MongoDB is a key player in data storage and analytics with a document-oriented setup. Last year, we dug into what this actually means. I think that’s an important read for those wanting to understand MongoDB. The piece can be found in section 3 of this article (mostly section 3 part e).

The firm’s most exciting product is called MongoDB Atlas. This is a cloud-native database service that implements a group of servers (or a cluster) to actually store data for app creation within its platform. The nature of MongoDB’s product allows clusters to be easily added to or subtracted for easier flexing up & down as needs fluctuate. It also offers MongoDB Realm as a mobile environment for app creation, MongoDB Stitch to build apps without servers or any needed infrastructure maintenance and MongoDB Search for data querying. Finally, it offers MongoDB Data Lake specifically for unstructured data, which directly competes with players like Snowflake.

Last quarter, the company shocked the street with sharp annual guidance misses. It pocketed $80 million in unused Atlas revenue and multi-year licensing business last year. That $80 million will not recur. That chunk was essentially pure margin, which makes the profit hit larger than demand. Keep this in mind as we go through its financials and why they look underwhelming.

MongoDB is a sandbagging king. It loves to aggressively lower the execution bar only to sharply beat that lowered bar. Most (including me) assumed that it did the same thing here. In my mind, this is why the annual guidance reduction (discussed later) was so harshly punished. The initial disappointment last quarter was related to strange, non-recurring items. This incremental guide down was related to core fundamental trends and so is more notable. Let’s dig in.

Please note that Atlas proliferation drives gross profit margin (GPM) contraction. Its legacy on-premise business is higher margin.

We’ll cover why that happened later in the article.

MDB trades for north of 100x this year’s earnings. Earnings are expected to shrink 32% Y/Y. When zooming out, it earned $0.81 per share two years ago. That represents a 2 year earnings CAGR from then to now of 67%. It way over-earned last year. Earnings growth should be around 35%-40% Y/Y next year.

We’ll get into the temporary headwinds holding this business back, but the durable structural tailwinds for its niche remain firmly in place. Its document-oriented approach to data ingestion, organization and utilization is perfect in today’s world. It allows for more flexible onboarding of different, non-static formats to power easier cloud app creation. And now? In today’s multi-modal GenAI world, it thinks this flexible style of data ingestion will be perfect for powering GenAI-infused applications. Gen AI apps and models require the ability to query from a diverse set of data types. Models need to be trained and apps need to pull from these trained models. For this to happen, stakeholders need a database that supports diverse data ingestion and source code languages. That is what MongoDB provides.

Again, MongoDB’s annual guidance offered last quarter was unanimously seen as overly conservative and intentionally sand-bagged by sell-siders. That is why its negative guidance revisions missed estimates by a materially larger margin than its own forecast. The actual quarter wasn’t the issue… the guide was the issue. It lowered a bar that many considered to be the “kitchen sink.” So what happened? 

First, Atlas consumption trends were worse than expected. Recent cohorts expanded usage more slowly than expected. While this business model does benefit from subscriptions, consumption/usage swings in its products also can have material impacts on revenue generation. This was partially blamed on macro, while most other software names have said the same thing. The other factor was misplaced focus:

“While we acquired a record volume of workloads last year… we are now seeing those cohorts grow more slowly than expected. In the process of winning new workloads, we unintentionally lost some focus on workload growth potential. We've made adjustments in our processes and incentives to strike a better balance.” – CFO Michael Gordon

New workload acquisition trends were also worse than expected. Not ideal. Macro was blamed again, but MDB took some personal responsibility too. Like so many other companies (including Zscaler, which we’ll cover later in this article), it overhauled its go-to-market last year to focus on workloads with longer growth runways. It was able to address some of this weakness as the quarter progressed, but not all of it.

“New business performance for Q1 wasn’t up to our standards.”

CEO Dev Ittycheria

Slower new workload growth and workload expansion were the sources of the annual guidance reduction. That’s the bad news. 

The good news is that retention rates are strong, competitive win rates are stable and it’s seeing progress in fixing sales processes. What do I think is happening here? I don’t think MongoDB is a bad company… far from it. I think its current revenue cycle timing is poor. Companies are racing to add GPU capacity, with hyperscalers enjoying faster growth as a result of this too. GenAI app creation and production is still in its infancy. The same is true for models to a lesser extent. The software boom from this GenAI boom has really not yet begun. MongoDB doesn’t benefit from massive GPU orders. MongoDB does benefit from added demand for data services and app creation. That part of the GenAI wave is coming… just not yet.

“We see a lot of experimentation, but we're not seeing AI apps in production at scale. And so I think that's the delta between the results the hyperscalers produce versus what we are seeing in our business.”

CEO Dev Ittycheria

Novo Nordisk is now using MDB Vector Search to accelerate drug discovery. This is shrinking drug report times from 12 weeks to 10 minutes. Vector search queries needed data to power broad app creation. It offers semantic search too, which allows clients to seamlessly scrape insight from data. It offers theme-and-idea-based querying rather than just word-based. It also provides retrieval-augmented generation (RAG). This pushes semantic search results into associated large language models (LLMs) to uplift querying precision.

MongoDB is getting more proactive in nurturing channel partnerships. It finds “disproportionate returns” from winning the kinds of enterprise contracts that these relationships routinely provide. This quarter, it announced the MDB AI Applications Program (MAAP). This provides architectural templates and instructions (with a boatload of 3rd party integrations) to support GenAI app creation. Again, this is how the GPU goldrush will potentially benefit this specific business in the future. It wants that future to come sooner. Enter MAAP. Accenture is its first major system integrator partner in the program. That firm will create a “center of excellence” specifically for MongoDB app projects. 

Broadly speaking, displacing legacy relational databases is MDB’s biggest opportunity. It’s great at automating the preparation and data movement of these antiquated products. It’s not great at rewriting application code as that data enters its ecosystem. It plans to lean into GenAI products to improve here. It thinks this will significantly diminish migration friction for new customers. Anything it can do to motivate experimentation with its document-style system should be great for this business, considering how much better this architecture works vs. older, archaic relational databases.

“We can use AI to meaningfully reduce time, cost and risk of modernizing legacy apps.”

CEO Dev Ittycheria

This was not good. I think the weakness is somewhat understandable, given its business model and the current backdrop. But? This was still not good. The sources of the disappointment last quarter were a lot easier to digest than the issues cited this quarter. Still, MDB’s document-oriented niche is highly relevant and should find healthier overall growth as macro brightens. It’s easy to see how performance here could sharply improve, and that’s basically a requirement at this point based on the multiple you’re playing. I still view this as a high quality company. It’s just struggling a bit right now.

A key theme of this chat was Airbnb’s conviction in the runway being very long… even for its existing travel marketplace product in its most mature markets. While brand awareness is sky-high, that’s merely step one. Step two is turning that grand awareness into purchase consideration, high engagement frequency and eventually more revenue per aware customer.

How do they drive this awareness? In a few places. First is listing quality. Some customers are uncomfortable with the idea of booking a vacation property based on a few pictures from a random host. Listing fraud, dishonesty and availability inaccuracies are all still somewhat prevalent issues. Staying in a Marriott hotel, where quality, amenities and experience are wildly consistent, is simply more appealing to some customers.

Airbnb cannot match that uniformity and does not want to match that uniformity. Listing uniqueness is a big part of its pull. Luckily, it doesn’t really need to. There’s no discernible difference between professional and individual host list ratings. That surprised me a bit. So? It can more intentionally and safely sift through poor quality and create groupings of its highest quality, top-rated listings. It’s doing just that.

Aside from making the core product better, expanding more deeply into its underpenetrated international markets is a focus. Germany was the example used, as Airbnb hasn’t quite nailed down the marketing playbook for that unique country. It thinks it’s figuring things out and, along with other nations like Brazil, is starting to see real fruits from this labor.

The last area for growth has been the most disappointing one to date. Airbnb Experiences attempts to match people with world-class talents or knowledge with guests to bestow this wisdom/skill upon them. This really hasn’t taken off much since the 2019 debut. In fairness, it did greatly pull back on efforts here throughout the pandemic, and now it’s ready to lean back in. It’s newest “Icons” product matches people with celebrities and majestic listings (many Disney-themed) to wow them. This is really for the most affluent Airbnb client, but expansion here should be rapid now that the team thinks it has nailed down the foundation. It has a $73 billion annual volume run rate with one product. There are so many more things Airbnb can do to match guests with compelling housing… but also so much more beyond housing. That opportunity is entirely ahead of the firm. Outside of Icons, Airbnb wants to do a better job with forecasting the timing of travel accommodation purchases. It wants to be a bigger piece of facilitating transportation, excursions and more.

The team was asked about Vrbo competition. In the nicest way possible, they hinted at Vrbo being a pandemic darling. Its subscale nature means a lower-quality marketplace, while its larger professional host skew means fewer unique options. There’s little listing overlap and the strengths of Airbnb’s model have turned out to be more structural in nature than for Vrbo and other smaller competitors in the USA.

2024 is looking like the first “normal” year for Airbnb since the pandemic thus far. Leisure travel demand seems to be settling in after violent swings over the last 4 years. Still, there are some pandemic accelerants that don’t seem to be going away. Long-term and non-urban stays continue to index well above 2019 levels, and that seems to be more perpetual than most assumed.

Airbnb continues to see a somewhat price-sensitive consumer hunting for deals. This is why its commitment to listing price stability over the last several quarters has resonated so well. Its price increases are tracking well below hotel peers. This is also why its price transparency change and push for hosts to eliminate cleaning fees while driving affordability have been so popular. Most individual hosts simply don’t know what the optimal listing price is. Comparable pricing tools have helped a ton here.

As an interesting aside on pricing, Airbnb is working on helping its hosts to better understand market demand. It hinted at toying with dynamic features to automatically toggle prices based on demand levels. It’s a very similar concept to Uber and Lyft.

Gross margin realized 120 bps of leverage from product and air freight disinflation as well as lower inventory provisions. Markdown rates did rise 50 bps Y/Y, but the company reiterated its expectations of flat markdown rates for the year.

“In terms of U.S. product margin, I don't see that changing over the long term, we run a highly full-priced business. We have no plans to change our strategy there. So I would view some of the current challenges with assortment and slightly higher markdowns as temporary.”

CFO Meghan Frank

Selling, general and administrative (SG&A) was 38.1% of revenue vs. 37.4% Y/Y. 70 bps of deleveraging was much better than 135 bps that it guided to. This was related to expense timing, but also revenue outperformance and cost management.

Lulu trades for about 24x this year’s GAAP earnings (doesn’t make non-GAAP adjustments here). Earnings are expected to rise by about 16%-17% this year, depending on where estimates shake out.

The key concern stemming from Lulu’s last earnings report was weakness in the USA. Many assumed this was because brands like Alo and Vuori were eating its lunch, taking its share, and killing its brand momentum. As I argued then and will argue now, that is just not realistic. Athleisure is a $300 billion industry. To see other players taking some market share as a red flag is just something I vehemently disagree with. But I will say… a company not having a monopoly in a $300 billion space is a rather easy bear case to palette. So thank you, bears.

What went wrong in the USA and sparked the breathtaking stock decline? I’m glad you asked. There were two issues from Q2, which were both telegraphed on its last call. The first is macro. General consumer uncertainty is weighing on overall spending levels. And while that convenient cop out was cited, the team made it clear that this isn’t the predominant factor. The biggest issue is women’s assortment. Lulu messed up this year in terms of stocking its site and stores. The team will explicitly tell you that. They didn’t have the right leggings color variety or enough smaller sizes for women. They also didn’t have enough inventory for their popular bag releases. Revenue was left on the table. Traffic trends in its stores and on its site were quite strong. Conversion suffered because of its own blunders. Again, they will candidly tell you that. As this situation was partially resolved, conversion rates directly and materially improved; wherever it was able to plug inventory gaps, guests responded quite positively.

So? While this issue is self-inflicted, it’s not overly concerning to me. It’s actually somewhat encouraging. This team has a pristine track record of winning, and one blunder does not change that. Furthermore, this scenario shows you there’s nothing wrong with this brand’s momentum. People are going to their stores… trying to buy things… and buying things when they are available. Give them what they want. That’s a very easy fix and the fix is being implemented now. Despite this error, U.S. female market share trends were stable and things should improve from here. 

For more evidence, it took significant men’s market share in the USA, as it didn’t deal with the same micro-level assortment headwinds. All loungewear launches like “Steady State” are performing well. “Zeroed In” and “Pace Breaker” on the technical front are both thriving. Its shoe launch continues to outperform expectations (especially casual footwear). New sweat-wicking performance polos and hoodies are performing well. This segment is entirely fine. This is not an Alo issue or a Vuori issue… this is a Lulu needing to execute better issue. I’m confident that they will. There is zero change in the team’s long term growth expectations for the USA market (or any other market). It’s as confident as it was when the stock price was setting all time highs.

“Competition has always been there. I haven't seen anything dramatically shift or change.”

CEO Calvin McDonald

Every other geography besides the USA is killing it for Lululemon. The same playbook that worked so well in North America is proving to be globally relevant. Run highly productive stores… locally activate brand awareness with geographic marketing… give them all of the assortment they want. It’s a very typical process that we see from others such as Airbnb. The journey is yielding faster brand awareness gains than it saw in North America and faster adoption from men too.

Ex-USA strength includes Canada (+12% Y/Y), where its unaided brand awareness is highest and its opportunity is the most mature. That again shows you the U.S. issues are related to its inventory blunders. I love evidence.

Lulu sees international revenue rising from 27% of its business to 50% over the coming years. As untapped as the U.S. opportunity remains (30% unaided brand awareness vs. ubiquitous for Nike) that’s even more true outside of North America; international awareness is routinely below 10%. And when these international consumers are becoming aware, they are spending. China revenue rose by 45% Y/Y (52% FX neutral) with the “Rest of World” region rising 27% Y/Y (30% FX neutral). Comp sales in both geographic categories rose by more than 20% Y/Y.

Lulu’s guidance implies a stronger 2nd half of the year. This optimism is coming partially from easier comps. Furthermore, expectations to finish resolving the current inventory issues during the second quarter are helping too. The year is also significantly back-half weighed in terms of new product and category launches.

During the quarter, Lulu spooked many with a press release talking about leadership turnover, an organization shift and needing to “accelerate innovation.” As I said a few weeks ago, that bothered me. As I also said a few weeks ago, I tend to over-analyze every single word published by firms in the coverage network. It’s a personal strength AND a personal weakness that I must guard against. So I sat on my hands and did nothing. I discussed it here.

Today’s release made the changes seem like they were coming more from a point of strength than that press release did. The team knew this transition was coming; it had long prepped the succession plan for the departing chief product officer. Lulu sees the new shape as bolstering creativity, accountability and speed of innovation. Sounds just like what Disney is doing with its film business. It’s connecting brand, designer and merchant teams to more consistently and reliably ensure proper inventory levels.

As another small piece of organizational change, it purchased its Mexican franchise partner and its 15 stores for $160 million. It sees gaining control of this business as accelerating its ability to expand there.

Mr. Market is especially weird when it comes to earnings season. If CrowdStrike posted this kind of quarter, it would have been down 20%+. Sentiment and positioning heading in matter almost as much as the actual data… that is for traders. Sentiment has been overwhelmingly negative for Lulu, and so a very average quarter is being rewarded.

For investors, this is all noise. What actually matters? Multi-year data trends. Those trends remain strong for Lululemon. That was true 6 months ago… 3 months ago… and is true today. There’s more margin expansion left to be secured, there’s a miles longer runway to enjoy and there’s a team with a demonstrated ability to win. I think inventory blunders will be short-lived, comps will normalize, macro will brighten a tad and this thing will be right back to a 15%+ revenue compounder with leverage. This is a pristine brand and pristine brands don’t die overnight because their stocks fell and Twitter pundits loudly concocted new risks. All things considered, I thought this went well.

Zscaler is a large player in network security. It competes with Palo Alto’s next-gen suite, Cloudflare and many others. ZScaler’s Zero Trust Exchange (ZTE) is its latest and greatest cloud security platform. It blazes a trail between users, apps and devices across eligible networks. It also secures data at rest and in motion.

Zero Trust is exactly what it sounds like: never trusting a device or end user. The exchange vets and verifies all traffic as it moves within a company’s perimeter. It does not allow bad actors to breach the most vulnerable piece of infrastructure and freely move about it thereafter without any subsequent verification. ZScaler uses risk scores to assess needed levels of security for requests. That makes sure it’s only creating user friction when there’s actual security concern.

This Zero Trust approach routinely cuts infrastructure costs for customers. How? By shrinking the attack surface down to grant permission to one app, one user and one piece of traffic at a time. Permissions are based on client policy. This replaces an antiquated firewall and virtual private network (VPN) based philosophy in which every device & user within a perimeter gets perpetual and unconditional access. So? Zero Trust is safer, cheaper AND allows remote employees to responsibly work from anywhere. Zero Trust is rapidly replacing firewalls and VPNs for these reasons.

The aforementioned Zero Trust Exchange is the overarching platform layer tying all of this utility together. This is what drives vendor consolidation and better outcomes for all stakeholders. Now let’s dig into the results.

Retention rates were impacted by a larger part of its incremental business coming from brand new customers vs. client expansions.

Gross margin expansion was aided by extending the useful life of some assets, just like last quarter. Similarly to other firms, it’s likely attempting to milk more value out of its general compute infrastructure as the high performance compute wave unfolds. They also probably just wanted the margin boost like Meta, Cloudflare, Amazon and Google did.

For fiscal year 2025, it sees higher data center CapEx and some billings weakness from go to market changes discussed later. When asked about the guidance, the team explicitly said “we like to be prudent.”

Zscaler trades for 56x fiscal year earnings. Note that it’s now in Q4 of its current fiscal year. It trades for about 51x next year’s earnings. Earnings are expected to rise by 10% Y/Y next year, following nearly 70% Y/Y growth this year.

The billings results being much better than expected is important. It offers a telling sign of robust forward-looking demand amid an uncertain backdrop. The desire to displace firewall-based systems has never been stronger… along with the desire to do so through Zscaler’s offering. Continued zero day (new) exploits of competitors continue to bolster the appetite for displacing ineffective, costly systems with something that actually works. Phishing attempt activity is rapidly rising Y/Y, incumbents are failing to prevent lateral threat movement and Zscaler is taking full advantage. For some evidence, $1 million+ ARR customers rose 31% Y/Y as it expects demand to “stay strong.”

Despite continued budget scrutiny and macro anxiety, its approach is deeply resonating with its clients. Like CrowdStrike, Zscaler in network security offers a compelling ability to consolidate point solutions across ZIA, ZPA, ZDX, data protection and more emerging product categories. The end-to-end zero trust exchange platform routinely costs more than archaic firewalls. Still, the cost-to-value dynamic is far more compelling as Zscaler drivers better coverage, better automation, easier usage, better interoperability and superior outcomes. That’s why a fortune 2000 financial services firm picked Zscaler. Its “superior architecture and protection” meant things “just worked.” Customers may pay more today… but they save more tomorrow. Zscaler allows companies to do more with less and turn investments into near-term profit drivers. That’s how companies are bucking weak trends so far in 2024. Winning today requires that capability.

And again, all of this success is despite macro difficulties. Specifically its sales cycle has elongated from about 10.5 months to 12 months.

ZScaler also won a large European retailer contract during the quarter. The client went with Zscaler for its Zero Trust approach to Software-Defined Wide Area Networks (SD-WAN). SD-WAN is a digital manager of connectivity across networks. Legacy vendors here routinely deal with troubling lateral threat movement, which makes some clients assume Zscaler’s tool does too (simply because it’s also called SD-WAN). The zero trust, minimum permission, constantly vetting approach Zscaler infuses into SD-WAN significantly minimizes this issue. A key priority has been educating large customers on this reality, and this contract is a good sign that things are working. This retailer will use branch-level protection, which is a newer Zscaler product discussed next.

Zero Trust Segmentation localizes and separates networks. This treats individual stores/factories/buildings as secure islands to prevent open sharing across locations. That lowers the risk of lateral threat movement. To expand its presence here, it purchased Airgap Networks for its location-level network security tools. By treating local subsections of infrastructure as individual assets, threat actors cannot solely pick on the weakest piece of an enterprise to access everything else.

Emerging products outside of ZIA, ZPA and ZDX represented 25% of all business during the quarter as it rounds out its end-to-end platform. Its AI solutions like Risk360 and Business Insight netted them multiple large contract wins. These tools can more quickly uncover, rank and help resolve vulnerabilities. Its purchase of Avalor, which unifies data ingestion and ranks these vulnerabilities, should be a welcomed addition.

It added a new ZDX Copilot to automate the “detection and resolution of network performance issues.” It also introduced data security posture management (DSPM) into general availability to better tag, categorize and protect cloud-native data. Generally speaking, it continues to infuse 3rd party GenAI work and its own GenAI work into products to drive efficiency gains.

Zscaler sees no pricing pressures. It was asked about more specialized, smaller network security providers, but it isn’t seeing them in its competitive bids. Its win rates remain stable and “very high.” It continues to price to value and lean on its superior outcomes to justify what is often a higher price than other vendors. The broad array of firewall and VPN-based vulnerabilities now must be more openly disclosed (per SEC mangates). This is highlighting poor outcomes and product offerings from other vendors and is accelerating demand for Zscaler.

If only I had a nickel for every company talking about tweaking go to market…

Like many other companies this year, Zscaler has made significant changes to its go-to-market processes. It hired Mike Rich away from ServiceNow (where he was its Americas President) as its new Chief revenue officer. Rich just finished filling out his management team during the quarter. What is he changing?

Zscaler is getting more focused on building channel partnerships, hiring industry-specific sales reps and shifting focus from general opportunities to specific accounts. This led to some higher-than-expected sales attrition (some involuntary) for employees likely not finding new compensation structures compelling. There were also just some salespeople with an improper skillset following the changes. Zscaler is now ramping up the pace of sales hiring to plug this gap. All of this change led to some selling and billings volatility during the quarter, as last quarter’s guide assumed. The large demand outperformance shows you that Zscaler was able to endure the obstacle reasonably well. It sees all of the new systems and hires understandably taking some time onboard. Based on this, it told investors to expect a “few point headwind” to billings growth for its next fiscal year. We’ll get more formal guidance here next quarter.

This was a great quarter. Besides CrowdStrike, it’s hard to find anyone else in the overarching cybersecurity space performing at a higher level. Traction in workload protection and its AI services is encouraging in terms of future demand runway and this team simply continues to execute. I think forward guidance sets them up for another easy quarter of outperformance in Q4. At that time, all eyes will turn to its 2025 guidance. It was smart to call out the billings headwind this quarter for 2025. That will likely create expectations that are easier to surpass. And for a shallow, under-promise-over-deliver-obsessed Wall Street, that matters.

Uber wants investors to shift focus away from incremental EBITDA margin and towards total EBITDA generation. Why? Because it wants the degree of freedom to get aggressive with growth investments when compelling opportunities surface. It does not want to be a slave to needing to meet a margin number, when that could mean lower profit generation overall.

I like leverage, and I don’t think we should expect leverage to suddenly halt based on these words. CFO Prashanth Mahendra-Rajah reiterated Uber’s three year targets set at its last investor day. These targets imply EBITDA as a % of gross bookings rising from 3-4% to 5%+. It has a longer term goal of getting to 6%+. I personally see profit growth leading revenue growth for many more years to come.

We also got a bit more detail on these three year targets. The mid-to-high teens volume CAGR it offered includes low teens growth for its core UberX business. That’s its most mature bucket.

I could have answered this question for Mahendra-Rajah and I’m surprised he was asked it. Uber’s differentiation lies in its ability to offer a broader array of compelling services. It doesn’t just move people – it moves people, food, iPhones, clothing and sweatshirts so I have to see an ex-girlfriend face-to-face again. Thank you Uber. It is a last-mile delivery company… not an on-demand taxi company.

More services mean better retention and lifetime value. More services mean happier, wealthier, busier drivers, which means a better, faster consumer offering. More services mean 25% of its overall bookings now come from highly visible Uber One subscriptions. This is how Uber becomes more profitable than anyone else in its space and justifies more aggressive investment to keep taking more market share. It’s how the team plans to reach an investment grade credit rating next year. More services help everywhere. It is where Uber stands out from the pack and how Uber profitably wins. This doesn’t even mention the fact delivery is already a solid EBITDA, GAAP EBIT and free cash flow driver for this business.

Uber has top positions in 5 of its 10 largest delivery markets and is “generally gaining category position in all of them.”

Mahendra-Rajah thinks he messed up a little bit at the beginning of his tenure. His newness with data trends and demand patterns led to more caution than he needed. In the wake of uncertainty, I always prefer caution, but this did cost them some revenue. It’s now remedying that situation by leaning back into growth spend in all of the compelling areas it had been forgoing. I like hearing brand new CFOs take accountability. It would have been very easy to just blame macro or other people.

In the U.S., just 30% of its addressable market has tried Uber. It is pushing hard to raise this number and its trips/month level as we speak. Part of that is its focus on suburban demand. Its new partnership with Instacart and its own grocery business are helping a lot here. Its new ride-sharing products catered specifically for commuting should help too. But it’s also simply getting more proactive when it comes to attracting customers earlier. The new Uber for Teens account is enjoying “absolutely explosive” growth out of the gates. This not only means more revenue today, but more revenue tomorrow for longer runway consumers. These younger consumers are also generally more comfortable with navigating apps and digesting interface updates, which should be good for diversity of product consumption. Multi-product users are 3.5x more valuable to Uber vs. users of one product.

Supply is a massive part of Uber’s success. It means lower wait times and lower surcharge rates for customers. It also makes Uber a more valuable business-to-business partner, as it controls a gigantic fleet to be leveraged. Uber constantly runs driver surveys and moves to improve experiences with small tweaks like an Apple CarPlay integration. During his tenure, CEO Dara Khosrowshahi has been an Uber driver (love this) to get hands-on experience with how processes could improve. It’s an iterative process, and iterations netted Uber a 30 bps Y/Y increase to driver retention rates.

Consistent readers know how I feel here. Uber is more of a monopoly than any AV program will ever be. I don’t see one player ever owning anything close to the entire market. Not Google, not legacy automakers, not Amazon… and not even king Tesla. I could be wrong, but that’s what I think.

So? If I’m right, AV players will need to search for partners to drive optimal utilization rates. Who can provide that? Maybe the largest consumer transportation network on the planet with millions of merchants to plug into for delivery demand. There will be a lot of moving pieces as cars go fully autonomous. I am confident that Uber will not suddenly lose its entire value proposition from this change. It will likely have to share a large chunk of revenue with these AV partners, but it will also pocket the hefty take rate that its drivers currently demand.

The transition to AV will also not be overnight. Maximizing fleet profitability is tricky as Mahendra-Rajah explains. Do these providers build out excess capacity to meet peak demand at all times? Or do they underbuild to ensure lack of dead-weight loss, but less revenue. Pairing AV fleets with more flexible driver supply is the answer to this. It’s how fleets can avoid unsustainable margins while still taking advantage of as much demand as they realistically can. That gives Uber an even better shot of being a dominant AV player as the slower evolution will give it more time to adapt. It’s almost like streaming/cord cutting, but without a pandemic to vastly accelerate pace of transition overnight.

“We have a number of AV partnerships. We have conversations ongoing with a number of more partners.”

Mahendra-Rajah

GAAP net income is going to remain an irrelevant metric for Uber. Mark to market fluctuations in its long term equity investments greatly cloud that number. When its owned stocks are rising… GAAP net income inflates and vice versa. It’s dumb GAAP accounting rule. This is why EBITDA, GAAP EBIT and FCF are my go-to metrics here. But even for GAAP EBIT, periodic legal charges like we saw last quarter can make that normally reliable metric quite noisy too.

Have a great night.