Revenue growth did outpace bookings growth. This is not a matter of Uber squeezing drivers, as driver earnings also rose faster than bookings growth. Instead, this is related to lower rider incentives and more advertising revenue.

“Our performance thus far should give investors confidence in our ability to deliver on our 2026 commitments.”

CFO Prashanth Mahendra-Rajah

Uber trades for 23x forward earnings. EPS is expected to compound at a 35% clip over the next two years.

The core Uber business remains healthy. Drivers and couriers rose to 7.8 million, as it accelerated Q/Q adds. Uber’s growing supply lead means lower wait times, lower surcharges and better customer service. Riders follow supply. Strong MAPC growth was also nicely complemented by continued 4% Y/Y growth in trip frequency. All of this enables continued 20% Y/Y FXN bookings growth at massive scale. Secular tailwinds, including a shift to services and rising “consumer preference for on-demand convenience,” are firmly intact; Uber is taking advantage better than anyone else. As an important aside, its merchant supply growth on the delivery side works wonders in bolstering selection and raising conversion rates for the company and its restaurant partners. Supply matters so much.

It’s taking sharpening edges in other ways too. The Uber One subscription is now up to 25 million members, representing 70% Y/Y growth. For some context of how impressive that is, consider another wildly popular consumer brand in Robinhood. Robinhood’s Gold subscription (not the same product but a similar idea) saw 65% Y/Y growth at less than 10% of the scale this quarter. Uber One is a budding monster, with traction directly enabling the company to flex its value proposition by affordably offering deeper deals. How? These members consume more products (hence the retention benefit) and spend 3X more per month than non-members. All of this creates the most efficient base of revenue in the ride-sharing and delivery spaces, which allows Uber to invest more in growth, product perfection, discounts and awareness. Uber One now accounts for 35% of total core business bookings. Higher revenue visibility and quality… with lower churn… and lower marketing intensity.

This means best-in-class margins, liquidity and now cost of capital to offer yet another unique advantage vs. Lyft and others. This quarter, Uber achieved investment grade credit ratings from S&P, Moody’s and Fitch. It then issued $4 billion in new credit to replace older borrowings and reduce its cost of debt by 125 basis points. It thinks there’s more to do here to ensure its access to liquidity is cheaper than for competition.

Finally, Uber’s cross-selling engine is unmatched, meaning a superior lifetime value to customer acquisition cost (LTV to CAC) dynamic vs. others like Lyft and DoorDash. It offers a true platform for moving people and goods with far more options to cross-sell. Similarly to Uber One boosting retention, multi-product adoption generally speaking also improves revenue quality.

The cross-selling advantage will merely grow as Uber successfully debuts more products. Its existing brand awareness and consumer traffic give it a better chance of finding traction than anyone else in the space, and that’s apparent in its results. The firm’s new bets portfolio is now up to a $20 billion volume run rate (not so new anymore). Impressively, the bucket continues to overindex in terms of new customer contribution, as 25% of first time trips during the quarter were from these offerings. New bets are also fostering incremental growth and are not overly cannibalistic to existing use cases. Great to hear.

Aside from new products, more complete geographic coverage is a key focus area. 45% of the USA still doesn’t have access to reliable Uber service. This includes my densely populated Detroit suburb, where I can’t get a ride in less than 20 minutes. Its leading scale gives it a great chance to rationally pursue this demand, while higher willingness to spend and wait from these suburban consumers helps too. To capitalize on these observed preferences, Uber is debuting its Reserve and UberX priority products.

As Dara put it in the Q&A, rapid growth in the suburbs has almost been “accidental” and he sometimes “kicks himself” for not seeing this opportunity sooner. They’re now leaning in.

The pristine health of Uber’s core business gives it the clear right to go pursue these newer opportunities and its early traction is promising.

For mobility, the UK, Germany and Argentina yielded the strongest growth. This was augmented by the addition of commercial fleet partnerships like Black Cabs in London and a new relationship in Denmark to re-open that market to the company. It has added 3 new markets in the last year through these types of arrangements. Some municipalities are more dedicated to protecting cab drivers and Uber can easily help them do that while adding supply in large chunks. Win-win.

The teen account launch is also going well and is “highly incremental” to the overall business. Trips rose 40% Q/Q (small base) here. Uber added new parental controls like guardian booking to give parents more authority over bookings. Lastly, while the shuttle launch was heckled by some investors, it’s going very well. Most of the routes are selling out with little to no marketing effort and Uber will expand the product going forward.

For delivery, this quarter marked Uber’s 6th straight period of Y/Y MAPC growth acceleration. Notably, “order frequency in each of the last 6 annual cohorts has grown every year.” Merchant growth remained strong at 16% Y/Y to ensure adequate marketplace selection. As a result of all of this, Uber thinks it set new all-time highs for market share across the “vast majority of top markets.”

Finally, freight’s growth was powered by revenue per load rather than volume recoveries. Sector headwinds remain strong. No change there. It’s in wait for macro to brighten mode.

Advertising revenue rose 80% Y/Y and contributed to more operating leverage. Like Meta with Reels and Alphabet with AI overviews, Uber already has the traffic and the impressions to sell. It’s now just about getting paid. To effectively do so, it knows it still has work to do on targeting, measurement and formats. This quarter, it added a “First Impression” format to let advertisers “temporarily take over the home feed.” On the measurement side, its T-Mobile partnership will allow it to upgrade geo-targeting and data processing to improve impression decisioning.

I think the firm’s expanded partnership with Waymo was a bit overlooked. Why? Because now Waymo is exclusively offering rides through the Uber app in new cities. I guess Uber’s ability to maximize fleet utilization, optimize management and advance routing algorithms all matter a little to them. I guess Uber’s network effect… even for a mega-cap with 7 products at 2 billion+ MAUs and 2 gigantic map offerings… is still valuable.

This is encouraging. During the quarter, it inked new partnerships with Cruise, WeRide and Wayve to bring its total partner roster to 14. Partners include Waymo, BYD, GM, Toyota, Hyundai and more. Dara teased more announcements coming soon. 

“On M&A, we remain extraordinary disciplined and I want to emphasize that all opportunities are reviewed with a rigorous value creation mindset; Uber's bar for M&A has never been higher. The best deal is not having to do a deal... So we are excited to continue on our exceptional path of organic growth while sticking to our firm commitment of capital returns.”

CFO Prashanth Mahendra-Rajah

Very strong showing. The core business is thriving and its competitive leads are expanding. A tiny guidance miss is not going to make me turn sour on this fundamental darling… not even close. Margins are moving briskly higher, cost of capital is tanking, its cash pile is ballooning and its revenue quality is improving. Everything that can be going well for Uber is going well. I see what the stock price did today, and I am not fretting. I actually used proceeds from my recently significant Uber trims to add back to my existing stake (as shared earlier today in the Max note).

Expectations could not have been much lower heading into this report. And furthermore, results for Intel don’t really matter all that much. Nobody expects it to effectively compete with Nvidia or AMD in data center GPUs or CPUs. It’s all about how well it can position itself to compete with Taiwan Semi years down the road in its foundry business.

AWS growth (especially AI growth within AWS) continues to be supply constrained, not demand constrained:

“I believe we have more demand that we could fulfill if we had even more capacity today… And so while we're growing a pretty big business at a very rapid rate… I actually believe that the rate of growth there has a chance to improve over time as we have bigger and bigger capacity.”

CEO Andy Jassy

Amazon guided to $75 billion in 2024 CapEx. Jassy also told us this would likely grow in 2025. They view this as a once-in-a-lifetime opportunity and remain adamant that the true risk is underinvesting. Strong demand signals inform this point of view. And while GenAI revenue will come with lower margins today, upfront costs are needed to provide the service and capacity needed to win here. And? Amazon is fully confident in strong margins and returns for these investments down the road. 

Amazon trades for 44x forward earnings. Earnings are expected to grow by 63% this year and compound at a 24% clip over the following two years.

Amazon’s customer-facing GenAI shopping assistant called Rufus is expanding to several more markets across Europe and Canada. This product is still pretty nascent, which is quickly apparent if you try it out. It will only get better from here, with new introductions like “AI Shopping Guides” to help with discovery.

Amazon Q, its coding companion for AWS, is about to get several “practical AI game changers” for developers. Amazon has been fixated on listening to its developer community to understand exactly what they want from it this year. Jassy is very excited about the upcoming Q update.

Project Amelia is another shopping assistant geared towards sellers. It offers detailed analytics and nudges best practices to perfect marketing materials (through GenAI tools like video generation), conversion rates and revenue growth.

Alexa will soon get a large, GenAI model-powered upgrade.

For review:

This quarter, Bedrock’s leading selection of foundational models grew. It added Anthropic’s and Meta’s new models, as well as new products from Mistral and Stability AI. Amazon continues to fixate on model choice, which seems to be the correct mindset. It’s finding that most companies want to use several models within Bedrock for individual apps and even for single workloads. Bedrock’s open, interoperable architecture makes this very easy to do.

SageMaker added a new capability called “Hyperpod” during the quarter. This optimally distributes workloads across AWS’s wide range of diverse compute capacity to cut model training cost by 40%. It also automates the tracking of this distribution and fixes any blunders on its own.

In terms of Amazon’s full-stack AI push (end-to-end vertical integration of all pieces of GenAI opportunity just like Alphabet), it added new Elastic Compute Cloud Instances with its newest Graviton4 CPUs. Instances mean virtual servers to rent. These instances offer 75% memory boosts and 30% better compute performance vs. the previous generation.

In partnership news, as expected, the Oracle Database and AWS integration launched during the quarter. This allows for easier workload migration, lower data transfer cost and broader multi-cloud interoperability. It also is now working with Databricks (DB) to build custom models on AWS, while DB will use Trainium chips as its “preferred” AI chip. Per Amazon, DB went with AWS as part of a push to improve price performance for building models and apps.

Speaking of which, Amazon continues to feel more pressure from customers to improve price performance for their AI workloads and training/inference work. Amazon sees its chipsets as offering “very compelling dynamics” here, with its new Trainium chip already selling out. It had to go back to Taiwan Semi to order more of them multiple times. 

Still, while Amazon is working hard to offer cheaper complements to Nvidia’s Hopper and Blackwell hardware, the two companies remain close partners and this will not change. Amazon is a key reseller of Nvidia GPUs and AWS was also selected for Nvidia’s “Project CEBA” to provide infrastructure for its new supercomputers. AWS’s operational performance and security were the deciding factors.

GenAI is accelerating demand for AWS. In the high performance compute world, the volume of data processing is exploding in size. This makes data silos within disparate on-premise offerings unattractive and makes lower-latency, hyper-scalable, lower-cost querying vital. That’s the only way to ensure these GenAI apps and models have the context they need to drive inference and value. It’s the only way to stretch engineering resources and talent far enough to deliver elite experiences and expanding margins. It’s the only way that needed data processing won’t mean hefty cash burn.

Amazon’s suite of world-class relational and no standard query language (NoSQL) database products mean a company can bring all of their data… across multiple clouds… structured or unstructured… right to their development environment. That’s how you minimize cost and latency while maximizing price performance per query and the utility of GenAI work. Moving data and workloads to the cloud has always routinely driven efficiency and cost advantages. GenAI is merely making that more accurate.

This reality is facilitating an explosion in AI-related revenue for AWS. Specifically, growth for the AI segment was over 100% Y/Y. That’s nice, but it’s growing from a small base. Even more notably, AI growth at this revenue scale is 3x faster than AWS was delivering at the same scale. My ears perked up when I heard that — despite built-in install base advantages enjoyed today that AWS didn’t have early on. It seems like Jeff Bezos got in the ear of new CEO Andy Jassy to speak a lot more about GenAI financial progress, as anecdotes like these were more numerous on this call than past periods.

Advertising continues to be powered by sponsored listings. Amazon’s ability to connect top-of-funnel discovery with bottom-of-funnel, targeted offers is quite powerful in driving return on ad spend (ROAS) and buyer revenue. It thinks it has a lot more to do with improving its targeting and measurement capabilities to keep attracting more dollars here. Considering its Thursday Night Football opener drove a record for NFL streaming viewership, more tech upgrades paired with a large base of traffic should be quite popular for advertisers.

Amazon is orchestrating more progress on cost to serve reductions, with plenty of room left to go. It has done extremely well in rapidly localizing its outbound fulfillment network, and now thinks it can do the same thing for inbound fulfillment (from merchants and suppliers) to continue lowering miles to fulfill, raising goods per box shipped and improving overall efficiency. It has made 100s of changes for inbound processes in recent months and is already improving inventory placement quality by 25%. It also keeps boosting its footprint of same-day delivery facilities. These not only offer the best service, but the lowest cost too, so it’s no wonder as to why it’s getting aggressive here.

Within fulfillment center automation through robotics, we’re in the first inning. It announced its 12th generation fulfillment center layout (starting in Louisiana), which comes with a bevy of new robotics tools and automated processes to cut costs further. This is already reducing expenses by another 25% during peak hours. That should mean better marketplace margin during the holidays.

Cost to serve reductions and faster delivery speeds through localization don’t just mean a more profitable Amazon, but unlocking more sections of U.S. commerce too. Reductions are making it rational to offer more low cost everyday essentials on its marketplace. This has been a theme for about a year now and is directly bolstering order frequency and customer lifetime value (LTV) for Amazon. This also is having two different impacts on financials that should be mentioned:

Great quarter from Amazon. Their positioning in GenAI has been vastly underrated, just like with Alphabet. In reality, this company has a world-class suite of tools across every layer of the opportunity and is already driving material financial gain. Between AWS accelerating, the untapped advertising opportunity, continued marketplace efficiency work, exploding cash flow and brightening macro, there’s a ton to like here. In a perfect world, the AWS beat would have been larger, but it would have been if Amazon had more capacity (same as Azure). I’m very pleased.

GAAP net income was quite noisy this quarter. Apple’s effective tax rate was 50% of pre-tax income vs. 15% Y/Y. This was related to a $10.2 billion charge paid to Europe to reverse illegal aid paid to it. It’s a one-off event and led to the -34% Y/Y EPS growth. With a stable 15% Y/Y tax rate, it would have earned a little less than $1.65 per share, beating $1.60 estimates by about $0.05. Some financial media publications had not adjusted GAAP net income and EPS expectations to reflect this unique item and still had $1.60 estimates. Those that had changed their numbers had aggregate estimates of roughly $0.95, which was roughly in line with $0.97 results. If you see some reporting a “sharp miss” they’re wrong (see the GAAP EBIT beat).

Apple guided to mid-single-digit revenue growth for next quarter, which compares to 6.7% Y/Y growth estimates. GPM guidance of 46.5% beat 45.9% estimates. For EBIT guidance, if we assume mid-single-digit growth means 5% Y/Y, this represents a small miss. If we assume it means 6% Y/Y, it would represent a small beat. I wish they would just give us a revenue number range.

“As you know commodities are important for gross margin… I can tell you that both for the September quarter and what we expect for the December quarter, most commodities (besides dynamic random-access memory (DRAM) which is good for Micron) are going to move down in price.”

CFO Luca Maestri

EPS is expected to compound at a 10% clip over the next two years.

As Apple investors know, Apple earnings calls offer very little detail and newness. They’re always on the brief side of things. They’re the masters of not showing their cards and saying exactly what they need to… nothing more.

Apple Intelligence is the company’s GenAI-powered “intelligence system” offering the power of world-class models from OpenAI and the strict privacy controls that Apple users are accustomed to. As reviewed in my coverage of the 2024 Worldwide Developer Conference (WWDC), Apple is “minimizing data sharing while maximizing the amount of data processed on device.” It isn’t impermissibly collecting and using personal data to season GenAI models. And when more processing power is required for a query than a device can provide, Apple’s private cloud (run by Apple Silicon M4 chips) will field traffic. It will do so without any data leakage or storage beyond what it needs at that moment to respond. Privacy first for Apple Intelligence.

For iPhone 15 models and newer running on iOS 18.1, the first tools were just released. These include writing tools, a revamped Siri, movie generation from text and more. More products will come out in December.

“If you just look at the first 3 days, which is all we have obviously from Monday, the 18.1 adoption is twice as fast as the 17.1 adoption was a year ago. And so there's definitely interest out there for Apple Intelligence.”

CEO Tim Cook

Apple continues to see itself as the only mega cap without a pressing need to ramp CapEx. It will lean on chips and models from partners like Google and OpenAI, its massive consumer ecosystem and its world-class developer network as its value propositions for those partners. 

“On CapEx, I mentioned before a number of times, we have a bit of a hybrid model in the way we run our data centers. In some cases, we use our own data centers. In some cases, we use third-party providers. So our CapEx numbers may not be fully comparable with others.”

CFO Luca Maestri

iPhone continued to dominate top-selling smartphone lists across the globe. Per Kantar research, it was the top seller in the UK, the USA and Urban China. Urban China is especially notable considering Ming-Chi Kuo (popular Analyst in Taiwan) reported iPhone 16 orders being cut by 10 million units. A lot of people assumed this was related to weakness in China. There has also been concern about shrinking lead times hinting at less pent-up demand for this cycle than previous cycles. To leadership, this is simply a byproduct of supply chain normalization and not dealing with all of the pandemic’s weirdness. It’s not a matter of shrinking demand, which is evidenced by the strong quarterly number for this product. Customer satisfaction for the product remains sky-high at 98% vs. 98% Q/Q.

Its newest Macs are equipped with its latest M4, M4 Pro and M4 Max processors. These Mac chips offer a great leap forward in Apple’s neural network capabilities and extend the complexity of workloads able to be processed on device. Customer satisfaction remains elite at 95% vs. 96% Q/Q. Incredibly, 50% of customers this quarter were still brand new to Mac. Nvidia is now using Macs for its workforce with 10,000 computers in use.

iPad’s back-to-school campaign was highlighted as a quarterly standout while its new iPad Mini was recently released to the public. Aside from robust core market growth, it delivered 10%+ growth in Mexico, Brazil, India and more countries. Customer satisfaction was steady at 97% Q/Q and the product continues to collect 50% of its sales from brand new customers.

For Wearables, Home & Accessories, the new Apple watch added satellite calls, hiking trail maps, sleep apnea aids and more. 50% of customers for this product are brand new to it and customer satisfaction slightly fell Q/Q from 97% to 96%. Its new hearing aid tool for the AirPods 4 has been a game-changer for some consumers early on. It also added noise cancellation to this product for the first time.

Finally, on Vision Pro. There were a lot of rumors about Apple stuttering this project due to lack of demand. Not so fast, my friends (said as Lee Corso for you college football fans). It’s launching the product in Korea and the UAE and remains committed to investing here. This is the correct decision. As I said in the Meta review, Apple must insulate themselves from the risk of smartphone mobile devices no longer being the main consumer technology form-factor. And while the Vision Pro (and Quest) are too big, heavy and uncomfortable for ubiquity, it can’t wait until Meta gets there to start taking this seriously itself. We all know Apple saw the Orian glasses that Meta debuted this year, which is far ahead in terms of technological miniaturization. Apple must keep its foot on the gas pedal here, despite that probably meaning hefty losses today (like for Meta).

Apple’s hardware install base growth continues to extend the monetization runway for its services business. Additionally, transaction accounts and paid accounts both rose by more than 10% Y/Y. Apple Card also remains the top ranked card by consumers per J.D. Power for the 4th straight year (only been out for 5 years). In Apple Pay, it’s adding seamless rewards redemption and credit access right at point of checkout.

“The recurring (AKA higher quality) part of services is growing faster than the transactional part.”

CFO Luca Maestri

This was a fine quarter. Margins were good and iPhone revenue was much better than feared. Services did lag a bit, but all in all I’d call this positive. The fortress ecosystem continues to deliver device growth and set the table for more service cross-selling. The buyback machine will not slow down. This is an iconic company trading at a multiple and a PEG ratio that candidly makes very little sense to me. I don’t think investors will get hurt investing in King Apple, but I do think there’s better risk/reward to be found within mega-cap tech and elsewhere. Just me.