances available, but we've heard loud and clear from customers that they relish better price performance.”
Amazon has been developing customized machine learning chips like:
🎓 Trainium for training.
🧠 Inferentia for inference.
It also offers Graviton for generalized CPU chips at a competitive price.
The second versions of these chips could bring significant improvement in price performance.
2) Models Layer (Bedrock)
Amazon offers “LLM as a Service.” In short, AWS customers can access large language models and customize them with their data. AWS handles the security and privacy for them.
LLMs require huge investments. That’s why many companies building Gen AI apps will use Amazon Bedrock to leverage existing LLMs.
Jassy explained:
“Bedrock has the largest selection of models, the best generative AI capabilities in critical areas like model evaluation, guardrails, RAG and agenting, and then makes it easy to switch between different model types and model sizes.”
Bedrock models include Anthropic’s Claude 3.5, Meta’s new Llama 3.1, and Mistral's new Large 2.
3) Apps Layer (Amazon Q)
The top layer includes consumer-facing apps like ChatGPT or Gemini.
Amazon released its own workplace-focused AI-powered assistant, Amazon Q. Pair programming is one of the most obvious use cases for gen AI, as illustrated by the success of GitHub Copilot for Microsoft.
Jassy sounded very upbeat about the launch, calling it a “game changer:”
“Q has the highest known score and acceptance rate for code suggestions, but it does a lot more than provide code suggestions. It tests code, outperforms all other publicly benchmarkable competitors on caching security vulnerabilities and leads all software development assistants on connecting multiple steps together and applying automatic action.”
Many of these tools will be commoditized, but Amazon has the advantage of operating as a market leader today, giving the company a potential adoption edge.
3. Key quotes from the earnings call
CEO Andy Jassy on AWS growth tailwinds:
“First, companies have completed the significant majority of their cost optimization efforts and are focused again on new efforts.
Second, companies are spending their energy again on modernizing their infrastructure and moving from on-premises infrastructure to the cloud. […]
And third, builders and companies of all sizes are excited about leveraging AI.”
While the cost optimization was temporary, the secular shift to the cloud combined with the surge in demand for AI solutions is a powerful combo for durable growth. So far, the incumbents in cloud infrastructure are the main beneficiaries.
On advertising:
“Sponsored ads drive the majority of our advertising revenue today and we see further opportunity there. Even with this growth, it's important to realize we're at the very beginning of what's possible in our video advertising.“
Amazon can deliver exceptional ROAs (return on ad spend) because of its closed ecosystem, purchase data, and high-intent audience. Prime Video ads could drive specific business outcomes and bring performance marketing to CTVs.
On Project Kuiper:
“We're accelerating satellite manufacturing at our facility in Kirkland, Washington. […] We continue to feel significant demand for the service from enterprise and government entities. We expect to start shipping production satellites late this year and continue to believe this could be a very large business for us.”
Jassy previously said Kuiper shares similarities with AWS, notably in revenue potential.
CFO Brian Olsavsky on capital investments:
“For the first half of the year, CapEx was $30.5 billion. Looking ahead to the rest of 2024, we expect capital investments to be higher in the second half of the year. The majority of the spend will be to support the growing need for AWS infrastructure as we continue to see strong demand in both generative AI and our non-generative AI workloads.”
It looks like money well spent as AWS growth re-accelerates.
4. The NBA deal and Amazon’s reach
Amazon's advertising revenue reached $12.8 billion in Q2, representing:
26% of Google Search advertising revenue (+1pp Y/Y).
33% of Meta's advertising revenue (-1pp Y/Y).
147% of YouTube ads (+8pp Y/Y).
Prime Video ads started in Q1 2024. All Prime members would see ads by default unless they spend an extra $2.99 monthly. In short, Amazon turned all Prime members into ad-supported subs.
Why is this move significant? Prime is expected to reach 180 million US members in 2024. As a result, Amazon created a giant ad-supported streaming platform overnight. For context, Netflix had 84 million paid memberships in North America at the end of June, and only a small fraction of those opted for the ad tier.
Sure, not all Prime members watch Prime Video. But when it comes to reach and distribution, Amazon is in the driver’s seat with potential deals with publishers. That brings us to the NBA.
🏀 The NBA deal in a nutshell:
Value: $76 billion media rights package (Disney + NBC + Prime Video).
Duration: 11 years starting with the 2025-2026 season.
Distribution: Includes the US, Mexico, Brazil, France, Italy, Spain, Germany, the United Kingdom, and Ireland.
Exclusive Amazon games: 66+ regular season games, select playoff games.
This strategic move expands the NBA's reach to Amazon's massive subscriber base. Critically, Amazon already has tens of millions of Prime members outside the US, giving the tech giant a crucial advantage relative to Warner Bros. Discovery.
According to Nielsen, Prime Video captured 3.1% of US TV Time in June (a decline of 0.1 points Y/Y). Prime Video captures just over a third of Netflix’s market share (and more than Disney+ and Paramount+ combined).
As Amazon continues to invest in live sports and expand its content catalog, Prime members may find themselves spending more time with the service they already pay for. Prime Video may have started as a loss leader, but if it can become the go-to streaming platform for ad-supported content, it could evolve into a significant revenue driver, even for a behemoth like Amazon.
That’s it for today.
Stay healthy and invest on!
Disclosure: I am long AMZN, GOOG, and META in the App Economy Portfolio. I share my ratings (BUY, SELL, or HOLD) with App Economy Portfolio members.
Author's Note (Bertrand here 👋🏼): The views and opinions expressed in this newsletter are solely my own and should not be considered financial advice or any other organization's views.
Quarterly and other short term numbers aren't the ideal way to evaluate a business like Amazon. The key metrics to focus on are customer acquisition and retention.
Consider AWS as an example. Cloud computing has a high level of customer stickiness. Large enterprises, which handle massive amounts of data, face significant challenges in migrating that data to the cloud, including potential security risks. To address this, Amazon introduced a 100-terabyte, highly secure data transfer device called "Snowball." This device was sent to a company’s office, where the data would be transferred from on-premises infrastructure to the Snowball, which was then sent back to Amazon for cloud upload. As computing needs grew, Amazon introduced "Snowmobile," a truck filled with Snowballs, to transfer even larger quantities of data to the Amazon cloud. This process could take months.
To illustrate the scale, Amazon itself spent 13 years, from 2006 to 2019, transferring its data from Oracle infrastructure to AWS.
Once this process is complete, customers have little incentive to switch providers.
It's often said that the value of a business is the discounted value of future cash flows. The focus isn't on current earnings but on optimizing future earnings. Each new customer significantly increases future cash flows, which compounds over time. When you calculate the net present value of those future cash flows, you see a company that is consistently growing in value, regardless of recent quarterly earnings, which may be depressed due to investment in acquiring more customers.
It is all about the lifetime value of the average customer multiplied by the number of customers!
You may remember when Bezos was once asked by an analyst about his quarterly numbers on an earnings call. Bezos dismissed the question, stating "this quarter was baked in three years ago, I'm currently working on a quarter 5 years from now".
If this is the way management are running the business, which is the correct approach in my humble opinion, then investors need to think about it the same way. Otherwise it is impossible to make sense of the numbers. In other words, don't look at the most recent trees to be planted, look at how the entire forest is evolving.
Food for thought.