The majority of this week’s content has already been sent. In case you missed it:

Housekeeping:

Next week, earnings reviews will be sent on – Palantir, Hims, Datadog, AMD, UBER, Shopify, Trade Desk, Mercado Libre, DraftKings, Cloudflare, Lemonade, Coupang. Snapshots on Axon, Celsius, AppLovin and Coinbase will also come on Saturday.

The remainder of my Spotify & Airbnb reviews will come in the coming weeks. Not next week, considering there are more than 10 reviews I need to write, but they’re coming. Part one (detailed financials) was already sent for each.

I do not plan on writing a review for Block. I think Square and Cash App are good assets. I think leadership needs to be completely overhauled. They’ve had more than enough time and things are getting worse. I don’t think giving this company a lot of attention is worth it until a new team steps in.

Demand:

Profits & Margins:

Balance Sheet:

Guidance & Valuation:

Block trades for 17x forward EPS. EPS is expected to fall by 55% Y/Y this year and grow by 58% Y/Y next year.

Demand:

“We're operating in an uncertain environment. Consumer and business sentiment has weakened primarily due to concerns surrounding the impact from tariffs and geopolitical tensions. On the other hand, so far this year, the fundamentals that support consumer spending have been solid and our drivers are generally stable.”

CEO Michael Miebach

“The headline is that our business remains strong and consumer spending remains healthy. On the macroeconomic front, the fundamentals that support consumer and business spending have been solid to date. Specifically, unemployment rates remain low and for the most part, wage growth continues to outpace the rate of inflation. At the same time, increased economic and geopolitical uncertainty has weakened sentiment and creates risks. But remember, our business is diversified.”

CFO Sachin Mehra

Profits & Margins:

Beat $3.73 EPS estimates by $0.17 and met its low double-digit growth guidance range. It beat low-teensFXN EPS growth guidance with 15% Y/Y growth.

Balance Sheet:

Guidance & Valuation:

Mastercard trades for 34x forward EPS. EPS is expected to grow by 10% this year (I think that will rise a bit more) and by 17% next year.

Demand:

It’s important to note that Leap Day made for tougher Y/Y comps. Growth has sped up sequentially so far this quarter, partially thanks to Easter and Ramadan timing.

“Our key business drivers were strong. Even with the lapping of leap day from last year and consumer spending remained resilient in an uncertain and dynamic environment.” – CEO Ryan Mclnerney

“Focusing on the U.S., in Q2 and through April 21, we have not seen any signs of overall consumer spending weakening. While spending growth differs among consumer spend bands, with the most affluent growing the fastest, all spend bands remain resilient and consistent with past quarters… Outside the U.S., we see similar stable trends.”

CEO Ryan McInerney

Profits & Margins:

Guidance & Valuation:

Visa trades for 29x forward EPS. EPS is expected to grow by 17% this year and by 11% next year.

Balance Sheet:

Low-to-mid single-digit revenue guidance for next quarter. If we assume that means 3% Y/Y growth, it missed estimates by 5.6%. 46% GPM guidance missed 46.7% estimates. EBIT missed estimates by 4.2%. This includes a $900 million cost from tariffs (about $0.05-$0.06 in quarterly EPS. If trade war news keeps brightening and tariffs ease, that could change. Without this $900M hit, the EBIT miss would have been around 1%. More on this later.

As of right now, Apple trades for 29x forward EPS and likely closer to 30x following this report. EPS is expected to compound at a nearly 9% clip for the next two years.

Tariffs – Financial Impact:

Apple expects a more pronounced tariff cost hit than any other mega-cap so far this earnings season. While Amazon talked about getting ahead of inventory ordering to stockpile goods and delay tariff impacts, Apple doesn’t really have quite the same degree of flexibility. Its products are more specialized and it doesn’t have 2 million merchants to mix and match inventory availability. It’s a lot harder to move a supply chain for a complex $1,500 smartphone than it is an $18 t-shirt. Still, it is getting creative where it can to try to delay tariff impacts. It ordered extra inventory (which will be reflected in next quarter’s results) to hopefully outlast trade battles. Beyond the June quarter however, the impact from these policies will ramp beyond $900M barring a positive change. How far beyond? They were not willing to say.

Furthermore, that $900M impact for the June quarter an educated guess, not a formal guide. Nobody knows how trade policy will evolve and the team explicitly and rightfully acknowledged that.

Tariffs – Supply Chain Diversity:

To counteract this chaotically changing backdrop, Apple will invest $500B in USA manufacturing over the next four years. Quite the commitment, with several states standing to economically benefit and a lot of its sourced chips this year coming from Arizona. Thank you, Taiwan Semi. Candidly, news like this makes me very happy. The company already does source some components like glass from the USA and it will increase domestic presence from here. It’s also rapidly shifting more iPhone production to India, where geopolitical risk is far lower and labor dynamics are compelling. It plans to source most of its iPhone sold in the USA from India by the end of 2026. It’s already nearing 50%, while it also expects 50%+ of Mac, iPad, Apple Watch and AirPod manufacturing to come from Vietnam. For hardware earmarked for nations outside of the USA, it will continue to do most of its manufacturing in China. As long as it isn’t importing those goods to the USA, it will avoid some tariffs. Again… they are far from immune. But they also aren’t helpless. Cook told us that almost all of its China tariff exposure will remain at the 20% level, while it’s exempt from higher rates at this time for most products. Some accessories, however, are subject to a 145% China tariff.

Tariffs – Changing Consumer Behaviors?

Apple did not see any material changes in consumer behavior or any demand pull forwards during the quarter. Its guidance expects stable macro and consumer behavior.

iPhone:

The 16e entry-level addition to the iPhone lineup has been well received; the more expensive models have been a “hit” with its users. This helped drive more than 10% Y/Y iPhone upgrade growth during the quarter. Kantar Research named it the top seller in the USA, Germany, Australia, Japan and, notably, Urban China for another quarter. While the trade war rages, it’s good to see Apple’s products continuing to sell in China. Customer satisfaction scores remain sky-high at 97% vs. 96% Q/Q.

Importantly, iPhone markets with Apple Intelligence continue to materially outperform markets without it.

Other Hardware:

Apple Intelligence tools on iPad like email summarization, “smart replies” and more are a “perfect match” for the Apple Pencil Pro. This has broad functionality in image editing. 50% of iPad customers remain brand new to the product, which I find so impressive at its scale. Customer satisfaction scores remain elevated at 95% vs. 96% Q/Q.

The new MacBook Air delivers a “massive performance boost,” as you’d expect from a brand new model vs. a predecessor. Its new Mac Studio model is its highest performing laptop to date, and unlocks the usage of 600 billion parameter LLMs and extensive on-device data processing and querying. Mac continues to enjoy a lofty 95% customer satisfaction score.

Airpods continue to wow customers with the built-in hearing feature. I think this is some of the most impactful innovation Apple has shipped in years. More of this please. 50% of Apple Watch buyers are brand new to the product, while the wearables segment overall has a 97% customer satisfaction rating. This includes Vision Pro, which led to tougher Y/Y comps (along with new watch releases) for the segment and -5% Y/Y growth. Speaking of Vision Pro, Apple continues to build out the content menu, with new Metallica content experiences and the new VisionOS 2.4, which brings initial Apple Intelligence features to that device. Dassault Systèmes is using it in their 3D design platform.

Software:

iOS 18.4 is bringing Apple Intelligence to more languages and nations across the globe. Through Apple’s custom silicon work, partnerships with frontier model builders and its own work on specialized models, it’s enjoying rising adoption of its AI tools. Its on-device data processing and secure public cloud paired with ChatHGPT integrations when more processing is required is, in their mind, the right combination.

 The company talked about emoji generations, image editing and writing tools like it has for the last few quarters, but there wasn’t much new on the software feature front. As I always say, Apple tries to tell investors as little as it can in its earnings reports. It does not like to show its cards, which makes sense considering competition routinely copies its work. I’m hopeful that we will hear much more about software and AI app iterations at its Worldwide Developer Conference (WWDC) next month. And speaking of wanting a faster pace of newness, the “more personal Siri features” that have been significantly delayed are still on the way. 

Services:

Chinese Nationalism?

It’s good to see Q/Q trends brighten so much in the market most aggressively targeted by U.S. tariffs. While Apple has been struggling in China for a while, it doesn't seem like this geopolitical drama is making it struggle more. FXN growth was 0% Y/Y while GAAP growth improved 9 points Q/Q. It’s still negative, but -2% is a lot better than -11%. Subsidies surely helped, but they helped everyone, and again, market share trends per Kantor Research in at least urban China are good.

More Notes:

All things considered, it was a fine quarter. The $900M impact is quite modest and gives Apple a chunk of time to hopefully outlast current trade disputes. While I deeply admire this company and live on their hardware, I don’t see anything exciting here anymore. The highlights of the calls are discussed longer battery life and better cameras for their hardware. I would hope those things improving annually is a given, rather than a flex.

I see Meta clearly ahead of it in consumer smart glasses and arguably Quest as well. I see a company in Apple that isn’t really doing much to drive durable growth outside of a gigantic buyback program. That buyback program, considering its size, can mask slow innovation and a lack of top-line growth for a long time. It can create plenty of EPS growth in the quarters ahead. But? I can’t help but think this company is in need of a new leader to reignite the innovation engine and make sure things like long Siri delays do not happen. I think Tim Cook is one of the greatest financial engineers of our time. But I also think Apple is in need of a Steve Jobs-like leader to usher in the next decades of growth. They need to get bolder and stop resting on their laurels. The world will not live on the iPhone forever. Apple is not going anywhere any time soon… but at 30x earnings and with a 3-year revenue CAGR that is struggling to stay positive, I see better opportunities. 

Duolingo is a leader in language learning. It’s now expanding into music and math. While this is a learning-based platform, the learning is meant to be fun, competitive, social and engaging. This is a proven formula to keep users coming back. The foundation driving this compelling reality is in rapid, obsessive iteration of every single piece of its product to ensure it’s always getting better. Constantly split-testing every single variable is in its DNA. It leads with product, rather than advertising and relies on word-of-mouth growth to power the vast majority of its success. It then supplements that with efficient marketing (from social media to the Super Bowl) to create viral moments and demand accelerants.

My Duolingo Deep Dive can be found here.

Duolingo acknowledged quarter-to-date currency favorability, but it did not include that in its guidance. If the dollar doesn’t aggressively strength over the next 60 days, that will be a source of upside.

Duolingo trades for 50x forward FCF and likely a few turns lower following this report. FCF is expected to compound at a 39% clip over the next two years. Those growth estimates will be revised higher.

Affordable Marketing Machine:

Duolingo is the king of creating viral moments on social media that garner billions of impressions with virtually no variable cost. This quarter was more of the same. Its “Dead Duo” promotion teased fans with the funeral of their beloved own and racked up 1.7 billion organic impressions (no content promotions or boosts). Just more of the same for Duolingo.

Split-Testing Machine:

Duolingo continues to infuse AI into its already rapid split-testing engine to expedite it further. This is accelerating insight gleaning… product improvements… engagement… and in turn… masterful financial success. It’s to a point where it can test several hundred iterations simultaneously and optimize for the most subtle of variables. This quarter, that yielded new visual updates and lesson reminders. These sound like very basic changes because they are. But? All of these changes compound to yield the results you see above.

AI & Duolingo Max:

Well I guess Max isn’t a “mirage,” Citron. This subscription tier again drove the bulk of the quarterly outperformance. It’s now 7% of total subs vs. 5% just one quarter ago.  Adoption of this tier was again powered by the highly popular AI FaceTime feature. As a reminder, this lets students practice conversational skills, with an AI avatar (Lily) that has memory, emotion and a lack of judgement when someone makes a mistake. This naturally makes people feel more comfortable with practicing their skills when compared to chatting with a native speaker. In the near future, Duolingo will add call history so students can revisit prior conversations. 

Max is the centerpiece of the company’s new aim to be AI first. The company released a new note this week that was eerily similar to what Shopify released a few weeks prior. Going forward, team members will need to prove AI can’t do the work of more human talent requests. It has long focused on this new technology, which is partially why results remain so incredible, but those efforts are being turbo-charged. AI is what enables video chats and Duolingo’s ability to build, prep and debut 148 new language courses in a single year. It took them over a decade to add their first 100. That’s not possible without GenAI and its tightly-knit partnership with OpenAI. Best-in-class models + the largest language learning dataset by a mile and a half = a great combination. While most content creation included heavy manual lift in recent years, it’s now approaching 100% automated content creation. Hello more margin upside.

At the same time, AI token costs are not cheap. Last quarter, Duolingo said it would deprioritize focus on near-term margin optimization in order to lean heavily into GenAI-inspired product growth. It wanted to release products as quickly as possible, rather than obsess over releasing them at maximum initial margin. It also told us that the work to optimize these costs was wonderfully tangible, and that it would recover the lost gross margin by the end of the year. That expectation was reiterated. The shift back to a margin focus was supposed to begin in Q3. But? Some of this work was “so straightforward” that it was “implemented ahead of schedule.” That’s a cliché for this team. They optimized model costs and performance with their Role Play tool, which provides a great foundation for expanding efficiency work to the rest of the AI suite.

The consistent bear case over the last few years is fear of real-time AI translators displacing the need to learn a language. To me… as I’ve said before… this is like someone saying “why watch a movie when I can just ChatGPT it? Why go for a walk outside when I can ask the AI what that walk would be like?” This risk ignores that people enjoy learning things and enriching their skills. It ignores that romantic matches and job interviewers prefer talking to prospects without a machine. It also ignores that Duolingo thrives because it’s entertaining, social and competition-fostering. Those translators have existed for years, yet some think they’ll suddenly now kill this business. I just couldn’t disagree more.

“We really want to stay ahead with AI. And we believe that we are far ahead of everybody else on this.”

Co-Founder/CEO Luis von Ahn

More on Max & Advanced English:

Max continues to over-index with English learners, partially thanks to more advanced English content. That’s important. 80% of language learners are learning English. 50% of Duolingo learners are learning English. There’s a large opportunity. 

The company thinks it’s in a far better spot to teach complex English than a few years ago. But? All of its growth comes from word-of-mouth. As the team said, its reputation became “good for beginners but not much else.” That’s no longer true, but it will take time for that to be organically communicated across its user base. The team plans to “speed that up” with some targeted marketing.

Chess & Non-Language Subjects:

As recently announced, Chess will join languages, literature, math and music and Duolingo’s latest subject. Duolingo is confident that it has earned the right to offer more classes because it has built a reputation of trust when it comes to combining entertainment and knowledge. That’s why math and music have rapidly scaled to 3 million DAUs in a short period of time. That was last quarter’s disclosure; this quarter they just said it’s higher than 3 million and growing more quickly than language learning DAUs. Like always, this will take time to scale and monetize to a point of moving the financial needle, but they’re confident it will get there.

For another example of how impactful AI is for this business, consider the chess launch. This course was created with two people who didn’t even know how to code. AI guided them and equipped them with the tools to build an entire course.

Apple:

The aforementioned app store ruling in the Apple section could be big news for Duolingo. If they can route app users to web-based checkout, they can sidestep the vast majority of their cost of goods sold. At the same time, as the team mentioned, Apple is wonderful in terms of vaulted payment information cutting conversion friction. So it will test pushing people to higher margin, higher friction checkout options, but it is not committing to meaningfully doing so just yet.

Macro:

Duolingo is seeing zero signs of consumer weakening. Nothing. This is the luxury of selling a highly affordable subscription that boasts excellent engagement levels.

DAU Ceiling?

Duolingo continues to enjoy some of its fastest DAU growth in its most mature markets. This is explicit evidence of the growth ceiling being entirely unknown. That’s a very good thing.

Masterful quarter… again. You know things went well when their prepared remarks include a joke and a 30 seconds on “read the shareholder letter to see how well we’re doing.” This is a special business that deserves a 50x FCF multiple. Why? Because profit growth is so consistently elite and a 1.25x growth multiple is quite modest for this quality of consistent execution. I am not trimming shares here, but I would probably entertain doing so if it keeps running. If that happens, it will remain a core holding. The potential trims will be small.

Hopefully Andrew Left and Citron are reading this review so they understand how elite this business model is and how erroneous their bear case is. I won’t count on it. You think this is Chegg 2.0? Are you kidding me? Some people just don’t like saying when they’re wrong. Which is funny… because all investors are wrong regularly. Myself included. But with that said… boy does it feel good to be right on this one.

Uber and May Mobility inked a multi-year partnership to deploy thousands of AVs over the coming years. The first launch will happen this summer in Arlington. In Europe, Uber and Momena signed a partnership to begin AV deployments in 2026. As I talk about a lot, Uber is the network effect and demand aggregator for vehicle supply. That doesn’t change because the cars are driverless, as it still controls the vast majority of consumer mindshare. The way it can be supplanted is if one competitor owns the whole market. That makes its network effect unimportant. Pieces of news like these bode well for a future of more perfect competition, which bodes well for Uber. There are many AV deployments happening this year from companies not named Waymo or Tesla. Finally,  Uber Freight, in partnership with Arora, also announced that it has been testing autonomous truck rides from Dallas to Houston.

Waymo and Toyota announced a landmark partnership to bring driverless technology to personal vehicles.

DraftKings enjoyed an encouraging bounce-back week in New York State gambling with an 11% hold rate and 19% Y/Y handle growth. And market share vs. Fanduel in that important market continues to trend in the right direction.

Deutsche Bank channel checks point to an outperforming Shopify quarter.

SC Awards named SentinelOne the best Endpoint and Cloud Security vendor for 2025. SentinelOne also debuted an upgrade to its Purple AI product (called Athena), which infuses agentic AI into endpoint protection and security information and event management.

CrowdStrike debuted 3rd-party data integrations for its managed threat hunting to augment the reach of its talented team. This is a first-of-its-kind product.

The -0.3% GDP print is very misleading. While some pushed back against it being a byproduct of import front-loading, it is. The specific import surge in question came from gold. That’s key. Gold is not counted in consumption or investment, which is where the import offsets for other goods come. That’s likely why the net export hit to GDP was about 1.5 points larger than the investment/inventory boost. A slowdown in government spending growth amid federal cuts also hurt. Final sales to domestic private purchases (which excludes this noise) was steady at 3% Y/Y growth.

Output Data:

Consumer & Employment Data:

Inflation Data: