
1. The Trade Desk (TTD) – Earnings Review Continued
The Trade Desk is the leading buy-side player in open internet advertising. Its platform allows advertisers to buy with precision, scale and open reporting.
In case you missed it, I published the first piece of this earnings review yesterday.
e. Call & Release Highlights
Some History & Another Advertising Landscape Transition:
Jeff Green’s prepared remarks are always chalked-full of wisdom about the advertising industry. This quarter was no different. He took us through a history lesson on how the Great Financial Crisis (GFC) and pandemic both created heightened sense of urgency to maximize return on ad spend (ROAS). That, in turn, drove an accelerated shift to programmatic, biddable advertising, which benefits The Trade Desk. Economic and global shocks have a way of pushing the pace of change as companies are forced to do more with less.
And there’s another massive shift currently playing out in the world of advertising. This time, it’s being driven by regulatory crackdowns on “draconian” walled garden tech giant policies. Whether it’s the current Texas Attorney General’s Google lawsuit or the upcoming Department of Justice trial, there’s building scrutiny about how walled gardens control their ecosystems. The added attention here, per Green, has “put the spotlight” on some of the practices from Google specifically. He sees growing awareness around the role walled gardens play in knowingly matching buyers with lower grade inventory; he sees rising pushback against walled gardens routinely placing impressions next to sensitive, low quality user generated content (UGC). This could mean irreparable brand harm as well as poor returns.
While walled gardens still command more ad dollars, consumers spend 60% of their time on the open internet — with better audience authentication, better data leveraging and better ad returns. People are waking up to this reality, and open internet budgets are getting more love.
This reality is what is turbo-charging the current transition that Green spoke on. It’s pushing advertising dollars to the “best of the open internet.”
How The Trade Desk Wins Amid this Shift – Kokai:
There are three main ideas I wanted to cover here for this quarter. There are other factors contributing to TTD’s outperformance vs. everyone else, but these items were the ones focused on this quarter. First is Kokai.
Kokai is TTD’s brand new bidding platform. Kokai emulates the ease of data onboarding, campaign creation and measurement that walled gardens offer… it just does so in a fully open, transparent fashion. Advertisers get the best of both worlds, as Kokai combines convenience with better returns and honest reporting. One interesting innovation coming from Kokai recently is its new approach to audience-based buying. Thanks to seamless first party data onboarding and Kokai, advertisers can get more from their customer profiles. They can take insights from loyal fans, and let TTD create large cohorts of copy-cat customers to go target. Pretty cool. Advertisers are no longer refined to guessing which types of consumers will watch which title. TTD can tell them exactly where their highest-value eyeballs are. Unilever used this for one of its brands in Asia to drive a 229% reach boost with an 81% improvement in conversion rates. Target also used this in Australia to raise its conversion rate by 66% and cut cost per action by 36%.
A new “AI-fueled” ad performance forecasting tool should further support more outcomes like these.
There’s one more important item to cover within Kokai for this quarter. Walled gardens have always been able to better aggregate demand compared to TTD and others. It’s one thing to offer Fortune 500 brands a few hundred impressions with strong returns. But that’s not needle-moving for them. They need massive, concentrated opportunities. Kokai has a feature called the Sellers and Publishers 500 Plus Index. This collects high quality inventory (thanks partially to TTD’s TV Quality Index) from 500 of the largest open internet players. Whether it’s streaming impressions, Spotify or countless publications, this gives the big boys walled garden-like scale on TTD’s platform.
How The Trade Desk Wins Amid this Shift – UID2:
Kokai tells advertisers who they should target while helping them do so; Unified ID 2.0 (UID2) ensures buyers know exactly where high value customers are. The two complement each other perfectly.
While UID2 is only a couple years old, “ubiquitous” was the term used to describe open internet adoption levels. And if you think about it, that makes sense. 3rd party cookies are (maybe) being deprecated. That means a lack of authentication signal for web-based browsing. Just like with the GFC and pandemic (and also Apple’s data sharing restriction), this forced the industry to rapidly seek a solution. UID2 has been that solution. It replaces cookies, but goes even further than that by bringing precise authentication to all forms of digital content consumption — not just the web.
“We are building the new identity and authentication fabric of the open internet.” – Co-Founder/CEO Jeff Green.
Unwind Media enjoyed a 47% boost to ad impression value by integrating and tagging their inventory with UID2.
Disney’s head of addressable sales credited a 3x-4x ad matching rate improvement to UID2, while directly saying it uplifts ad impression value. Funding content costs is not cheap; this is how to make it profitably work.
“I don’t know that I’ve ever seen the industry in such a state of transition. I sense some panic from others about what to do next. For us, this facilitates our recent outperformance and why I'm so confident about 2024 and the years ahead.” – Co-Founder/CEO Jeff Green
Partnerships and CTV Momentum:
Disney and TTD expanded their partnership to integrate the Disney Real Time Ad Exchange (DRAX) into OpenPath. OpenPath allows publishers to directly connect to TTD demand and to practice their own yield management.
New Roku partnership to plug data and supply into TTD’s demand network. Roku is expected to integrate with UID2.
More than 90% of Ad Age’s top 200 advertisers have run a campaign on TTD over the last year.
NBC will offer Olympics inventory on TTD’s platform for the first time.
TF1 and M6 (large broadcasters in France) adopted the European version of UID2 called EUID.
DISH Media adopted UID2 for DISH TV and Sling TV.
TTD doesn’t have access to Amazon Prime Video just yet, but it thinks this will eventually happen. Amazon has the issue of conflict of interest. It’s hard to imagine (I say as a shareholder of both companies) that their impression placement is objective, considering they own a lot of the inventory they’re placing. This is further complicated by its private label business, which also purchases impressions from Amazon Prime Video. Green sees Amazon evolving exactly like we just saw Roku did to embrace open bidding for its impressions. This would dramatically improve impression value by improving the supply/demand dynamic to help fund expensive content costs on that streaming service.
f. Take
This was an excellent quarter that again shows you why this company deserves to be the most expensive in its space. There’s nothing to pick at. Flawless is the word that comes to mind.
2. Duolingo (DUOL)
Duolingo is the leader in online language learning. It uses its best-in-class dataset to drive rapid product split testing, improving engagement and easy word-of-mouth growth. It’s quickly expanding into new subjects like math and music.
Duolingo’s earnings calls are always quite brief. This would bother me a little more if it didn’t consistently deliver one masterpiece quarter after another like it does.
a. Demand
Beat bookings guidance by 4.8%
Beat revenue estimates by 1.1% & beat guidance by 1.3%.
Subscription revenue rose by 53% Y/Y; ad revenue rose by 11% Y/Y; Duolingo English Test (DET) revenue rose by 28% Y/Y; in-app purchase revenue rose by 26% Y/Y.


b. Profits and Margins
Beat EBITDA estimates by 19% & beat guidance by 20.5%.
A small portion of the EBITDA outperformance was powered by delaying some operating expenses from Q1 to Q2.
Beat $0.26 GAAP EPS guide by $0.31.


c. Balance Sheet
$829M in cash & equivalents.
No debt.
Comp dollars rose 19% Y/Y. There’s about 13% of total shareholder dilution left in founder awards and restricted stock units (RSUs). These will vest over the coming years.
d. Annual Guidance & Valuation
Raised bookings guidance by 2%.
Raised revenue guidance by 1%, which slightly beat expectations.
Raised EBITDA guidance by 5.2%, which beat expectations by 3.3%.
Sees incremental EBITDA margin of 39% for the year vs. its long term target of 35%.
Reiterated annual net dilution of 1% for 2024.
Second quarter guidance was 0.4% light on revenue & in-line on EBITDA. This was largely due to ramping FX headwinds.
It also told us that Y/Y DAU growth should remain around 60% for the foreseeable future. Find me another product with 31+ million DAUs sporting that kind of consistent growth at these margins.
Duolingo trades for 51x 2024 earnings, with earnings expected to compound at a forward 2-year clip of 43% (1.19x PEG ratio).
e. Call, Presentation & Letter
Engagement:
Engagement continues to move from strong to stronger. 32.1% of its monthly active users (MAUs) are now DAUs vs. 28.0% Y/Y. This is not a matter of more external marketing; it’s a matter of constant product split testing to ensure the app is always getting better. That’s the beauty of its leading scale. It has more data than anyone else to guide more informed split testing, a better end product and viral word-of-mouth growth. The byproduct? Duolingo is one of the only 40%+ revenue compounders with positive GAAP net income in the entire market.
Advanced English:
It sees a real opportunity to further grow usage with more advanced English learning content. Despite most language learners around the globe studying English, only half of Duolingo’s users are. That is now changing with the introduction of more advanced content, which reached 1 million DAUs just a few months after launching. This is a big growth opportunity, just like new subscription tiers. It just released another large batch of English content to nurture this momentum. It views this as getting learners to much higher proficiency and a key “step one.” Now, it’s about making sure these learners are beginning courses at the right level of difficulty. This is more challenging for English vs. its other courses, as most people come into English learning with some baseline of knowledge. It’s important to make sure Duolingo doesn’t frustrate them with content that’s too hard or easy.
3rd party research consistently shows that Duolingo is more effective for English beginners than traditional education; this is its way of widening and expanding that edge.
The new English content was created via GenAI, which brings us to another important idea. GenAI is already helping Duolingo improve its app and shrink time to project completion. That, in turn, is emboldening it to entertain more experiments, considering failures aren’t nearly as costly.
Monetization:
Duolingo Max (new subscription) and its Family Plan were cited as two reasons for the guidance raise. It sees several “avenues” to bolster monetization and bookings, and these are two of those examples. The early Max uptake has been promising. Now that it’s a year into this launch, Duolingo is ready to more broadly roll out the tier to more countries and more languages. As of now, it’s still only available on iOS for 2 languages in 6 countries (about 5%-10% of its daily users).
“We’re seeing a lot of evidence that people are willing to pay a substantially higher price for Max.” – CFO Matthew Skaruppa
Leadership also spoke a lot about the family plan. When the tier launched, it wasn’t given all that much thought. Duolingo didn’t have a formal team in place and didn’t spend much time optimizing it until very recently. The company now has that team in place and is confident that it can make success here even more powerful than it already is. It’s now adding social features and family quests to the product following the popular launch of its friend quests product. 8% of its total subscribers are now on this higher retention subscription tier.
Music & Math:
It is still not monetizing math or music courses. It takes a few years for Duolingo to build the scale and product-market fit to get comfortable with charging. That was true for language learning, and it will be true here too. Still, early signs for both courses are giving the team conviction in these being successful products over time.
For now, it’s only live for iOS in a few languages. It’s focused on adding more content and building more scale before turning on the money faucet.
Social Media:
Duolingo has been intentionally diversifying away from TikTok for a year. It doesn’t expect a ban to be impactful to its business for this reason. As evidence of that being the case, the TikTok ban in India simply led to impressions moving to other platforms where Duolingo is popular.
f. Take
This quarter wasn’t good… it was fantastic. Nothing here is anything but positive. Duolingo is becoming one of the highest-quality names in software based on its combination of growth at scale, strong margins (16% GAAP net income margin) and rapid operating leverage. The only thing it’s missing is a massive runway, but continued successful expansion from language learning to overall learning should take care of that over time. Excellent results for a wildly volatile stock.
3. Airbnb (ABNB)
Airbnb needs no introduction.
a. Demand
Beat revenue estimates by 3.9% & beat guidance by 4.4%.
Beat gross booking volume (GBV) estimates by 1.8%.
Easter timing drove the Y/Y take rate growth seen in the second chart below.


b. Profits & Margins
Beat EBITDA estimates by 30%.
More than tripled $30M GAAP EBIT estimates.
Beat $0.23 GAAP EPS estimates by $0.18.
Note that Q4 2023 net income and free cash flow (FCF) exclude unique, one-off tax charges. More net interest income, fixed cost leverage and general cost discipline drove the outperformance.


c. Balance Sheet
$11.1B in cash & equivalents.
$2B in debt.
Diluted share count shrank by 3% Y/Y. It has $6 billion left in share repurchases under the current program. It sees stock comp growth of 20% Y/Y for 2024, with growth expected to slow thereafter.
d. Guidance & Valuation
Q2 revenue guidance was 1% light, while EBITDA margin guidance was in line. This means EBITDA guidance was also a bit light. The revenue guide represents 9% Y/Y growth, with revenue expected to accelerate for the remainder of the year. Leap Day, Easter timing and FX headwinds are the sources of the slower growth guide for Q2.
Airbnb reiterated its 2024 EBITDA margin guidance of at least 35%.
e. Call & Letter
Supply:
Airbnb removed thousands of low quality listings from its marketplace during the quarter. This lowered supply growth by 200 bps from 17% Y/Y to 15% Y/Y. Supply growth was over 10% Y/Y in all regions. Faster supply growth in Latin American and Asia Pacific continued to directly lead to faster nights booked growth too. Supply growth continued to be evenly balanced between professional hosts and individuals.
To help quality supply stand out, Airbnb recently launched Guest Favorites. This is a curated collection of listings, with inclusion leading to stronger occupancy rates for hosts. Another way to push guests toward higher-quality listings is the newer Guest Favorites badge. Listings here come with an average rating of 4.93 stars and this tweak makes them more visible on the platform. More generally speaking, Airbnb continues to fine-tune search and discovery algorithms to raise conversion and extract more revenue from its giant base of traffic.
Expand Beyond the Core:
Airbnb’s expansion beyond the core will come from two places – expansion to new markets and new products. Its expansion markets are performing well, with nights booked growth rates doubling its core market growth.
It recently launched Icons as a “new category of extraordinary experiences” hosted by celebrities like Kevin Hart. Some experiences include a stay at the Disney Up House or in Marvel’s X-Mansion (a lot of Disney in here), with many more options set to be added throughout the year. This launch is expected to help Airbnb penetrate more untapped markets and demographics, while marking a major step in its renewed expansion beyond travel.
Demand:
Airbnb’s mobile app is killing it. Downloads rose 60% Y/Y while nights booked through it rose 21% Y/Y (to 54% of total vs. 49% Y/Y). This was driven by brand awareness and word-of-mouth, not paid marketing.
The company has enjoyed some nice demand tailwinds from a few unique places. First, the total solar eclipse drove a 2x Y/Y spike in nights booked along its path. The Summer Olympics have driven a 5x boost to nights booked Y/Y, while the Euro Cup in Germany this summer has delivered a 2x boost.
By geography, North America (UCAN) and Europe enjoyed strong non-urban and large group growth, with urban growth called “stable.” Average Daily Rate (ADR) rose 3% Y/Y in the U.S. due to FX and mix shift and 7% Y/Y in Europe due to the same factors, as well as listing price inflation. LatAm and Asia Pacific nights booked growth was 19% and 21% Y/Y respectively. Cross-border nights booked represented 46% of total vs. 45% Y/Y. 90+ day stay bookings rose 25% Y/Y based on the affordability changes Airbnb made last year (discounting/removing fees beyond 90 days).
Group Bookings:
Airbnb released a slew of new and updated tools to make group bookings easier on the platform. Considering 80% of its volume is from larger groups, this focus area is well placed. It debuted shared watchlists and trip invitations while revamping its group message tool for both hosts and guests to more easily communicate.
f. Take
Fine quarter with more profitable compounding and the beginnings of a renewed expansion beyond travel. Strange comp items were the source of the slightly weak guidance, and it was good to hear that growth will accelerate in Q3 and Q4. Yet another somewhat ugly share price reaction that I don’t think long-term investors should be fretting about.
