Deep|The Growth Potential of Adtech: Applovin, Unity, and Zeta

Adtech Is the Best Industry for AI Applications

Introduction:

In our previous article, Why is Applovin so awesome, but others haven't learned from it?, we summarized the core reasons behind Applovin's market position and increasing market share in the DSP domain:

• Near-monopoly in Mediation: This allows Applovin to act as both a player and a referee in ad bidding, helping DSPs obtain bidding information from other DSPs through Mediation and subsequently improve their machine learning algorithms.

• Exclusive data from Adjust: As the second-largest player in the MMP market with nearly 30% market share, Applovin can access Adjust's data regardless of whether game publishers agree or not, while other DSP vendors need the publishers' consent to obtain MMP data.

• Feature iterations of Axon2: More exclusive data drives more frequent updates to machine learning algorithms. Applovin also shortens the model training cycle for traffic apps by integrating training data from Meta and others, enabling customers to launch CPA campaigns faster.

As the market enters this stage, companies playing the DSP role in Adtech have different stories. In this article, we will briefly discuss the stories, growth potential, and core advantages of different companies.

1.Applovin

Applovin's past story mainly revolved around matching advertisers with game publishers. Since its traffic mainly came from the gaming industry, advertisers were also primarily from the gaming sector.
Although e-commerce advertisers have always wanted to try out Applovin, they could only be displayed in lower positions because their bids couldn't surpass those of gaming advertisers.
However, starting from the beginning of this year, Applovin began frequently engaging with large e-commerce advertisers and achieved significant success.

Let's start with the world's two largest e-commerce advertisers.
Most readers may not imagine that the world's two largest e-commerce advertisers both come from China: TEMU and SHEIN.

First, let's look at SHEIN's ad placement:

• Applovin started reaching out to SHEIN in 2019 and encouraged SHEIN to start testing Applovin's ad placement in 2021, but it wasn't until the first half of 2024 that SHEIN officially launched campaigns.

• Applovin's ad budget share remained very small until mid-2024, accounting for only 0.2-0.3% of SHEIN's online ad budget. However, by the end of 2024, Applovin's ad budget share had reached 2%.

• The 2% ad budget share ranks 7th to 8th among SHEIN's ad channels, behind META, GOOG, Tiktok, SNAP, PINS, and a mysterious ad channel.

• Considering the excellent ROI of SHEIN's current ad placement on Applovin, with Applovin's CPC being only 30-40% of Facebook's or 25-35% of Google's, Applovin's ad budget share in SHEIN will continue to increase.

• According to our research, this share may reach an astonishing 5-6% by the end of next year, which means it may surpass PINS and even SNAP. By 2026, the share may approach 10%, but breaking through the 10% budget share may be extremely difficult.

Now, let's look at TEMU's ad placement:

• Applovin engaged with TEMU later than SHEIN but achieved better results.

• So far, TEMU's ad spend has reached 2-3 times that of SHEIN, and TEMU is also Applovin's largest e-commerce customer globally.

• TEMU's ad placement on Applovin is smoother because TEMU has more product categories. Applovin's accurate user profile tags allow TEMU to find suitable product categories as ad creatives.

• If SHEIN is still solving the problem of matching categories with traffic profiles, TEMU has already solved the category problem and is further addressing the traffic issue.

• Observing TEMU's ad placement, its budget on Applovin still has a 2-3x growth potential next year.

Why is Applovin, as a 3-rd party DSP, the first to be accepted by e-commerce advertisers?

• This is mainly due to Applovin's more precise user segmentation.

• For example, Unity and Ironsource can only perform simple game scenario segmentation for gaming users, while Applovin can segment users into tags more suitable for product ads, such as beauty enthusiasts, office workers, and sophisticated moms.

• This helps e-commerce advertisers understand gaming user groups. Behind gaming users are also consumers of different e-commerce platforms, but there was a lack of accurate matching mechanisms.

• Meanwhile, Applovin also described the massive update of Axon3 to e-commerce advertisers next year, which includes further detailed descriptions of shopping profiles using GenAI technology to build a more comprehensive tag system. Applovin will also provide advertisers with full-process user tracking, such as the reasons for customer churn.

The above reasons make Applovin more favored by e-commerce advertisers compared to DSPs like Unity. We also see sell-side firms like Jefferies frequently organizing research on e-commerce advertisers placing ads on Applovin. We believe the current progress is still in the early stage and has significant room for growth.

Let's do a simple calculation of the growth potential:

• In 2025, the global ad budget of SHEIN and TEMU alone may reach as high as $15 billion.

• Assuming 6% is used for Applovin ad placement, that's a $1 billion ad budget.

• Applovin usually charges a ~20% take rate for other gaming advertisers, but for super advertisers like SHEIN and TEMU, it may adopt a more aggressive take rate strategy, such as only charging ~10%, which is equivalent to $100 million in net revenue.

• Among META's e-commerce customers, SHEIN and TEMU account for about 20%. If other e-commerce advertisers have the same progress, it could bring Applovin an incremental revenue of $500 million, roughly equivalent to a 15% increase in Software Platform Revenue.

• Looking further ahead, Applovin's budget share and take rate among e-commerce advertisers may continue to grow, providing a multi-year growth engine.

2.Unity

After discussing Applovin, let's talk about an Adtech company that has been consistently underperforming. However, if this company successfully turns around, its opportunity next year may be even greater than Applovin's.

There is no doubt that Unity's organizational and product capabilities have been weaker than Applovin's in the past. The company is filled with various negative impacts, including:

• Ongoing internal struggles in the advertising business. After acquiring Ironsource, there have been rumors of discord between the Ironsource team and the Unity Ads team. The Ironsource CEO also left Unity in early 2024.

• From customer feedback, Unity's advertising algorithms have rarely been updated in the past three years, except for a failed IDFA feature update that caused losses for advertisers and took nearly half a year to fully fix.

• During this period, Unity's original CEO was also forced to resign due to issues with game engine fees, further disrupting the company internally. Both Unity Ads and Ironsource experienced team layoffs.

However, this seems to have changed after the new CEO of Unity Ads, Jim Payne, took office:

• Jim Payne is the founder of Mopub, the former Top 4 Mediation Platform, which was eventually acquired by Applovin.

• Jim Payne knows very well how to do Mediation and understands Applovin's key strategies.

• After joining Unity in mid-2024, Jim Payne quickly restructured the product team and reformed the sales team. In the past, Unity did not have an independent ad sales team in some regions and relied on the game engine team for cross-selling. Now, more independent sales teams have been formed, and the service level has been greatly improved.

On the product side, Unity has started to communicate with customers about the massive update in the middle of next year, which will significantly narrow the gap between Unity's advertising algorithms and Applovin's Axon2 algorithms:

• The new algorithm will allow real-time ROI calculation and daily ad placement summaries. Damn it, Unity didn't have this feature before, which allowed customers to increase ad placement on Applovin in a timely manner, but they had to wait until the next day on Unity.

• The new algorithm will also refine tags, striving to build user profile tags outside of game scenarios and label consumption behaviors, which may leverage some of Unity's game engine capabilities.

• The new algorithm will also have more complete backend analysis functions, integrate with Unity's other ad placement components, and try to narrow the gap with Adjust.

• Considering so many product updates, Unity even promised some customers that its new generation of algorithms could have a 15% ROI update, which also means that advertisers may invest more on Unity next year.

So far, we cannot assert whether Unity can truly turn around. But a smart new Adtech leader + the right product direction can definitely narrow the capability gap with Applovin in the short term. The Adtech industry has always been very fair. Better ROI drives more budget, and competition is not a zero-sum game. If the marginal ROI meets customer requirements, customers will continue to increase their budget levels.

3.ZETA

Next, let's discuss an Adtech company with possibly the greatest potential.
ZETA is a very mysterious company. For a long time, we thought ZETA was a DMP+CDP company, and its competitors were companies like Salesforce Marketing Cloud, Adobe Experience, and Braze.

However, ZETA has also been rapidly improving its DSP capabilities in the past two years, and its main target is the largest advertiser industry - e-commerce.