News of the Week (June 5 - 9)

Apple vs. Meta; SoFi; The Trade Desk; Amazon; CrowdStrike; Mastercard; Progyny; PayPal; Revolve; JFrog; GitLab; Uber; Macro; Portfolio

Today’s Article is Powered by Savvy Trader:

1.  Meta Platforms (META) & Apple (AAPL) – 2023 Worldwide Developer Conference (WWDC) Keynote

a) Apple VisionPro

The clear highlight of Apple’s WWDC keynote was its long-anticipated reveal of its first extended reality (XR) headset. Brilliantly, it is marketing this product as a “spatial computer” to distance itself from the negative sentiment surrounding Meta’s “Metaverse” vision. Here, we’ll dive into the actual product, market dynamics and the ripple effects on Meta’s business.

The design features lightweight materials and a single panel of laminated glass lining the front of the device. This transparent material allows the extensive cameras and sensors within the hardware to have a clear view of a user’s surroundings. The glass is encompassed by an aluminum frame with a button on it to take 3D pictures. The frame also features a crown (just like on the watch) which can be turned to toggle experience immersion levels to suit preferences and dynamic needs. Straps keep audio ports close to ears.

From the WWDC Presentation

Somewhat surprisingly, the hardware is not wireless, unlike Meta’s Quest line-up. It must be plugged into an external battery pack with 2 hours of battery life. Separating the battery pack from the headset makes it smaller & lighter. This design works entirely fine for static work and gaming use cases but not so much for movement-based gaming or fitness. Apple will invariably remove the wire over time and this will not be a durable Quest edge.

VisionPro’s battery life is the same as the Meta Quest 2 device when enabling face tracking mode and compared to 3 hours of Quest 2 battery life without that turned on.

The Apple spatial computer functions to blend the physical and digital world together in an AR-based style as Apple’s “first product to look through instead of looking at.” When putting on the headset, users see a familiar display of their installed apps. They experience all of these apps with rapid hand and eye tracking, eliminating the need to use controllers while their eyes are used for things like app selection. Meta has these same features but insiders we spoke to said Apple is generally ahead in this department. For gaming, the Apple device still requires separate controllers.

Apps are displayed on virtual screens suspended within the physical environment. These apps are fully size-adjustable and have “environments” that serve as backgrounds, fully submerging the user into the space. This can help people stay more focused while working or emulate the feel of an actual movie theater while they watch a film. Pretty cool. Multiple apps can also be displayed and used at the same time.  Moreover, non-IOS apps like the Microsoft Office suite will be available (same with the Meta Quest device). And, as always, the VisionPro works seamlessly with all other Apple hardware to sync work, pictures and more via iCloud. That’s a key, built-in advantage of Apple simply having most of North America on its devices. This perk alone will be enough to win over some users.

In true augmented reality (AR) fashion, objects can be pulled from the physical environment and infused  into the digital workspace while MacBooks can also be merged into these spaces by looking at the computer. Users can use the device to remotely interact through the exact same app (like Google Docs but a lot more powerful) to make working from anywhere more productive. Two interesting FaceTime features include spatial audio which means if you’re talking to several people, the audio will come from the direction of whoever is speaking to you. Secondly, FaceTime comes with sensor-powered digital twins of users so that people see who they’re talking to directly, rather than through a headset. AKA… an avatar that looks quite similar to Meta’s next-generation Codec Avatars. Like “Spatial Computer” vs. “Metaverse”, not using the term Avatar was another brilliant marketing ploy. Meta isn’t ready to roll this next-gen Avatar out broadly, so Apple may be a few months ahead here.

The most cutting-edge piece of this launch is the contextual pass through lens that it features. This lens hides a user’s eyes when they’re fully entrenched in an app, yet can emulate a digital twin of their eyes (that a user can see through) when someone else enters a room. Meta is still working on perfecting this and Apple appears to be ahead here for now. Quest 3 will reportedly get a large boost in this department, so we’ll have to see how the new model compares to VisionPro.

Apple Silicon semiconductors are an instrumental piece of bringing this product to life. They freed VisionPro to pack 64 pixels into the “space of a single iPhone Pixel” and gave it the capacity to provide sharper imaging and higher quality display than 4K resolution. This is objectively sharper than what Meta offers at its lower price point. VisionPro is powered by Apple’s M2 chip for elite yet quiet performance and uses a brand-new Apple chip (called R1) to facilitate real-time sensor processing. This eliminates sound and display latency that has made motion sickness an issue for some users of Meta’s (three year-old) Quest 2 hardware. Wall Street Journal came out with a piece saying motion sickness is still an issue for Apple’s device, but this could potentially diminish that problem.

• All apps and software run on an operating system called VisionOS which Apple built as an extension of its current world-class operating systems for iPhone, Mac & more.

• Two interesting partnerships came from this launch. First, Disney’s Disney+ will be available on it immediately while Apple will work with Unity Software to integrate its apps as well.

• The device will be available early next year in the U.S. (late 2024 globally). It will start at $3,500 a pop vs. $500 for the new Quest 3 which will launch this year.

b) The XR Market & Meta Ripple Effects

Meta dominates the headset market today with Apple now set to enter. Meta’s Quest product sales have eclipsed 20 million units and it will soon debut a new Quest 3 version this year at an accessible price point with more powerful use cases. How will Meta fare vs. Apple? A few notes on this.

First, the XR market will not be winner-take-all. You can ask whoever you want… this is a widely-held opinion that we passionately agree with. The sector will likely be concentrated, but not to a monopolistic extreme. So? In a way, the billions of dollars in cash burn Meta has invested to get to this point has been validated by Apple following suit in the space. Apple’s investments serve as a stamp of approval for headset potential and will inevitably motivate a lot more developer time and attention to build apps for the budding sector. Some of those apps will be built on Meta to enrich its content ecosystem as well as Apple’s. At the very least, Apple sees this as meaningful enough to justify a hedge against XR replacing some of its hardware use cases in the long term.

Next, the XR market is still in the top of the first inning. To drive ubiquity, the hardware will likely need to look like a pair of sunglasses, not heavier goggles.  That’s why Meta partnered with RayBan. Both companies are several iterations away from this technology becoming compelling to the masses and are several years away from needle-moving results. This is likely why Apple has supposedly cut its production target on VisionPro from a million to 150,000.

Many won’t be compelled by these new models (including us). While iterations continue to advance, Apple and Meta will both copy each other’s best features for their own systems. Apple has had the luxury of Quest’s tech being public knowledge to inspire its own creation for years. Now Meta (A master of copying successful features) will have that same opportunity. Whether one headset has a unique feature vs. another today is all but irrelevant. It’s about what each is building towards. And making this product comparison more irrelevant is the fact that VisionPro will START at 7x the cost of Meta’s new Quest 3. Meta could obviously add more sensors and more cameras into its Quest 3 headset if the revenue per sale was factors higher than what it will be -- see the unsuccessful Quest Pro for evidence. The approaches are simply different.

The edge that Apple will enjoy is its decades of experience in building world-famous hardware, elite operating systems and an extensive ecosystem of content partners and developers. Meta will need to continue working hard to win the time and attention of developers and will need to strike compelling partnerships of its own to motivate hardware sales. Paramount and a rumored collaboration with Roblox would be two important near-term relationships with many more needing to come (and they will).

c) Quest 3 vs. VisionPro -- Right Now, Who Cares?

In terms of product features, the hardware is somewhat similar. The contextual lens we think gives Apple the slight edge vs. Meta today which, again, it will be emulating. It’s hard, however, to see any other tools VisionPro offers that Quest 2 didn’t already have. Hand and eye tracking is not unique although some call Apple’s eye tracking tech more powerful than Meta’s older Quest 2 device. The lack of an external battery pack needing to be plugged into the device is a Quest strength which Apple will inevitably match over time. Apple’s chip is more powerful than Quest 2’s chipset. The Quest 3 will get a large upgrade via Qualcomm’s Snapdragon line but Apple’s chipset will still be a lot more powerful. For gaming, Meta has the clear edge with 500+ titles vs. Apple’s 100+ available on the arcade along with the aforementioned lack of cord. There was also nothing in the presentation about 3D-enabled gaming from Apple which is Quest’s most popular use case today.

Again, whoever finds success in this space will find it via what the products look like in several years… not what they look like today. We can argue and bicker for hours about Meta’s ability to find success in hardware, but whoever is right likely won’t become clear for some time. XR has a long, long way to go to being as popular as having a smartphone in our pockets.

“I’ve seen this before -- me during the entire keynote.” -- Former Senior Product Designer at Meta Jon Malkemus

“There’s nothing here that we haven’t seen. It’ll come down to execution and how it all comes together.” -- Lead Designer at Meta’s Reality Labs Jon Lax

The last point we want to make is perhaps the most important. The Metaverse is currently irrelevant to Meta’s results and will be for a while. The health of its apps and its ability to continue improving its AI-powered ad targeting will be the first, second and third most important piece of how well this investment thesis works for now. And if it becomes clear that Apple will win in this area, that simply will lead to Meta pulling back on investments and watching its margins explode higher as a result elsewhere. That’s a worst-case scenario we can easily accept.

d) How Does This Shake Out?

How will this market actually shake out? We can only speculate today. After speaking with several people more familiar with the technology, our favorite opinion came from the twitter account @SupBagholder. Bright guy. His view is that Meta will continue to pursue hardware commoditization. It’s doing the same thing with the open sourcing of its AI models and did the same thing with WhatsApp.

Meta will continue to flood the market with at cost (or even negative gross margin) hardware to drive adoption and accessibility on its journey to 1 billion devices sold. Doing so makes it inherently more difficult to compete with but also drives a compelling network effect which could establish a sustainable niche. That traffic will foster developer motivation to build and create content on Quest. Paramount among other key partners serve as a great jump start for populating compelling content, but independent developers will need to join as well. Traffic is that motivation and at-cost hardware lowers the barrier to building traffic scale.

In this increasingly likely Meta approach, the company will not worry about hardware being a source of EBIT. Instead, it will chase market share initially at all costs and later monetize with ads (like everywhere else) along with apps and app sale take rate. Winning a large piece of the hardware market will free it to most effectively extract value from higher margin aforementioned revenue streams.

Apple can’t really follow this play book and doesn’t appear to be doing so with its $3,500+ initial price point. The reason is that its hardware sales last quarter generated a robust 65% gross margin. Hardware is a large profit driver for this firm and racing to the bottom in terms of price per unit could damage the iconic reputation Apple has painstakingly built.

Apple also doesn’t need to sell at-cost hardware. Here I sit writing to you from my MacBook, with my AirPods in, my Apple Watch on my wrist and waiting for a message on my iPhone to tell me my food is here. Apple owns a massive chunk of the tech hardware ecosystem and has for a long time. With VisionPro integrating into the devices many of us live on, Apple can leverage that stickiness to fetch hefty premiums that Meta simply cannot justify. You can lose friends over sending a text with a green bubble vs. through iMessage. We’re kidding… sort of.

Finally, hardware is not purely incremental for Apple like it is Meta. There is a world where XR hardware advances to a point of demotivating smartphone demand. We are not remotely close to that point, but it’s a possibility. This would make Apple more reliant on a hardware-based EBIT driver vs. Meta where this endeavor is brand new for it and is purely incremental to its financials.

So, 2 different approaches:  Apple is leveraging a controlled, interrelated ecosystem to command immediate high margin hardware sales. Meta’s long term strategy is selling at cost to gain market share and ubiquity for later monetization. Both can work.

e) Zuck’s Point of View

Mark Zuckerberg’s comments on the device were leaked by The Verge. These comments mirrored our point of view. Here they are:

f) Other WWDC Announcements from Apple:

While XR was the main piece of this event, Apple also announced a slew of other upgrades to its products. Here were the highlights:

Hardware:

New MacBook 15 Inch Air with an M2 Chip making it 12x faster than any Intel-based product. Comes with 18 hour battery life. Its new MacBook Pro equipped with an M2 chip was also called out as 6x faster than any Intel-based product. There were several shots at Intel throughout the event. It also launched a new desktop product equipped with Apple Silicon’s chip IP.

Operating System (OS) Upgrades:

• New iOS 17 perks include transcribed voicemails, FaceTime voicemails and added text search filters.

• MacOS now allows you to remote start your car from your computer.

• New WatchOS with tools like mapping where on a hike you last had service and movement tracking within things like golf swings.

More Software Upgrades:

• AirDrop now supports contact and content sharing simply by putting an iPhone close to another.

• Health App is now available on iPadOS.

• Added a new game mode to prioritize performance and de-prioritize background apps not in use.

• New Safari profiles to separate work and personal search cookies.

• AirPods have a new adaptive audio mode which allows for easier toggling between noise cancellation mode and transparency mode (making it easier to hear surroundings).

• AirPlay will be live in certain hotel chains to allow casting of screens to hotel televisions.

g) More Meta News

• At an all hands-on deck meeting this week where Zuck expressed confidence in Quest’s positioning, executives reportedly celebrated the legacy Facebook app doing a better job attracting younger users in North America. That is fantastic news.

  • Leadership discussed rapidly closing the TikTok vs. Reels time spent gap from 21% of TikTok to 50% of TikTok -- a massive increase in just a year. Again, this is far more important to the viability of this investment for the foreseeable future vs. Quest.

  • Morgan Stanley came out with another note reiterating what we already know: Reels continues to rapidly grow adoption and engagement.

• The Twitter copy-cat app is coming soon.

• WhatsApp is launching channels to allow users to follow their favorite brands and people in a somewhat more social use case for the app. This will be separate from personal conversations to ensure proper privacy.

• Instagram and WhatsApp are both working on AI-powered chat and content bots. This will populate sites with more content to season its discovery and matching algorithms and so more precisely ensure users are shown what they want. It could also motivate more Meta verified demand (extremely high margin revenue) to stand out from AI. Meta verified launched in India this week.

2. SoFi Technologies (SOFI) -- CEO Anthony Noto Interviews with Piper Sandler

a) Piper Sandler Interview

This was a brief conversation without many new details shared. Here were the highlights:

On Guidance:

Noto remains “very confident” in the firm’s full year guidance and its path to positive GAAP net income “by” the fourth quarter. Over the longer term, nothing has changed about his view of long-term revenue compounding in the 25%-35% range. Importantly, he also reiterated that the technology segment will reach 15%-20% growth by Q4 while accelerating more thereafter. He’s fully confident in SoFi adding over $2 billion in deposits this quarter while maintaining that minimum rate of growth this year.

On Lending:

Noto told us that SoFi -- with Wyndham Capital purchased -- is now ready to “step on the gas” for its home loan marketing program. Rates aren’t conducive to marketing acceleration right now, but its product is ready to go with net promoter score (NPS) soaring for when that macro headwind abates. He also hinted at small business lending being a future product launch.

Noto is excited about student loan payments restarting. He sees a whopping $200 billion in outstanding federal loan originations at 6-7% coupons that SoFi can refinance at meaningfully lower rates. $1 billion in volume is safely worth about $40 million in high margin revenue for this dominant market participant.

He sees pent-up interest in tweaking loan duration as another source of demand as this resumes. In today’s high rate environment, SoFi can still collect immediate variable profit to self-fund the student loan product.

On the Balance Sheet:

Part of this interview was spent on SoFi’s balance sheet practices and loan accounting. The comments Noto gave were identical to his response 2 weeks ago which we’ve linked here. Its 18% leverage ratio leaves significant wiggle room (10.5% regulatory minimum) to continue collecting deposits and internally (& profitably) funding more originations. Its flexibility allows it to seek out the highest return on that asset which today is enjoyed via holding.

On Financial Services:

Noto again spoke on SoFi’s banking charter freeing it to maintain checking and savings APYs better than the competition as rates start to fall. Its ability to use deposits to profitably fund originations and its generally lower cost of capital vs. non-banks are two key advantages. These edges mean it can profitably pass on a chunk of its contribution margin to customers via APY maintenance. That should further separate it from the pack when we invariably get to the rate cutting part of this cycle.

All of SoFi’s financial services verticals have reached a point where contribution profit is meaningful enough to fund acquisition cost and marketing spend. This will accelerate marketing efforts further here.

b) Student Loans

As Student Loan re-financing returns in the coming months, SoFi will greet that launch a Student Loan Verification (SLV) service. This will be part of its “SoFi at Work” product suite which provides an employee benefits program to large organizations.

This is in response to Congress passing the “Securing a Strong Retirement Act” which now allows employers to match (similar to IRA matching), student loan payments with payments to retirement plans. This can be easily layered on to an existing benefits program. SLV ensures this payment matching isn’t met with rampant fraud and that eligible borrowers are the only employees being matched. Employers set eligibility parameters and SLV does everything else.

SLV is yet another example of SoFi recognizing, reacting and leading the market with new products in response to changing economic opportunities. Whether it be in response to this legislation or a soaring federal funds rate (and its checking APY reaction), this company moves quickly.

---

Savvy Trader is the only place where readers can view my current, complete holdings. It allows me to seamlessly re-create my portfolio, alert subscribers of transactions with real-time SMS and email notifications, include context-rich comments explaining why each transaction took place AND track my performance vs. benchmarks. Simply put: It elevates my transparency in a way that’s wildly convenient for me and you. What’s not to like?

Interested in building your own portfolio? You can do so for free here. Creators can charge a fee for subscriber access or offer it for free like I do. This is objectively a value-creating product, and I’m sure you’ll agree.

There’s a reason why my up-to-date portfolio is only visible through this link.

---

3. The Trade Desk (TTD) -- Kokai Launch

Introduction:

The Trade Desk hosted its Kokai launch event this week. Kokai means “open for business” or “open waters” in Japanese and is a “new media buying platform” infused with the power of AI. It extends the AI-powered tools that Koa (launched 4 years ago) features to assist optimal campaign design in new ways. The event included upgrades to several other existing products as well as a few interesting launches.

The overarching goal of the new platform is to equip advertisers and agencies with as much granularity, jurisdiction and power as possible while simultaneously making the work flows as intuitive as possible. It also aims to make The Trade Desk a better ecosystem partner and to inspire the supply chain to work more efficiently and cohesively. This, in Founder/CEO Jeff Green’s mind, is the only way to foster a competitive edge over Walled Gardens (like Google & Facebook). Let’s dig in.

Galileo Upgrade:

Galileo is The Trade Desk’s first party data onboarding product. First party data is the only way the open internet can attract dollars vs. the legacy monopoly players. Why? Walled gardens have exponentially larger datasets to work with. First party data gives the open internet a data RELEVANCE edge to combat the data scale edge. Relevance trumps scale.

The most important data is on a firm’s actual customers and their preferences. Galileo makes it easier to onboard and put that data actionably to work. This perk is now being used by over half of all TTD clients just 5 months into launch. This first party data is called a “seed” by Green. Seeds are concentrated data sets on an individual’s various relevant touch points. This data can be extrapolated in look alike models to uncover similar users maybe not on a brand’s radar to extend its rational, profitable marketing reach.

Galileo announced 14 new data partners, 7 new customer data cleanroom partners and user experience upgrades as part of the event.

Unified ID 2.0 (UID2):

If first party data is bread, identification is butter. CTV’s always signed-in format means identity is easy there. UID2 is an open internet identifier built by TTD to help with identity in all other settings (as well as CTV). It ensures an advertiser knows who they’re targeting with first party data so they know they’re targeting the right person at the right time with the right message. That luxury was jeopardized by Google’s 3rd party cookies degradation in web. UID2 replaces and disrupts cookies while working across ALL channels.

Adoption is reaching a tipping point where more and more impressions will be UID2 tagged to ensure CTV, web, mobile and all channels can always responsibly identify a user. UID2 means I know who I’m targeting while data ensures I know what they want. Powerful combo. This tandem ensures concentrated data seeds grow into mighty redwoods… and quickly. As an example of what this looks like in retail media, The Trade Desk can now take a conversion and tell you high probability instances of countless more conversions for other brand new customers.

More consistent identity authentication and data leveraging means buyers are willing to pay more for an impression as the value from those impressions spikes. So? Supply side players make more, demand side players get more value, and yes, The Trade Desk makes more too.

• UID2 is adding Azure as a data partner to join Snowflake, Google Cloud and AWS among others. These partners guarantee data can be leveraged while control over that data is maintained by it never leaving a buyer’s own environment.

• UID2 will soon launch as an open currency to extend use cases beyond traditional advertising. Green cited companies like Shopify potentially using it for win back campaigns as an example.

OpenPath:

OpenPath is The Trade Desk’s tool to “directly convey to publishers what we’re willing to pay for an impression.” This is not its entrance into the supply side, it’s a service allowing publishers with internal supply side capabilities to more directly communicate with demand. In a way, while it doesn’t replace supply side players, it does allow publishers to more easily supplant them.

OpenPath has now added 20 aggregated sellers to the program with plans to add another 100 over the next 12 months. These 100 vendors represent about half of all marketing spend; adding them to OpenPath could “dramatically improve the open internet supply chain” via better, more objective communication. This part of a key Kokai’s goal of making all stakeholders work more closely together. The launch has gone wonderfully with OpenPath traffic doubling Q/Q in each of its first 5 quarters.

OpenPath announced a few upgrades to its bid communication tools. First, it now allows publishers to block buyers that don’t culturally fit with a publisher at the domain level. This will diminish instances of frustrated bidders that won an auction but weren’t given the slot. It also debuted competitive separation capabilities to automate the spacing out of impressions from direct competitors.

Tweaking its AI Model Approach:

Koa essentially functioned as one massive, centralized AI model for media buying. Quantifying human reaction to ads is hard to grasp and using one model to convey it just didn’t work optimally (in Green’s mind). Kokai takes a more distributed approach to AI models. Like Koa, AI for Kokai doesn’t replace the need for human decisioning. Instead, it is the “co-pilot” to a media buyer pilot.

TTD broke down its media buying algorithm into several different models for things like predictive clearing (ensures proper bid pricing), budget prioritization, campaign optimization, increased resilience (signal without cookies or Apple cross app data sharing) and forecasting. These all feature language learning models somewhat related to ChatGPT, but more purpose-built for ad buying. The models are “orders of magnitude smaller” than ChatGPT, but “orders of magnitude faster.” With real-time bidding, several-second latency doesn’t work and TTD removed it.

With Kokai, every impression is “treated like a snowflake” (each unique) with it routed through all of these models to ensure the right buyer is paying the optimal price for the best slot. It pointed its new predictive clearing model at a failed campaign and immediately doubled the return metrics for an idea of how needle moving these enhancements can concretely be. “Our AI is good” is one thing. “Our AI doubled campaign performance” is a far more meaningful conclusion.

New Features:

• Launched its partner portal to guide integration partners through onboarding. A consistent complaint from partners had been difficulty with executing this. The portal should help remove friction.

• Launched the Retail Sales Index to aggregate retail media data for a more holistic view of how marketing dollars (even in offline settings) translated into revenue. This will start with Walgreens, Albertsons and Dollar General as launch partners.

• Upgraded the TV Quality Index which “pinpoints value of professionally produced content to shift budget from UGC to premium.”

• Quality Reach Index launched to quantify how similar a marketing audience is to your actual buyers to gauge marketing efficacy more closely.

All of these features will be part of its new shopping cart functionality where all decisions within buying immediately are reflected in a relevance and budgeting score among other key performance indicators (KPIs). It generates reports on this immediately while reports used to be delayed. That time lag routinely discouraged more campaign optimization so this should boost TTD’s return metric edge further.

Finally, Kokai now designs campaign and media buying like a periodic table (its “programmatic table”). This intuitively depicts variables while using Koa as the co-pilot to steer buyers through the most effective decisions. Buyers work through panels to design the best campaigns humanly possible.

4. Amazon (AMZN) -- Ads & Margins

Amazon will reportedly begin offering an ad-based tier on Prime Video (mirroring the developments at Netflix/Disney/HBO). The launch will start with 3rd party content (like Warner Brothers and Paramount) and will extend to 1st party content thereafter. This is part of our Amazon investment thesis which was centered around ample low hanging fruit for margin expansion. These sources include 3rd party selling, fulfillment optimization, and yes, advertising. With 40% of North American e-commerce, Amazon’s unrivaled traffic is ripe for ad-based monetization. It has begun to lean in with things like sponsored listings and grocery delivery placements -- this is another promising opportunity.

It’s very early innings here, yet Amazon is already generating roughly $10 billion in quarterly advertising revenue with 22% Y/Y growth as of last quarter. The massive base of revenue and passionate loyalty of its customer base is a wonderful combination for fostering cross-selling, added layers of monetization and fatter margins. Ads will be a large piece of that.

In terms of what this means for the advertising industry, we reached out to a few people to gauge the impact specifically on The Trade Desk’s business. Amazon does have its own demand side platform (DSP) on which it will rely at the start of this journey. As the segment scales however, it’s highly likely that other demand side players will be invited into the equation to populate open bidding auctions. This open bidding is how cost per impression is maximized and how this segment will be most successful. As TTD CEO Jeff Green constantly tells us, streaming is too fragmented for walled garden approaches to work. Content must be funded one way or another and open bidding is the best way to do it. That’s why virtually all other streamers are embracing this format. Because TTD represents a large chunk of these potential bids, this opening would serve as another CTV demand tailwind over time.

More news:

• Wells Fargo joined a long list of institutions upgrading the firm on optimism surrounding an AWS growth rebound in the second half of the year.

• UBS raised its price target on Amazon traction around its budding foundational Bedrock AI model.

• Amazon Prime members in Canada will get a free DashPass loyalty membership.

5. CrowdStrike (CRWD) -- President and CTO Adam Sentonas Investor Conference

Charlotte AI:

CrowdStrike recently debuted its own generative AI product called Charlotte AI. This merges prevention and remediation data (from 7 trillion threats analyzed per week) that CrowdStrike has fortified over the past decade+ with new conversational language models to help analysts. In Sentonas’s words, it gives every security analyst a powerful virtual assistant to transform beginner-level employees into experts. Sentonas called most ChatGPT-powered products glorified chat bots. Charlotte AI conversely points you to any endpoint vulnerabilities, prioritizes vulnerabilities, recommends expedient response and actually helps with remediation on its own (if you want). It’s more of a co-pilot (like The Trade Desk and Microsoft call their own AI tools) than a chat bot.

Microsoft:

Sentonas again called Microsoft Defender a signature-based technology with a few bells and whistles added on top. Signature-based detection leans on pre-existing themes and patterns of activity to flag deviations and vulnerabilities. In a world where adversaries constantly practice new techniques (especially with generative AI-support), this doesn’t work. That’s why Defender vulnerabilities are a routine source of lead generation for new CrowdStrike business. Microsoft dominates in its ability to bundle security into its giant operating system share. While that works for some customers, it doesn’t work for many more customers with pressing endpoint security needs. This is also why CrowdStrike’s win rate vs. Microsoft remains “extremely high.”

Microsoft’s vulnerabilities, constant update requirements and multiple consoles all foster complexity for teams to deal with. These updates -- like E5 licensing update requirements -- cost more than an entire year of a Falcon Complete subscription per Founder George Kurtz last month. That complexity vs. the single agent that Falcon platform boasts doesn’t only outcompete Defender with efficacy… but also cost. This is why CrowdStrike continues to grow its lead in endpoint market share despite not having that key operating system-bundling edge. Microsoft is omnipresent within large enterprises, yet CrowdStrike continues to thrive and become more ubiquitous there in its own right.

“Our win rates against Microsoft have stayed incredibly high. And when we do an evaluation to demonstrate our technology versus theirs, we demonstrate better outcomes technically and financially… We have to call it like it is. I mean, the big reason why there's so many compromises is because of the fact that adversaries are exploiting vulnerabilities in the Microsoft operating system. And they need to do a better job of fixing that and I'd like them to focus on that.” -- President/CTO Adam Sentonas

6. Mastercard (MA) -- Investor Conference

On the Consumer:

Mastercard’s North American president reiterated what we’ve been hearing from this consumer bellwether. The consumer remains “very resilient” with May volumes “consistent with expectations.” Travel, the secular shift to digital payments (less cash) and cross-border growth are all offsetting headwinds from inflation and banking turmoil.

On Partnership Momentum:

Mastercard’s wide range of services (like data analytics, loyalty programs and cyber tools such as encrypted tokenization) were credited for its recently robust new partner momentum. These services have been added to the firm’s arrow quiver mainly through tuck-in acquisitions like Baffin Bay and its AI-powered malware protection. These growth vectors helped it land Citizens Bank’s and Gap’s (formerly with Visa) business while newer products like its pay later tool was how it won SoFi’s business. These priorities were also the source of its recently expanded business with JP Morgan Chase while the momentum overall has been a key source of its doubling of accepted locations in just 5 years.

Open Banking:

Data telemetry helps business models everywhere. Endpoint security vendors are using it within extended detection and response (XDR) to augment threat protection via infusing more data sources and Mastercard is now offering it as a service within credit. It calls this service open banking which essentially functions to plug into the rest of a consumer’s data rich touch points to enhance underwriting decisions and authorization rates. Chase is using this technology today.

FedNow:

The Federal Reserve’s planned electronic peer-to-peer payment offering should go live this year. Mastercard leadership doesn’t see this solving any problems considering things like zero customer protections and a questionable interface. Mastercard does not see this as a threat to its P2P business which would be good news for vendors like PayPal and Square.

7. Progyny (PGNY), Revolve Group (RVLV) & PayPal (PYPL) -- Status Quo

a) Progyny

Progyny’s conference functioned to repeat what we’ve been recently told. Structural macro tailwinds remain, temporary macro headwinds aren’t impacting its new business or utilization, the selling season is going really well and its win rates vs. competition remain sky high. The only time it loses a customer to VC-backed vendors like Kindbody is when potential clients like Walmart force it to structure a plan that it’s not willing to offer (usually due to the low quality and friction-packed make-up). It could’ve had that business too if it wanted it. Progyny didn’t want the business and prioritized its reputation over a revenue driver.

Simply put, business is booming.

b) Revolve

For Revolve, we heard a lot more of the same as well. They remain on track to re-balance inventory by the end of this quarter. May revenue trends were down single digits Y/Y just like in April while comps start to get much easier in June. It remains fully confident in margin expansion from here as promotional activity wanes, it leverages the largely fixed G&A cost line, its distribution footprint scales and it infuses AI and AR into the shopping experience to foster more educated buys and lower return rates.

It also remains fully confident in long term revenue compounding at 20% Y/Y as it exits this challenging part of its history, markets like India rapidly grow, beauty thrives and it more aggressively cross-markets FWRD with Revolve. The new development here is AI generated designs on billboards actually being taken and used for Revolve’s owned-brand sales. It’s not quite ready to lean into owned-brands and wants to fix the inventory issue first (less flexibility with owned-brand inventory). Its newly launched owned-brands like “Helsa” are killing it.

c) PayPal

PayPal leadership hosted yet another chat to reiterate what we’ve been hearing from them as well. Volume growth is accelerating, branded share is stable, unbranded share is growing and unbranded margins have many tailwinds ahead. These tailwinds include expansion globally, serving smaller customers and layering on more services. Nothing new.

8. JFrog (FROG) -- Founder/CEO Shlomi Ben Haim Investor Conference

Current Demand Environment:

Shlomi reiterated what we heard on its last earnings call. The panic and pullback it saw during the second half of 2022 is gone. Sales cycles remain elongated, but usage growth has normalized and its new business has re-accelerated. Migrations to the cloud are also being delayed, but because JFrog offers an on-premise version of its software, this just means a delay in revenue proportion shifting to the cloud. It does not mean lost revenue in this case with on-premise margins actually being a bit better than cloud-based.

AI:

While JFrog already incorporates AI-powered automation into all corners of its operations, leadership spoke about a different perk of this transformation for JFrog’s business. AI is making developers far more productive and thus is making source code creation far more rapid. This source code must be converted to binaries for large scale deployment and maintenance. JFrog is indifferent to the source code (authorship (robots or people). All it cares about is more binary management requirements which AI will naturally support. This is an accelerant for its current business. It has no plans to independently monetize its AI product enhancements in the near term and just wants them to support its customer usability for now.

Low Hanging Fruit:

The most under-appreciated piece of JFrog’s business is that it built a revenue base of $200 million with virtually no enterprise sales investments or support. It was all bottom-up with developers powering adoption. Now, with JFrog’s strong balance sheet and more platform-level offerings, it is accelerating investments in top-down, go-to-market by bringing on more sales staff. This has been happening for the last few quarters and serves as a powerful growth driver.  

9. Gitlab (GTLB) -- Q1 2024 Earnings Snapshot

a)  Results

• Beat revenue estimate by 7.7% & beat guide by 8%.

• Beat EBIT estimate by 43% & beat guide by 43.4%

• Beat -$0.14 EPS estimate by $0.08 & beat guide by $0.09

• 128% net revenue retention vs. 133% Q/Q and 130%+ Y/Y.

b) Guidance

• Raised revenue guide by 2.1% & beat estimate by 1.7%.

• Raised EBIT guide by 28% & beat estimate by 26.8%.

• Raised -$0.27 EPS guide by $0.11 & beat estimate by $0.10.

• Next quarter guidance was comfortably ahead across the board.

c) Balance Sheet

• $930 million in cash & equivalents.

• No debt.

• Stock comp was 25.5% of sales vs. 27.3% Q/Q and 20% Y/Y.

• Share count rose 3.4% Y/Y.

10. Uber Technologies (UBER) -- New Service

Uber is launching a peer-to-peer car rental service in North America after doing so in Australia last year through M&A. Broken record alert: This is the luxury of Uber getting its cost structure and balance sheet in far healthier shape than they’ve been. It can now reaccelerate product roadmap expansion to drive cross-selling, subscriber growth and retention. Doing so while in cash preservation and survival mode is challenging (as Lyft is learning). Doing so from a point of strength is far easier and merely separates Uber further from more mono-line product competition. This is not astronomically important news, but is yet another modest growth vector that it is pursuing.

11. Final Headlines

• Shopify closed the sale of its fulfillment business.

• France is working to eliminate a tax perk that benefitted Airbnb and other short term rental firms as it endures a housing shortage.

• Visa is buying a Brazilian payments firm called Prismo.

• Lululemon Studio extended its content partnership with Xponential Fitness.

12. Macro

Output Data:

• S&P Global Composite Purchasing Managers Index (PMI) for May was 54.3 vs. 54.5 expected and 53.4 last month.

• Services PMI for May was 54.9 vs. 55.1 expected and 53.6 last month.

• ISM Non-Manufacturing PMI for May was 50.3 vs. 52.2 expected and 51.9 last month.

Employment Data:

• 261,000 initial jobless claims vs. 235,000 expected and 233,000 last report.

• Institute for Supply Management (ISM) Non-Manufacturing Employment for May was 49.2 vs. 51.0 expected and 50.8 last month.

5-Year Breakeven Inflation continues to hover just above 2%:

High Yield Option Adjusted Corporate Credit Spreads remain somewhat elevated but have been falling just a bit as of late:

13. Portfolio

I made no transactions this week.