Applovin Corp (APP) 2024 Q4 法說會逐字稿

Welcome to AppLovin's earnings call for the fourth quarter and year ended December 31, 2024.

歡迎參加 AppLovin 截至 2024 年 12 月 31 日第四季和年度收益電話會議。

I'm David Hsiao, Head of Investor Relations.

我是投資者關係主管 David Hsiao。

Joining me today to discuss our results are Adam Foroughi, our Co-Founder, CEO and Chairperson; and Matt Stumpf, our CFO.

今天與我一起討論我們的成果的有我們的共同創辦人、執行長兼董事長 Adam Foroughi；以及我們的財務長 Matt Stumpf。

Please note our SEC filings to date as well as our shareholder letter, Financial Update and press release discussing our fourth quarter annual performance are available at investors.applovin.com.

請注意，我們迄今為止提交給美國證券交易委員會 (SEC) 的文件以及我們的股東信函、財務更新和討論我們第四季度年度業績的新聞稿均可在 investor.applovin.com 上查閱。

During today's call, we will be making forward-looking statements, including, but not limited to, the future development and reach of our platform, our expected growth opportunities, the result and timing of our strategic transactions, the efficiency of our operations, the expected future financial performance of the company and other future events.

在今天的電話會議中，我們將做出前瞻性陳述，包括但不限於我們平台的未來發展和影響力、我們預期的成長機會、我們策略交易的結果和時機、我們營運的效率、公司預期的未來財務表現以及其他未來事件。

These statements are based on our current assumptions and beliefs.

這些陳述是基於我們目前的假設和信念。

We assume no obligation to update them, except as required by law.

除非法律要求，我們不承擔更新它們的義務。

Our actual results may differ materially from the results predicted.

我們的實際結果可能與預測的結果有重大差異。

We encourage you to review the risk factors that are most recently filed Form 10-Q for the third quarter ended September 30, 2024.

我們鼓勵您查看最近提交的截至 2024 年 9 月 30 日第三季的 10-Q 表的風險因素。

Additional information may also be found in our annual report on Form 10-K for the fiscal year ended December 31, 2024, which will be filed later this month.

您也可以在我們將於本月稍後提交的截至 2024 年 12 月 31 日的財政年度的 10-K 表年度報告中找到更多資訊。

We will also be discussing non-GAAP financial measures.

我們也將討論非公認會計準則財務指標。

These non-GAAP measures are not intended to be superior to or a substitute for our GAAP results.

這些非 GAAP 指標並非設計優於或取代我們的 GAAP 結果。

Please be sure to review the GAAP results and the reconciliations of our GAAP and non-GAAP financial measures in our earnings release and financial update available on our Investor Relations site.

請務必查看我們投資者關係網站上的收益報告和財務更新中的 GAAP 結果以及我們的 GAAP 與非 GAAP 財務指標的對帳表。

This conference call is being recorded, and a replay will be available for a period of time on our IR website.

本次電話會議正在錄音，重播將在一段時間內在我們的 IR 網站上提供。

Now I'll turn it over to Adam and Matt for some opening remarks, then we'll have the moderator take us through Q&A

現在我將請 Adam 和 Matt 致開幕詞，然後主持人將帶領我們進行問答

Thank you all for joining us.

感謝大家加入我們。

Q4 was a major milestone, arguably our most foundational period since the AXON upgrade in 2023.

Q4 是一個重要的里程碑，可以說是我們自 2023 年 AXON 升級以來最基本的時期。

For the first time, we captured meaningful holiday shopping advertising dollars and witnessed the impact of an advertising category beyond solely gaming contributing to our growth.

我們首次獲得了意義重大的假期購物廣告收入，並見證了遊戲以外廣告類別對我們成長的貢獻。

I'm sure many of you are curious about how much revenue our e-commerce category contributed.

我相信很多人都很好奇我們的電子商務類別貢獻了多少收入。

While we're not breaking out revenue by vertical because that's not how we view our business, I'd like to provide some perspective.

雖然我們沒有按垂直行業細分收入，因為這不是我們看待業務的方式，但我想提供一些觀點。

We operate a platform that reaches over 1 billion people in mobile games daily, with their engagement times comparable to social networks.

我們經營的平台每天可覆蓋超過 10 億手機遊戲用戶，他們的參與時間可與社交網路相媲美。

Historically, most of our ads focused on advertising for other games, but now we're attracting a broader set of advertisers.

從歷史上看，我們的大多數廣告都集中於其他遊戲的廣告，但現在我們正在吸引更廣泛的廣告商。

Q4 results show that our models can perform in other categories, in addition to continuing to improve performance for our gaming customers.

Q4 的結果表明，我們的模型除了能繼續為遊戲客戶提高效能之外，還能在其他類別中表現出色。

This breakthrough is only the beginning.

這項突破僅僅只是一個開始。

We've now also validated that our platform success isn't only limited to direct-to-consumer brands.

我們現在也證實，我們平台的成功不僅限於直接面向消費者的品牌。

Early pilots have shown positive outcomes for a range of advertisers, suggesting that any business in any vertical can harness the power of our platform.

早期試點已為一系列廣告商帶來了積極成果，這表明任何垂直領域的任何企業都可以利用我們平台的力量。

This opens up a massive opportunity as there are over 10 million businesses worldwide you advertise online that could eventually use our platform profitably.

這開啟了一個巨大的機會，因為全球有超過 1000 萬家企業在網路上投放廣告，他們最終可以使用我們的平台獲得利潤。

By delivering incremental value, we position ourselves as an engine for growth.

透過提供增量價值，我們將自己定位為成長引擎。

It's a win-win for brands, consumers and shareholders.

這對品牌、消費者和股東而言是雙贏的。

These early results solidify our vision of building one of the most influential marketing platforms in the world.

這些早期成果堅定了我們打造全球最具影響力的行銷平台之一的願景。

Where we once focus on gaming, we're now positioning ourselves to serve the entire global advertising economy.

我們曾經專注於遊戲，現在我們的定位是服務整個全球廣告經濟。

Importantly, the users engaging with our network aren't just shifting existing purchases.

重要的是，與我們的網路互動的用戶不僅僅是轉移現有的購買。

They're discovering new products while playing the games they love, generating truly incremental demand.

他們在玩自己喜歡的遊戲時發現了新產品，產生了真正的增量需求。

By enabling these discoveries, we're expanding the global economy for consumers and advertisers alike.

透過實現這些發現，我們正在為消費者和廣告商拓展全球經濟。

Demand from advertisers wanting to join our platform is high.

想要加入我們平台的廣告商的需求很高。

Currently, our systems are still being fully developed and lack the full self-service capabilities needed to handle growth at scale.

目前，我們的系統仍在全面開發中，缺乏處理規模成長所需的完整自助服務功能。

Our priority this year is to develop and roll out more automated tools to allow countless new businesses to tap into our platform.

我們今年的首要任務是開發和推出更多自動化工具，以便無數新企業能夠利用我們的平台。

In line with this expanding focus on advertising, we've been assessing how best to invest our resources to serve the needs of a global client base.

隨著對廣告的關注度不斷擴大，我們一直在評估如何最好地投資我們的資源來滿足全球客戶群的需求。

Seven years ago, we began acquiring gaming studios to help train our earliest machine learning models, an invaluable step in shaping the AI that underpins our AXON platform.

七年前，我們開始收購遊戲工作室來幫助訓練我們最早的機器學習模型，這是塑造支撐我們的 AXON 平台的人工智慧的重要一步。

However, we've never been a game developer at heart.

然而，我們從內心來說從來都不是一個遊戲開發者。

We have immense respect for the creativity it takes to build games, including from teams in our studios.

我們非常尊重製作遊戲所需的創造力，包括我們工作室的團隊。

Today, we're announcing we've signed an exclusive term sheet to sell all of our apps business.

今天，我們宣布我們已經簽署了一份獨家條款清單，出售我們所有的應用程式業務。

Matt will share further details, but I want to emphasize to our teams.

馬特將分享更多細節，但我想向我們的團隊強調。

You'll soon be part of a company that specializes in and champions game development.

您很快就會成為一家專門從事並倡導遊戲開發的公司的一員。

While it's bittersweet to part ways, we're excited for your future and immensely grateful for your role in getting us to where we are today.

儘管離別的滋味令人難過，但我們對您的未來感到興奮，並十分感謝您為我們取得今天的成就所做的貢獻。

Finally, I'd like to highlight our favorite metric going forward, adjusted EBITDA per employee.

最後，我想強調我們最喜歡的未來指標，即每位員工的調整後 EBITDA。

As we're transitioning to a pure advertising platform, our focus will be on productivity, automation and building lean, high-impact teams.

在我們轉型為純廣告平台的過程中，我們的重點將放在生產力、自動化和建立精實、高影響力的團隊。

In Q4, we had approximately $3 million in run rate adjusted EBITDA per employee in our advertising business, and we expect that number to rise as we refine processes and scale our business.

在第四季度，我們的廣告業務每位員工的調整後 EBITDA 約為 300 萬美元，我們預計，隨著我們完善流程和擴大業務規模，這一數字將會上升。

This metric underscores our commitment to operational excellence.

這項指標強調了我們對卓越營運的承諾。

Thank you for your continued support and partnership as we enter this next phase of growth.

感謝您在我們進入下一階段發展時給予的持續支持和合作。

I'm more confident than ever that we're building a platform with the potential to transform global marketing.

我比以往任何時候都更相信我們正在建立一個有潛力改變全球行銷的平台。

With that, I'll turn it over to Matt for a deeper look at our financials.

說到這裡，我將把話題交給馬特，讓他更深入地了解我們的財務狀況。

Thanks, Adam, and good afternoon.

謝謝，亞當，下午好。

I'm happy to announce we had another strong quarter with total revenue increasing 44% from the same period last year to $1.37 billion and adjusted EBITDA increasing 78% to $848 million, achieving a 62% adjusted EBITDA margin.

我很高興地宣布，我們又度過了一個強勁的季度，總收入比去年同期增長 44% 至 13.7 億美元，調整後 EBITDA 增長 78% 至 8.48 億美元，實現 62% 的調整後 EBITDA 利潤率。

This represents an 89% flow-through from revenue to adjusted EBITDA.

這意味著收入流向調整後 EBITDA 的比率為 89%。

In the fourth quarter, we generated $695 million in free cash flow, up 105% year-over-year.

第四季度，我們產生了6.95億美元的自由現金流，年增105%。

Quarter-over-quarter, our free cash flow grew 28%, representing 82% flow-through from adjusted EBITDA to free cash flow.

與上一季相比，我們的自由現金流成長了 28%，這意味著從調整後的 EBITDA 流向自由現金流的比率為 82%。

Free cash flow grew slightly more than our adjusted EBITDA growth over the same period due to the timing of cash tax payments.

由於現金納稅的時間安排，自由現金流的增幅略高於同期調整後的 EBITDA 增幅。

At the end of the fourth quarter, we had $741 million in cash and cash equivalents and 340 million shares outstanding.

截至第四季末，我們擁有 7.41 億美元現金和現金等價物以及 3.4 億股流通股。

During the quarter, our advertising business continued to drive increased performance for our mobile gaming partners combined with positive early results for e-commerce advertisers during the holiday season.

在本季度，我們的廣告業務繼續推動行動遊戲合作夥伴的業績成長，並在假期季節為電子商務廣告商帶來了積極的早期業績。

The advertising business generated $999 million in revenue and $777 million in adjusted EBITDA, achieving a 78% margin.

廣告業務創造了 9.99 億美元的收入，調整後 EBITDA 為 7.77 億美元，利潤率達到 78%。

Quarter-over-quarter flow-through from revenue to adjusted EBITDA was 75%, which is slightly lower than our normal levels.

營收與調整後 EBITDA 的季度環比流轉率為 75%，略低於我們的正常水準。

As I previewed last quarter, this was due to a step function increase in our data center costs.

正如我在上個季度預測的那樣，這是由於我們的資料中心成本大幅增加所致。

Flow-through will normalize from here as we gain leverage on this increase in GPU costs.

隨著我們能夠利用 GPU 成本的增加來提高效率，流通將從現在開始正常化。

Before we get into the financial performance for our Apps business, as Adam mentioned, we're excited to announce we've signed a term sheet to divest our Apps business.

正如亞當所提到的，在我們了解應用程式業務的財務表現之前，我們很高興地宣布，我們已經簽署了一份條款清單，以剝離我們的應用程式業務。

Total estimated consideration is $900 million, including $500 million in cash with the remainder representing a minority equity stake in the combined private company.

總對價估計為 9 億美元，包括 5 億美元現金，剩餘部分為合併後私人公司的少數股權。

Subject to regulatory clearance, we hope to close this transaction in the coming quarter and look forward to seeing the success of this business under new leadership.

在獲得監管部門批准後，我們希望在下個季度完成交易，並期待在新的領導下看到這項業務取得成功。

Our Apps revenue for the quarter was $373 million, a 1% decrease from last year with $71 million in adjusted EBITDA, representing a 19% margin.

本季我們的應用程式營收為 3.73 億美元，較去年同期下降 1%，調整後 EBITDA 為 7,100 萬美元，利潤率為 19%。

Turning briefly to our annual results.

簡單介紹一下我們的年度業績。

Revenue for the year was $4.7 billion.

全年收入為 47 億美元。

That's an increase of 43% from last year.

比去年增加了 43%。

Adjusted EBITDA was $2.72 billion, that's an incredible 81% increase from last year at an adjusted EBITDA margin of 58%.

調整後的 EBITDA 為 27.2 億美元，與去年同期 58% 的調整後 EBITDA 利潤率相比，成長了驚人的 81%。

Free cash flow for the year was $2.1 billion, representing an impressive 76% flow-through from adjusted EBITDA of $2.72 billion.

全年自由現金流為 21 億美元，佔調整後 EBITDA 27.2 億美元的 76% 左右。

This quarter, we withheld a total of 1.6 million shares for a total cost of $508 million.

本季，我們共計扣留了 160 萬股，總成本為 5.08 億美元。

For the full year, we repurchased or withhold a total of 25.7 million shares for a total cost of $2.1 billion.

全年我們共回購或扣留了 2,570 萬股，總成本為 21 億美元。

I want to pause here to emphasize this point.

我想在這裡停下來強調這一點。

During the year, we had $2.1 billion of free cash flow and spent $2.1 billion on our shares at a weighted average price of approximately $83 per share, illustrating the commitment we've communicated to drive shareholder value through prudent capital allocation and an investment in our own shares.

本財年，我們擁有 21 億美元的自由現金流，並以每股約 83 美元的加權平均價格收購了 21 億美元的股票，這體現了我們透過審慎的資本配置和對我們自己股票的投資來推動股東價值的承諾。

Finally, turning to our financial guidance for next quarter.

最後，談談我們對下一季度的財務指引。

In light of the transaction we highlighted here, and the continued focus on the advertising business, we will provide guidance for each of our segments separately.

鑑於我們在此強調的交易以及對廣告業務的持續關注，我們將為每個部門分別提供指導。

In the first quarter of 2025, for the advertising business, we anticipate to deliver between $1.030 billion and $1.050 billion in revenue, with adjusted EBITDA between $805 million and $825 million, targeting an adjusted EBITDA margin of 78% to 79%.

2025 年第一季度，對於廣告業務，我們預計實現 10.3 億美元至 10.5 億美元之間的收入，調整後的 EBITDA 介於 8.05 億美元至 8.25 億美元之間，調整後的 EBITDA 利潤率目標為 78% 至 79%。

We expect Apps revenue to be between $325 million and $335 million at an adjusted EBITDA of between $50 million and $60 million.

我們預計應用程式收入將在 3.25 億美元至 3.35 億美元之間，調整後的 EBITDA 將在 5,000 萬美元至 6,000 萬美元之間。

Now with that, let's move to Q&A.

現在，我們進入問答環節。

(Operator Instructions) Clark Lampen, BTIG.

(操作員指示) Clark Lampen，BTIG。

I'm curious, Adam, a lot to dive into, I think, around commerce.

亞當，我很好奇，我認為在商業方面有很多東西需要深入研究。

But I want to make sure that I understood, I think, one point that you made in your prepared remarks, which is you're seeing early benefits for a range of different brands in a variety of verticals beyond just sort of DTC marketers.

但我想確保我了解您在準備好的發言中提到的一點，那就是您看到了一系列不同品牌在各個垂直領域中獲得的早期利益，而不僅僅是 DTC 行銷人員。

Does that suggest that as you scale sort of your nongaming business over the course of the year, is that as we move into self-serve as we're addressing more global audiences, it's not going to be just DTC marketers.

這是否意味著，隨著你們在一年內擴大非遊戲業務的規模，隨著我們進入自助服務領域，面向更多的全球受眾，我們將不只是 DTC 行銷人員。

Or is that an opportunity that you sort of identified as one that works, but maybe we're not necessarily capitalizing on that near term.

或者您認為這是一個可行的機會，但也許我們不一定會在短期內利用它。

Thanks, Clark.

謝謝，克拉克。

You basically answered it for me.

你基本上已經回答我的問題了。

But when we put the platform together, we really strategically made a choice to go after DTC commerce.

但當我們將平台整合在一起時，我們確實從策略上選擇了追求 DTC 商務。

But what we knew is that if the tech works there, it's going to work on everything.

但我們知道，如果這項技術在那裡有效，那麼它就能在一切事物上有效。

And so sort of privately and quietly as we've gotten onboarding from clients in other categories across a wide range of categories, we've been seeing it the models are working there.

因此，當我們私下、悄悄地從各個類別的其他類別的客戶那裡獲得支持時，我們就會看到模型在那裡發揮作用。

And what I was announcing just a couple of minutes ago is that we're seeing success across any category that comes on to the platform, which gives us a lot of confidence that as we go through the year, we release more tools and really go into more of a self-service and open state.

幾分鐘前我剛剛宣布，我們看到平台上所有類別的產品都取得了成功，這給了我們很大的信心，相信隨著時間的流逝，我們會發布更多的工具，真正進入更加自助和開放的狀態。

We're going to be in a position to go after the very, very large set of advertisers that are in the $10 million plus range of the world over the next come quarters, years and decade plus.

在接下來的幾季、幾年乃至十年裡，我們將有能力吸引全球金額在 1,000 萬美元以上的龐大廣告客戶群。

This is something that we aspire to do, and we're going to take our time on because I do want to caveat it with an important point that ties back to our culture.

這是我們渴望做的事情，我們會花時間去做，因為我確實想用一個與我們的文化相關的重要觀點來強調這一點。

We're spending quite a lot of time trying to build tools to automate the entire system so that we're not looking at a massive revenue opportunity in front of us and suggesting that we have to go hire people to go after it.

我們花了大量的時間來嘗試建立工具來實現整個系統的自動化，這樣我們就不會看到面前有巨大的盈利機會而去僱人去實現它。

We're going to continue to stay diligent and lean, and we're going to use tools and AI and automation to deliver the same type of solution and success for advertisers that come on board in the future but make sure they come on board into a platform that hasn't changed anything from the cultural values that have gotten us to this point.

我們將繼續保持勤奮和精簡，我們將使用工具、人工智慧和自動化來為未來加入的廣告商提供相同類型的解決方案和成功，但確保他們加入的平台沒有改變我們走到今天的文化價值。

Okay.

好的。

Within that same list that you just sort of talked about with the dashboard automation at the bottom of it was CTV advertising.

在您剛才談到的同一個清單中，儀表板自動化位於底部，是 CTV 廣告。

And I don't want to do a full audit of sort of the Wurl deal, but I'm curious if you could give us synopsis of sort of maybe what happened last time around that proved to be a little bit different relative to that initial thesis?

我不想對 Wurl 交易進行全面審計，但我很好奇，您是否可以向我們簡要介紹一下上次發生的事情，與最初的設想相比，這次交易有何不同？

And now maybe what's changed lately that at this point in time has prompted you to, I guess, shift focus or maybe focus more on that supply opportunity.

也許最近發生了一些變化，我想此時此刻促使你轉移注意力或更多地關注供應機會。

Yes.

是的。

So I'll say we haven't really focused on it yet at all.

所以我想說我們還沒有真正關注它。

When we bought Wurl, the idea was that they're connected to the media companies, and we wanted to get supply online, not too dissimilar from the MAX auction so that we have the access to supply.

當我們收購 Wurl 時，我們的想法是，他們與媒體公司有聯繫，我們希望在網上獲得供應，與 MAX 拍賣沒有太大區別，這樣我們就可以獲得供應。

Our core business and the vast majority of revenue comes from the DSP side.

我們的核心業務和絕大部分收入來自DSP方面。

So advertisers coming to us and us helping them place that.

因此廣告商來找我們，我們幫他們投放廣告。

But we need supply.

但我們需要供應。

Wurl's brought a ton of supply online.

Wurl 在網路上提供了大量貨物。

And last year, what we had available to us to show into the [ad slot] were gaming ads.

去年，我們在廣告位上展示的是遊戲廣告。

That part of the platform is live.

該平台的這一部分是直播的。

But if you think about gaming out on a full screen television asking user to go download a game on the phone, it's just not that compelling a full screen TV ad.

但如果你想在全螢幕電視上玩遊戲並要求用戶去手機上下載遊戲，那麼全螢幕電視廣告就沒有那麼引人注目了。

We didn't see a single gaming ad on the Super Bowl, right?

我們在超級盃上沒有看到任何遊戲廣告，對吧？

What we can expect though is that consumer adds in D2C commerce and then across a wide range of categories like fintech, automotive, et cetera, as these come on to our platform, extending that type of creative onto the big screen could be quite compelling.

然而，我們可以預期的是，隨著消費者對 D2C 商務以及金融科技、汽車等廣泛類別的關注，隨著這些內容進入我們的平台，將這種類型的創意擴展到大螢幕上可能會非常引人注目。

Now it's not without its challenges.

如今，它也面臨挑戰。

There are attribution problems.

存在歸因問題。

There is a lack of a call to action.

缺乏行動的呼籲。

So there are difficulties to make this work, which is why we're not saying, hey, this is a layup.

因此，要實現這一目標有一定的困難，這就是為什麼我們不會說，嘿，這是一個上籃得分。

It's a challenge.

這是一個挑戰。

If we are able to do it and do it successfully, it will open up another massive channel of performance advertising on the big screen, probably for the first time in history, and that's why I keep mentioning it because it is a huge opportunity and one that we're going to be going after building into this year.

如果我們能夠做到這一點並取得成功，那麼它將在大螢幕上開闢另一個巨大的效果廣告管道，這很可能是歷史上的第一次，這就是我一直提到它的原因，因為這是一個巨大的機會，也是我們今年將要努力打造的機會。

Ralph Schackart, William Blair.

拉爾夫·沙卡特、威廉·布萊爾。

First question is just on model enhancements.

第一個問題只是關於模型增強。

Adam, I know each quarter it could be volatile, unpredictable.

亞當，我知道每個季度都可能動盪不安、難以預測。

But can you just maybe give a sense this quarter, what you saw in terms of model enhancements and how much that may have been additive to growth?

但是，您能否介紹一下本季您在模型增強方面看到了什麼，以及這些增強對成長的促進作用有多大？

And then I have a follow-up.

然後我有一個後續問題。

Yes, Ralph.

是的，拉爾夫。

So we've sort of broken down the growth vectors across a couple of different facets.

因此，我們將成長向量分解到幾個不同的方面。

One is just ongoing learning where the model is just improving itself that we expect to capture every single quarter, we haven't seen a slowdown there.

一是持續的學習，模型不斷完善，我們預計每季都會捕捉到這一變化，我們還沒有看到任何放緩的跡象。

That contributed to growth in the quarter.

這促進了本季的成長。

There's then another facet which is iterative and incremental improvements on our pre-existing model.

另一個面向是我們現有模型的迭代和漸進式改進。

That sort of happens along the way as we go.

在我們前進的過程中，這樣的事情時常發生。

We don't break those out.

我們不會將這些打破。

Those are, again, unpredictable and not huge, but those are not those step functions that you've seen in past quarters.

再次強調，這些都是不可預測的，也不是很大的，但這些不是你在過去幾季看到的那些階梯函數。

We had some of that in the quarter as well.

本季我們也遇到了一些這樣的情況。

Then you've got these big incremental lifts to a new model.

然後你就得到了新模型的巨大增量提升。

So analogy I've drawn is to ChatGPT 4.0, 4.1, et cetera.

所以我將其與 ChatGPT 4.0、4.1 等等進行了類比。

And so as their models increment, the model gets more complex and smarter, we have the same thing.

因此，隨著他們的模型不斷增加，模型變得更加複雜和智能，我們也得到了相同的東西。

This quarter, we did not have one of those.

本季度，我們還沒有遇到這種情況。

So then you see this outsized growth.

所以你會看到這種超大的成長。

And then the question is, okay, where did it come from?

那麼問題是，好吧，它是從哪裡來的？

And I touched on this in my talk track.

我在我的演講中提到了這一點。

We've got seasonality in Q4, which traditionally we've never been able to take advantage of.

我們第四季受到季節性的影響，傳統上我們從未能利用這一點。

And that's built around shopping and shopping behavior.

這是圍繞著購物和購物行為建立的。

And that this quarter, we were able to take some advantage of because we're starting to see the e-commerce product really take off.

本季度，我們能夠利用這一點，因為我們開始看到電子商務產品真正起飛。

The other part of it is that in Q4, you also have people having more time to spend on mobile devices, especially during the holidays.

另一部分原因是，在第四季度，人們在行動裝置上花費的時間更多，尤其是在假期期間。

And during that time, people are more likely to consume and transact in mobile games.

在此期間，人們更有可能在手機遊戲中消費和交易。

So naturally, every single year, Q4 is a very good healthy quarter for marketing mobile games.

因此自然而然地，每年的第四季都是手機遊戲行銷的一個非常好的健康季度。

So it was a combination of the baseline, which is just our models are continuously getting smarter on their own, some team effort on incremental improvements on the models and then seasonality across both our existing games business and our new up and coming e-commerce business.

因此，它是基線的組合，也就是我們的模型本身不斷變得更加智能，一些團隊努力對模型進行逐步改進，然後是現有遊戲業務和新興電子商務業務的季節性。

Great.

偉大的。

Just maybe a follow-up and I'm not sure if it's for you or Matt.

也許只是一個後續行動，我不確定它是否適合您或馬特。

But just in terms of e-commerce, obviously, a lot of excitement here and things seem to be going really well.

但僅從電子商務方面來看，顯然這裡有很多令人興奮的事情，而且事情似乎進展得非常順利。

Do you still -- I think before you've talked about it contributing to be a material contribution in 2025.

您是否仍 - 我認為在您談到它之前，它將成為 2025 年的實質貢獻。

Just curious if you still I think that's true, A?

我只是好奇你是否仍然認為那是真的，A？

And then B, can you just give us a sense of just the potential for incremental growth or how you might frame the opportunity as we're sort of thinking about our models for '25.

然後 B，您能否向我們介紹增量成長的潛力，或者當我們思考 25 年的模型時，您如何建立機會。

We still feel very confident, Ralph and the ability for us to contribute a material portion of revenue from the e-commerce opportunity in 2025.

我們仍然非常有信心，拉爾夫，我們有能力在 2025 年從電子商務機會中貢獻出可觀的收入。

But again, it's very difficult to predict when and how much that growth is going to come in during the period in 2025.

但同樣，很難預測 2025 年期間這種成長何時實現、以及實現多少。

So I think we'll see it when it comes, and then we'll provide more details.

所以我認為我們會看到它，然後我們會提供更多細節。

And then what I've given consistently is, look, our business predictably has got 20% year-over-year growth in it.

我一直強調的是，我們的業務可預見地實現了 20％ 的同比增長。

And that's from just the baseline of operating within the category that we've always traditionally operated.

這只是從我們一貫傳統營運的類別的營運基礎來看的。

And then you've got these things that can drive material upside that's really exciting for us.

然後你就得到了這些可以推動物質上漲的東西，這對我們來說真的令人興奮。

They're unpredictable.

它們難以預測。

And when things are unpredictable, a lot of times people assume they're going to be unpredictably bad.

當事情變得不可預測時，很多時候人們會認為事情會變得不可預測的糟糕。

Well, in our business, we've got 2 things that could be unpredictably really good.

嗯，在我們的業務中，有兩件事情可能會出乎意料地好。

One is these model enhancements.

一是這些模型增強。

We're really, really early still in AI development and research, both inside our company and then obviously externally, too.

我們在人工智慧開發和研究方面仍處於早期階段，無論是在公司內部還是外部。

This is not a field that's mature.

這不是一個成熟的領域。

It's early in its existence.

它的存在還很早期。

It's only going to get better.

情況只會越來越好。

And when our models get better, we've all seen the impact to revenue.

當我們的模型變得更好時，我們都看到了對收入的影響。

And the other piece is, we've always been a closed managed platform, specifically for mobile games.

另一點是，我們一直都是封閉的管理平台，特別是針對手機遊戲。

We now have a lot of proof of life in e-commerce.

我們現在有很多電子商務生命力的證據。

You've seen it on Twitter, a lot of the noise from customers that are in.

您可以在 Twitter 上看到，許多來自客戶的噪音。

But we have not let a lot of customers onto the platform yet as we've been on pilot.

但由於我們仍處於試點階段，因此尚未讓許多客戶加入該平台。

As we go more and more open and start attracting thousands and tens of thousands, hundreds of thousands of customers to come on over the coming quarters and years, the business is going to continue to show compelling growth.

隨著我們越來越開放，並在未來幾季和幾年內開始吸引成千上萬的客戶，業務將繼續呈現強勁成長。

We look at it as one single business and better monetizing the 1 billion-plus daily actives that we see.

我們將其視為單一業務，並更好地將我們看到的每天超過 10 億的活躍用戶貨幣化。

We don't think about it as revenue from each category matters.

我們並不認為每個類別的收入都很重要。

Jason Bazinet, Citi.

花旗銀行的 Jason Bazinet。

Would it be possible to share how many people are on your pilot now for e-commerce?

您能否透露一下，目前有多少人參與了您的電子商務試點？

We've not broken it out.

我們還沒有把它分解開來。

I mean, if you do industry checks, look at Twitter, there's enough noise where you can assume it's not in the 10s.

我的意思是，如果你進行行業檢查，看看 Twitter，那裡有足夠的噪音，你可以假設它不在 10 左右。

It's certainly not in the thousands.

這個數字肯定不會達到數千。

We're limited by core team.

我們受到核心團隊的限制。

It would surprise you how small we maintain our teams.

我們維持的團隊規模之小可能會讓您感到驚訝。

I mean I do talk about this EBITDA by employee number.

我的意思是我確實談論了按員工人數計算的 EBITDA。

We really do take that seriously.

我們確實非常重視這一點。

We've got around 1,000 people on the entire advertising business, including all the segments of it, Adjust, Wurl and AppLovin businesses.

我們整個廣告業務約有 1,000 名員工，包括所有部門、Adjust、Wurl 和 AppLovin 業務。

The e-commerce go-to-market team is roughly 20 people.

電商上市團隊大概有20個人。

So to give you a sense of how lean and automated we run, that should give you a sense, but that team is not going to be able to manually onboard a lot of advertisers in a short amount of time.

因此，為了讓你了解我們的運行是多麼精益和自動化，這應該會讓你有一個概念，但該團隊無法在短時間內手動接納大量廣告商。

Great.

偉大的。

And this goes back to your self-serve comment that you made, right?

這又回到了你剛才發表的自私評論，對嗎？

That's really what's going to open up the aperture to onboard more clients.

這確實會打開大門吸引更多客戶。

Self-serve and automated tools are going to be really, really helpful.

自助和自動化工具將會非常非常有用。

And so if you look at just the numbers, -- what gets us really excited is in a limited pilot of a few customers on the platform, we're driving actually interesting revenue from this category.

所以，如果你只看數字，真正讓我們興奮的是，在平台上對少數客戶進行有限試點後，我們從這個類別中獲得了真正有趣的收入。

So as we start opening up, we think it's going to be really impactful to the businesses of our clients.

因此，當我們開始開放時，我們認為這將對我們客戶的業務產生真正的影響。

It's also very, very impactful for the publishers that we work with all this inventory is mobile games.

對於我們合作的發行商來說，這也產生了非常大的影響，因為這些庫存都是手機遊戲。

And it adds variety to the advertising that the customer is getting on a mobile game.

這也為客戶在手機遊戲中獲得的廣告增添了多樣性。

It also removes constantly the customer getting an ad on a gaming publisher for a competitor of that publisher.

它還會不斷消除客戶在遊戲發行商上看到該發行商競爭對手的廣告的情況。

So that's really, really valuable because as you think about the growth vectors that we've got just embedded into the business, better monetizing the inventory that we have, but also bringing on more supply.

所以這真的非常有價值，因為當你考慮到我們剛剛嵌入業務的成長向量時，我們可以更好地將現有庫存貨幣化，同時也帶來更多的供應。

A lot of gaming publishers, the biggest ones, have traditionally been fearful of bringing on supply because they do not want to show their competition ads in their games.

許多遊戲發行商，尤其是大型遊戲發行商，傳統上不敢引入供應，因為他們不想在自己的遊戲中展示競爭對手的廣告。

Well, now that we're able to show that we have expertise marketing to any category, that's going to open up a ton more supply to come online, especially into our MAX platform over the coming years as well.

現在，我們已經能夠證明我們擁有針對任何類別的行銷專業知識，這將在未來幾年內開闢更多的線上供應，尤其是在我們的 MAX 平台上。

Vasily Karasyov, Cannonball.

瓦西里‧卡拉紹夫，《砲彈》。

Congratulations.

恭喜。

Adam, I wanted to get this bigger picture question.

亞當，我想了解這個更大的問題。

As you are rolling out your e-commerce solution offering, are you seeing already any kind of response from the incumbent competitors?

當您推出電子商務解決方案時，您是否已經看到現有競爭對手的任何反應？

Because if I were one of them, I would look at your success in gaming, right?

因為如果我是他們中的一員，我會關注你在遊戲方面的成功，對嗎？

And we probably think I need to do something about it.

我們可能認為我需要對此採取一些行動。

So I was wondering how -- what you're seeing, are you prepared for this?

所以我想知道——您所看到的，您對此做好準備了嗎？

And in general, once you roll it out to full scale, what they think the industry will look like?

整體而言，一旦全面推廣，他們認為這個產業會是什麼樣子？

How will it be different from what it is today?

它與今天有何不同？

Yes.

是的。

So I mean, look, we don't look at competition all that much.

所以我的意思是，我們並不太關注競爭。

What I will say is that we're not a platform that's taking the same dollars away from someone else.

我想說的是，我們不是從別人手中奪走錢財的平台。

So let's compare it to social.

讓我們將其與社交進行比較。

If you've got a mattress manufacturer advertising on social today, and driving a certain amount of business and they come on to our platform, what they're seeing are new transactions from customers that they wouldn't have otherwise gotten to respond to their ads.

如果今天有一家床墊製造商在社群媒體上投放廣告，並帶來一定數量的業務，當他們進入我們的平台時，他們會看到來自客戶的新交易，而這些客戶原本是不會對他們的廣告做出回應的。

Whether those customers who are on social or not might be an issue.

那些是否參與社交的客戶可能會是一個問題。

But certainly, there's a lot of overlap but a lot of customers just won't notice ads in one environment.

但可以肯定的是，存在很大的重疊，但許多客戶在一個環境中不會注意到廣告。

Now in our environment, we have a full-screen video ad that captures attention, and they come on to our platform, and they're driving incremental sales.

現在在我們的環境中，我們有一個全螢幕影片廣告來吸引註意力，他們進入我們的平台，並推動增量銷售。

So what does that mean?

那麼這意味著什麼？

That means that if they were spending $5,000 a day on social, they're not going to come to us and say, we're going to spend $5,000 a day on you in a performance manner and take the $5,000 over here down to 0, they're now going to be spending $10,000 a day.

這意味著，如果他們每天在社交上花費 5,000 美元，他們不會來找我們說，我們將以績效的方式每天在你身上花費 5,000 美元，然後將這裡的 5,000 美元減少到 0，他們現在每天要花費 10,000 美元。

And this is what I keep talking about and referring to is we're expanding the economy because it's incremental and entirely performance-based.

這就是我一直在談論和提到的，我們正在擴大經濟，因為它是漸進的並且完全基於績效。

The advertisers are getting measurable profits from the marketing that's going to expand their business.

廣告商從行銷中獲取了可觀的利潤，從而擴大了他們的業務。

They're going to then have to go get more inventory.

然後他們將必須去獲取更多的庫存。

They're going to then be able to sell more product and the whole economy will expand without causing any detrimental harm to competitors in the

他們將能夠銷售更多產品，整個經濟將擴張，而不會對競爭對手造成任何損害。

(inaudible)

（聽不清楚）

Rob Sanderson, Loop Capital.

桑德森（Rob Sanderson），Loop Capital。

I Wanted to ask a question on your go-to-market for the e-commerce and now new verticals.

我想問一個有關你們在電子商務和現在的新垂直市場的營銷策略的問題。

Like the playbook seems really simple, and I guess that's part of the strategy.

就像劇本看起來很簡單，我猜這就是策略的一部分。

Can you kind of discuss how the strategy may have evolved a little since the pilot to not through holiday season, any key learnings or pivots.

您能否討論一下自試行以來，該策略可能發生了哪些變化，以及是否經歷了任何關鍵的經驗或轉變？

Obviously, self-serve is going to be a big unlock.

顯然，自助服務將成為一個巨大的進步。

But just how are you thinking about going to market?

但您究竟想如何進入市場呢？

And then anything you can share on priority areas for demand generation in 2025?

那麼您能分享一下 2025 年需求產生的優先領域嗎？

And then I have a follow-up.

然後我有一個後續問題。

Yes.

是的。

We've been really exclusively focused on sort of mid-market D2C.

我們一直專注於中階市場 D2C。

I call that somewhere in the neighborhood of $10 million to $250 million of GMV.

我預計 GMV 大約在 1,000 萬至 2.5 億美元之間。

These are companies that tend to move fast, and they don't need a big team on the other side to bring them on board.

這些公司往往行動迅速，並不需要對方的龐大團隊來帶領他們。

And the legal process is much, much shorter.

而且法律程序也短得多。

We haven't gone after the very large brands.

我們還沒有追逐過非常大的品牌。

So no one in this pilot are companies that are incredibly large.

因此，本次試點中沒有一家公司規模特別大。

And then we're not focused on the very, very small while we have to manually go after these folks.

然後，我們就不會把注意力集中在非常非常小的事情上，而必須手動追蹤這些人。

In the middle, because of the social noise and the feedback loop, the really positive results that customers are getting early in the pilot we've got just customers coming to us.

在中間，由於社會噪音和回饋循環，客戶在試點初期獲得的真正正面成果只是來自客戶本身。

So there's a long, long line out the door, which means we're onboarding as we can onboard.

所以門外排起了很長的隊，這意味著我們正在盡可能地接待乘客。

Nothing has changed through the holidays.

假期過後，什麼都沒有改變。

The teams just bombarded with work.

各個團隊都忙於工作。

They had to work a ton during Black Friday.

他們必須在黑色星期五期間付出大量工作。

But other than just the really the sheer volume that we're trying to process on helping clients get onto the platform and see that positive result, there's nothing else that's changed or will change until we get the tools out the door that will allow any customer to come in a self-service manner to be able to unlock the same type of result that the folks right now are getting with our managed support.

但是，除了我們試圖處理的大量工作，幫助客戶進入平台並看到積極成果之外，在我們推出工具之前，其他一切都不會改變，任何客戶都可以以自助服務的方式獲得與現在人們透過我們的託管支援所獲得的相同類型的結果。

Okay.

好的。

And then a question, how much category-specific optimization is required to move sort of these non-DTC adjacencies, like does AXON kind of figured this set on its own, you start driving conversions and the model starts to figure it out and maybe the team can steer it with vertical-specific tweaks or whatever?

然後是一個問題，需要多少針對特定類別的最佳化來移動這些非 DTC 鄰接，例如 AXON 是否可以自行計算這組數據，您開始推動轉換，模型開始計算它，也許團隊可以透過垂直特定的調整或其他方式來引導它？

Or how much product work is required to sort of keep expanding for these non-DTC categories?

或者說需要做多少產品工作才能不斷擴大這些非 DTC 類別？

Yes, I referenced it because it works out of the box.

是的，我引用它是因為它開箱即用。

So we don't have the resources to be building custom models for every single vertical at least today.

所以至少今天我們沒有足夠的資源去為每個垂直行業建立客製化模型。

The goal is to continue to build a more complex model that's smart enough to figure out how to market any product on the other side.

目標是繼續建立一個更複雜的模型，該模型足夠智能，能夠弄清楚如何在另一端行銷任何產品。

That's what we're seeing, where these companies in other categories across a wide range are just plugging in and it works.

這就是我們所看到的，各個領域中其他類別的公司只需加入進來就能發揮作用。

So the challenge is really just go to market and creating awareness of those types of advertisers and whatnot.

因此，真正的挑戰是進入市場並提高此類廣告商的知名度等等。

100%.

100%。

We'll have a challenge of get the tools out so that we can bring more and more companies on.

我們面臨的挑戰是如何推出工具，以便能夠吸引越來越多的公司加入。

And then we'll always have this challenge that we're trying to sell a product where they're able to advertise and do it profitably and have it be incremental and have all of that be measurable.

然後，我們總是面臨這樣的挑戰，我們試圖銷售一種產品，他們能夠做廣告，並且盈利，讓它逐漸增加，並且所有這些都是可衡量的。

So long as that is the result across every single advertiser relationship with our platform, there'll be a lot of positive noise out there about the platform and the offering, which will continue to bring us customers in an organic manner.

只要我們的平台與每個廣告商的關係都能產生這樣的結果，那麼這個平台和產品就會得到很多正面的評價，並將繼續以有機的方式為我們帶來客戶。

Jim Callahan, Piper Sandler.

吉姆·卡拉漢，派珀·桑德勒。

Congrats on the nice report.

恭喜您出色的報告。

So we've been getting some questions on how you're servicing some of the extra e-commerce demand.

我們一直在問您如何滿足一些額外的電子商務需求。

Can you talk about maybe like what kind of (inaudible) and if perhaps this is different from the inventory that converts better for gaming?

您能否談談可能是什麼樣的（聽不清楚）以及這是否與更適合遊戲的庫存不同？

It's all the same inventory.

全部都是相同的庫存。

It's full-screen video ad that sits on top of the 1 billion-plus daily active users playing games, but we don't have access to differentiated inventory.

這是針對每天超過 10 億活躍遊戲用戶推出的全螢幕影片廣告，但我們無法獲得差異化廣告資源。

It's a unified auction.

這是一次統一拍賣。

Sometimes an ad for a mattress will be the one that selected sometimes an ad for Candy Crush will be the one that selected.

有時所選的是床墊廣告，有時所選的是 Candy Crush 廣告。

It all depends on what the model does.

這一切都取決於模型的作用。

But the end user and the end supply is all the same across all categories.

但所有類別的最終用戶和最終供應都是相同的。

Got it.

知道了。

Okay.

好的。

That is helpful.

這很有幫助。

And apologies if you talked about this in the beginning, I was having some audio issues.

如果您一開始就談到了這一點，我很抱歉，因為我遇到了一些音訊問題。

It looks like you're rolling out 28-day models.

看起來您正在推出 28 天模型。

Has there been any sort of early feedback here on the gaming side?

在遊戲方面有沒有什麼早期回饋？

And could that drive kind of incremental spend this year?

這是否會導致今年的支出增加？

Those have been out for, I think, a couple of quarters now.

我認為這些已經推出幾個季度了。

And obviously, I think if you talk to game developers across the border, models tend to work really in a compelling way for all of them, the longer models are the most highly most performant models that we deliver.

顯然，我認為如果你與跨境的遊戲開發者交談，模型往往會以令人信服的方式為他們所有人發揮作用，我們提供的模型越長，性能就越高。

And so they have driven up the ability for the mobile customer to spend more profitably on our platform over the last couple of quarters.

因此，在過去幾個季度中，他們提高了行動客戶在我們的平台上消費以獲取更多利潤的能力。

James Heaney, Jefferies.

傑富瑞的詹姆斯希尼。

One of the big questions that we've been getting from investors is just around the take rate dynamics between gaming and e-commerce.

投資者向我們提出的一個大問題是有關遊戲和電子商務之間的接受率動態。

I understand that you don't charge a fixed rate like other ad tech platforms.

我知道你們不像其他廣告科技平台那樣收取固定費率。

But just would be curious how to think about just the overall mix of the business as you go more towards e-com.

但是我很好奇，當您進一步走向電子商務時，如何考慮整體業務組合。

Does that have any impact on the take rate?

這對接受率有影響嗎？

And then I had a follow-up.

然後我進行了後續行動。

Yes.

是的。

Look, like as we get better at monetizing the inventory that we buy the take rate naturally goes up.

看起來，隨著​​我們越來越善於將購買的庫存貨幣化，利潤率自然就會上升。

On the flip side, we don't optimize the take rate.

另一方面，我們並沒有優化接受率。

We never have.

我們從來沒有。

So the system is willing to buy a user of 1% or 99%.

所以系統願意購買1%或99%的用戶。

It really doesn't matter.

這真的不重要。

It's going out and buying users that it thinks it can out monetize from what it has to pay regardless of what the end result is, but it is important to understand that there is a market we're paying the highest price, the publishing can get from anywhere.

它會出去購買用戶，並認為無論最終結果如何，它都能從所支付的費用中獲得豐厚的利潤，但重要的是要明白，我們支付最高價格的市場可以從任何地方獲得。

And the better and better that we get, the stronger our business is going to get because that take rate can expand.

我們做得越好，我們的業務就會越強大，因為接受率會擴大。

Yes, that's helpful.

是的，這很有幫助。

And Matt, now that e-commerce is obviously becoming a much bigger percentage of your business, can you just talk about the seasonality and how we should be thinking about that going forward as it relates to both the revenue growth and the expense profile?

馬特，現在電子商務顯然已經成為貴公司業務中越來越重要的組成部分，您能否談談季節性因素以及我們應該如何看待它與收入增長和費用狀況的關係？

Yes.

是的。

On the expense side, it shouldn't affect us much whatsoever.

從費用方面來說，它不會對我們造成太大影響。

Obviously, the expenses are primarily data center costs and payroll.

顯然，費用主要是資料中心成本和工資。

So those are -- those costs are more driven by usage rate, and we've kind of communicated in the past that as we grow revenue that we're expecting roughly 10% on an annual basis is that drag from incremental data center costs as we grow revenue.

所以這些成本更多是由使用率驅動的，我們過去曾表示，隨著收入的增長，我們預計年增長率約為 10%，這是隨著收入的增長而增加的數據中心成本的拖累。

So there shouldn't be much fluctuation there.

因此那裡不應該有太大的波動。

And then on the e-commerce side of the business, as we grow and e-commerce becomes a larger portion of the overall pie and as we start to move towards it being a material portion of the overall revenue, obviously, we'll be subject then to e-commerce seasonality and everything we've talked about in the past, obviously, Black Friday and the holiday periods, we'll see slightly larger periods of revenue.

在電子商務業務方面，隨著我們的成長，電子商務在整個業務中所佔的份額越來越大，並且我們開始朝著讓電子商務成為總收入的重要組成部分的方向發展，顯然，我們會受到電子商務季節性的影響，以及我們過去談到的一切，顯然，在黑色星期五和假期期間，我們將看到收入略有增加的時期。

Chris Kuntarich, UBS.

瑞銀的 Chris Kuntarich。

Can you just talk a little bit more about the 1Q sequential growth drivers of the e-commerce business?

您能否再談談電子商務業務第一季環比成長的動力？

Is this -- should we be thinking about this more same advertiser spend or more driven so by new incremental advertisers?

這是——我們應該考慮更多相同廣告商的支出還是更多地由新的增量廣告商驅動？

Well, we haven't broken out e-commerce, right?

嗯，我們還沒有突破電子商務，對吧？

But 1Q -- in a traditional advertising business, typically, 1Q steps down.

但在傳統廣告業務中，通常情況下，1Q 會下降。

Two less days and you've got seasonality effect in the fourth quarter.

少了兩天，第四季就會受到季節性影響。

So we would have that for sure in the e-commerce category across the current advertiser base.

因此，我們肯定會在當前廣告客戶群的電子商務類別中實現這一點。

If we didn't add a single new advertiser, then by definition, that category could expand in Q1 versus Q4.

如果我們沒有增加任何一個新廣告商，那麼根據定義，該類別在第一季相對於第四季可能會擴大。

But as a whole, we're guiding to expansion in the business even in Q1 against Q4.

但總體而言，我們預計第一季的業務將比第四季擴張。

And frankly, that's just because our business is growing really quickly right now.

坦白說，這只是因為我們的業務現在成長非常快。

And we're guiding that way implying that I think it was a 4% quarter-over-quarter guide up with two less days.

我們這樣引導意味著我認為這是一個比上一季成長 4% 的指引，而且減少了兩天。

You have 2% less days because we're confident that just as a baseline level, regardless of seasonality or any bump that we had in Q4, you're going to have continued growth into Q1.

您的營業天數減少了 2%，因為我們相信，作為基準水平，無論季節性因素或第四季度遇到的任何波動如何，您的營業天數都會在第一季繼續增長。

Got it.

知道了。

And just one quick follow-up.

最後再跟進一下。

As you think about the cohort of e-commerce advertisers that have been engaging with you for a longer period of time versus some of your newer ones, what are you seeing from the willingness to spend into higher daily spend thresholds?

當您考慮那些與您的合作時間較長的電子商務廣告商群體和一些較新的廣告商群體時，您認為他們願意在更高的每日支出門檻上投入什麼？

I mean, it's volatile only because we got into the category right into Q4.

我的意思是，它之所以不穩定只是因為我們在第四季就進入了這個類別。

So a lot of these e-commerce companies go bonkers going into Black Friday and going into the holidays.

因此，許多電子商務公司在「黑色星期五」和假期來臨之際都會瘋狂購物。

And then just we're even learning the business model right now.

我們現在正在學習商業模式。

As you go into Christmas time, they're sort of out of inventory, and they got to go replenish inventory.

進入聖誕節時，他們的庫存已經售罄，所以他們必須去補充庫存。

And so they'll spike, they'll drop, they'll come back.

因此，它們會飆升、會下降、會回升。

But ultimately, we're a performance marketing business.

但歸根究底，我們是一家績效行銷公司。

And so as they get inventory replenished and they get into Q1, they ramp back up.

因此，當他們補充庫存並進入第一季時，他們就會再次增加產量。

And so we see volatility just due to the nature of the business. but they see great result on the platform.

因此，我們看到的波動僅是由於業務性質所致。但他們在該平台上看到了巨大的成果。

So that means over time, as you do normalize for that seasonality, there'll just be consistent lines of revenue for us that should be growing.

所以這意味著隨著時間的推移，當你將季節性正常化時，我們的收入就會穩定成長。

Not only should be growing because of their success on our platform, helping their businesses become larger, in which case, they'll reinvest more into marketing, but also because we're really early on this.

他們不僅應該因為我們平台上的成功而成長，幫助他們的業務變得更大，在這種情況下，他們會在行銷上投入更多，而且因為我們在這方面才剛起步。

It's only been a matter of few months so far

到目前為止才過了幾個月

we're still in one of the earliest versions of our model.

我們仍處於模型的最早版本之一。

That model in this category is a couple of years behind on the evolution of a model to the gaming model.

這一類別中的模型在向遊戲模型的演變上落後了幾年。

So we've got a long runway of a long road map of things to do to improve the gaming side and the gaming models are years ahead of now the e-commerce model.

因此，我們為改善遊戲方面製定了長期的路線圖，並且遊戲模式比現在的電子商務模式領先數年。

And so we're playing catch-up to the efficacy of that gaming model.

因此，我們正在努力提高該遊戲模型的有效性。

So our platform is only going to get better and better as we go on and go forward from here.

因此，隨著我們不斷前進，我們的平台只會變得越來越好。

Omar Dessouky, BofA.

美國銀行的奧馬爾·德蘇基（Omar Dessouky）。

Great.

偉大的。

Okay.

好的。

So I'm intrigued by this notion that your technology works for many different types of advertisers.

因此，我對你們的技術適用於多種不同類型的廣告商這個想法很感興趣。

It seems that everyone is calling the nongaming e-commerce, okay?

好像大家都把非遊戲類的叫電商好嗎？

And within e-commerce, there are different types of, I guess, categories, some products, some subscriptions, things like that.

我猜，在電子商務中，存在著不同類型的類別，一些產品，一些訂閱，諸如此類。

You put out the AXON pixel.

您放出了 AXON 像素。

So e-commerce merchants have that on their website now to help you attribute.

因此，電子商務商家現在可以在其網站上提供此功能，以幫助您歸因。

Is that enough to access all of these different types of e-commerce advertisers?

這足以接觸到所有這些不同類型的電子商務廣告商嗎？

That's the first question.

這是第一個問題。

So really, the question is, is your technology -- is all of your technology deployed at least on the attribution side, to access that 10 million advertisers in the world, okay?

所以真正的問題是，你的技術——你的所有技術是否至少在歸因方面部署，以接觸世界上的 1000 萬廣告商，好嗎？

That's the first question.

這是第一個問題。

And then given that your AI has shown such efficacy, is it now going to start working for in-app install for nongaming categories.

鑑於你的人工智慧已經表現出如此的功效，它現在是否會開始為非遊戲類別的應用程式內安裝工作？

This would be like nongaming apps.

這就像非遊戲應用程式一樣。

Is that now a growth segment in '25 because I think to date, that's been a very, very small part of your business, and I'm wondering if that's about to inflect as well.

這是否會成為 25 年的一個成長領域，因為我認為到目前為止，這只是您業務中非常非常小的一部分，而且我想知道這是否也即將發生變化。

Yes.

是的。

So great question.

這個問題問得真好。

So the first one, I think you said is that have we deployed all of our technology for attribution and being able to bring on customers on the platform.

所以第一個，我想你說的是，我們是否已經部署了所有的技術來進行歸因，並且能夠在平台上吸引客戶。

We certainly have a pixel integration that's going to be consistent for any website to be able to advertise on our platform.

我們當然有一個像素集成，它將與任何能夠在我們的平台上做廣告的網站保持一致。

But I wouldn't say that in any aspect of this business, that we're at maturity and all the technology that we can release, you can always do more to get to more attribution that's accurate with advertisers.

但我不會說在這個業務的任何方面我們都已經成熟，而且我們可以發布的所有技術，你總是可以做更多的事情來獲得對廣告商更準確的歸因。

And so we're working on continuously adding tools to that.

因此，我們正在致力於不斷添加工具。

So I will say, though, it obviously works.

所以我要說的是，它顯然有效。

It works in a category of websites and e-commerce.

它適用於網站和電子商務類別。

We're seeing it work in broader than that other categories of non-gaming on the web.

我們發現它的作用範圍比網路上其他非遊戲類別的更廣泛。

And so any type of customer, once we're self-service, we'll be able to come into the platform and advertise so long as they have a website.

因此，任何類型的客戶，一旦我們實現自助服務，只要他們有網站，我們就可以進入平台投放廣告。

Then you've got other types of advertisers.

然後你就有了其他類型的廣告商。

You've got game advertisers, and you've got non-gaming app advertisers, and then you've got websites who will also have an app.

你有遊戲廣告商，有非遊戲應用程式廣告商，還有擁有應用程式的網站。

Those last two, we have to do work to bring online, but it's work that we'll be doing throughout this year.

最後兩個，我們必須做一些工作才能上線，但這項工作我們將在今年全年進行。

So it's not something where I think you said will inflect in '25, it's not something that is a '25 thing.

所以我認為這不是您所說的會在 25 年發生變化的事情，這不是 25 年的事情。

It's just something that's in our road map that we're actively working on.

這只是我們正在積極努力的路線圖中的一部分。

And we do expect that then opens up even new categories as we go into later through this year and then next year and beyond.

我們確實預計，隨著今年稍後、明年及以後的發展，將會開闢出新的類別。

Martin Yang, Opp Co.

馬丁楊（Martin Yang），Opp 公司

First question is for Adam.

第一個問題是問亞當的。

Since you took oversight of the HR function last quarter, what are the changes you can talk about on recruiting, streamlining the operation, anything worth highlighting to us.

自從上個季度您接手人力資源職能以來，您可以談談在招募、精簡營運方面有哪些變化，有什麼值得向我們強調的嗎？

Look, I mean my focus has been on streamlining the team and processes and getting really focused.

看，我的意思是我的重點一直是簡化團隊和流程，並真正集中註意力。

And we did announce that app sales that we're working on as well in our talk script.

我們確實在演講稿中宣布了我們正在進行的應用程式銷售。

And really, this is all about just we see a lot of opportunity in front of us.

事實上，這只是因為我們看到眼前有很多機會。

We have a long road map, a lot of things to do.

我們有一個漫長的路線圖，有很多事情要做。

And when you have that organic opportunity, you got to get everyone focused on it.

當你有這樣的有機機會時，你必須讓每個人都注意它。

And so has really been around streamlining.

所以確實一直圍繞著精簡。

I mean we're one of the most financially lucrative businesses to be constantly announcing layoffs.

我的意思是，我們是最賺錢的企業之一，但卻不斷宣布裁員。

Now we don't want to constantly be doing that.

現在我們不想經常這樣做。

But since I took over HR, my job was to go, where are areas of the business that are not perfectly aligned with these organic opportunities, and let's start shedding those so that we can really narrow in and focus on what's in front of us.

但自從我接手人力資源工作以來，我的工作就是找出業務中哪些領域與這些有機機會不完全一致，然後開始擺脫這些領域，這樣我們才能真正縮小範圍，專注於我們面前的事物。

And so we're really getting to that place.

所以，我們確實到達了那個地方。

And once we get there, I think it's going to be even more exciting times here.

一旦我們到達那裡，我想這裡將會是更加令人興奮的時刻。

My another question is about -- you talked about nongaming helping with the supply.

我的另一個問題是——您談到了非博彩業對供應的幫助。

So do you expect market share gain on mediation and how will the additional supply over time help the gaming business?

那麼，您是否預期中介的市場份額會增加，以及隨著時間的推移，額外的供應將如何幫助遊戲業務？

Look, our market share is really high, okay?

你看，我們的市佔率真的很高，好嗎？

So let's -- setting aside market share, let's just talk about sheer inventory.

因此，我們先撇開市場份額，只討論純庫存。

We're getting new game publishers all the time that run game ads.

我們不斷有新的遊戲發行商發布遊戲廣告。

We're the best at monetizing it at this point.

目前，我們在將其貨幣化方面做得最好。

And so that aspect of the business growth in gaming supply is there.

遊戲供應方面的業務成長就在那裡。

Now that's just sort of organic and natural, but it's not some step function opportunity.

現在這只是一種有機和自然的事情，但它並不是某種階梯式功能機會。

What I was referring to is, if you look at the in-app purchasing apps, most of them where a lot of consumer time is spend on gaming don't actually run ads.

我指的是，如果你看一下應用程式內購買應用程序，你會發現其中大多數應用程式並沒有真正投放廣告，因為消費者在玩遊戲時花費了大量的時間。

They don't run ads because if they can monetize a user at really effective rates with not in-app purchasing, why would they run adds to their competition.

他們不投放廣告，因為如果他們能夠不透過應用程式內購買以真正有效的速度從用戶身上賺錢，那他們為什麼還要投放廣告來與​​競爭對手競爭呢？

And King was a perfect example of this, and they used to report underneath Activision when they were public pre-acquisition they launched ads with nongaming advertisers and created a lot of incremental revenue doing so.

King 就是一個很好的例子，他們曾在 Activision 上市前報告稱，他們向非遊戲廣告商推出了廣告，並因此創造了大量增量收入。

Traditionally in the mobile gaming mediation market, vast, vast majority of the ads were for games that were competitive to the publisher game.

傳統上，在行動遊戲中介市場中，絕大多數廣告都是針對與發行商遊戲具有競爭力的遊戲。

As we execute on bringing on more and more advertisers in nongaming categories, we think that it's going to be pretty reasonable to sell the MAX product and our DSP platform to these gaming publishers who are monetizing predominantly or exclusively with in-app purchasing, bring on that supply and that will create another expansion vector.

隨著我們在非遊戲類別中吸引越來越多的廣告商，我們認為將 MAX 產品和我們的 DSP 平台出售給那些主要或完全透過應用程式內購買盈利的遊戲發行商是相當合理的，帶來這種供應並創造另一個擴展向量。

Alec Brondolo, Wells Fargo.

富國銀行的亞歷克‧布朗多洛 (Alec Brondolo)。

I think you're describing a situation where the business is rapidly transitioning from demand constrained and supply constrained as you kind of expand the aperture of categories that you service.

我認為您所描述的情況是，隨著您服務的類別範圍的擴大，業務正在從需求受限轉變為供應受限。

I guess acknowledging that some games could increase their (inaudible) and that could increase available supply, how do you think about kind of the amount of spend growth, the current supply base that kind of the mobile game environment or ecosystem has today could absorb, like could the inventory be twice as valuable, 3x as valuable.

我想承認有些遊戲可以增加其（聽不清楚），這可能會增加可用的供應量，您如何看待支出增長的量，當前的移動遊戲環境或生態系統的供應基礎可以吸收多少，例如庫存價值是否可以翻倍，三倍。

Do you think that's kind of the constraint on growth going forward?

您認為這會成為未來成長的限制因素嗎？

Yes, I don't think we're saying that's a constraint.

是的，我不認為這是一種限制。

I mean, in the last year, supply has been pretty constant, and I think we grew 80%, 90%.

我的意思是，在過去的一年裡，供應量一直相當穩定，我認為我們的成長了 80% 到 90%。

We're pretty early on both sides.

雙方都還很早。

On the demand side, we're just getting into areas that are going to help us much better monetize our inventory.

在需求方面，我們正在進入一些能夠幫助我們更好地將庫存貨幣化的領域。

And we've given out examples in terms of just ARPDAU that some of the largest social properties are able to monetize that being significantly higher than the ARPDAU that we monetize this audience in today.

我們已經給出了一些例子，僅從 ARPDAU 來看，一些最大的社交資產能夠透過其貨幣化獲得的收入遠高於我們今天透過該受眾貨幣化的 ARPDAU。

And yet our ad unit is a full-screen video.

但我們的廣告單元是全螢幕影片。

It requires some amount of engagement and viewing with the video, and it's really high impact.

它需要一定程度的參與和觀看視頻，而且影響力確實很大。

The only reason we're not better monetizing today where we would say maybe we're supply constrained is because we don't have every type of advertiser across the world on our platform yet.

我們目前之所以無法更好地實現盈利，唯一的原因是我們可能受到了供應的限制，因為我們的平台還沒有覆蓋世界各地的所有類型的廣告商。

Over the coming quarters, years, decade plus, as we onboard more and more advertisers and we're talking in the millions plus, that's going to allow us to have a lot more demand density.

在未來的幾個季度、幾年、甚至十年內，隨著我們吸收越來越多的廣告商，數量將超過數百萬，我們將擁有更大的需求密度。

Our algorithms are only going to get stronger.

我們的演算法只會變得更強大。

Our data footprint is only going to get stronger.

我們的數據足跡只會越來越強大。

And forget supply increase, same exact supply, 1 billion-plus daily actives are just going to be better monetized.

忘記供應量的增加吧，同樣的供應量，每天超過 10 億的活躍用戶將會更好地實現貨幣化。

There's a long runway of growth there.

那裡有很長的成長空間。

The supply side is just gravy on top.

供給方只不過是錦上添花而已。

It's really just an extra area that we think is just going to be given to us for free that's going to come.

這實際上只是一個額外的區域，我們認為它將免費提供給我們。

We certainly don't think it's going to be anywhere near as impactful as that demand expansion that's going to drive our growth.

我們當然不認為它會產生像推動我們成長的需求擴張那樣大的影響。

Arsenije Matovic, Wolfe.

阿塞尼耶·馬托維奇，沃爾夫。

What is the thought process you guys have on bringing [Audience Plus] out of pilots and into self-serve, I guess, what have you learned with the pilots that will help you improve or make changes to the solution before making itself serve?

你們在將 [Audience Plus] 從試點階段推進到自助服務階段時，是怎樣的想法？

And I guess what's a good point to look at in 2025 on when we should expect it to go self-service?

我想，在 2025 年，我們預期什麼時候可以實現自助服務？

And then just one brief follow-up.

然後只需進行一次簡短的跟進。

I mean we'll launch self-service.

我的意思是我們將推出自助服務。

I think when we're ready for it.

我想當我們做好準備的時候。

The most important keys there are: one, we're going to have tools to automate every step of the process, including AI agents that are helping the advertisers get on board and get going.

其中最重要的關鍵是：第一，我們將擁有工具來自動化流程的每個步驟，包括幫助廣告商加入並開始行動的人工智慧代理。

They really help them through the process so that they feel like they're engaging with human beings, but they're engaging with bots and they're getting going.

他們確實在整個過程中為他們提供幫助，讓他們感覺自己正在與人類交往，但他們正在與機器人交往，他們正在開始行動。

The other key piece is you have to build a lot of tools and content moderation controls when you open up a platform as big as ours.

另一個關鍵點是，當你開放一個像我們一樣大的平台時，你必須建立大量的工具和內容審核控制。

And at this point, I mean, in the advertising realm, we're -- you've got Facebook, you've got Google, we're now the next really big platform out there.

而目前，我的意思是，在廣告領域，我們有 Facebook，有 Google，我們是下一個真正巨大的平台。

We opened that up.

我們將其打開了。

And we would get a lot of fraud that came on board if we didn't have a lot of tools written to prevent that.

如果我們沒有製定大量的工具來預防詐欺行為，我們就會遭遇大量詐欺行為。

So we're going to take it really seriously.

所以我們會非常認真地對待此事。

We'll be conservative.

我們會採取保守態度。

I don't know that we have any need to accelerate growth right now.

我不知道我們現在是否有必要加速成長。

We're growing really, really quickly as it is in controlling the pace, but we're developing all these tools in the background.

我們在控制步伐方面發展得非常非常快，但我們在後台開發所有這些工具。

And once we're ready, we'll bring it out.

一旦我們準備好了，我們就會把它拿出來。

We'll let you all know, and we think it will be very impactful, not just for a moment in time, but that will compound over quarters, years and decade plus.

我們會讓大家知道，我們認為它的影響將非常大，不僅是一時，而且會在未來幾季、幾年甚至十年內累積起來。

Thanks for providing the segmented guidance.

謝謝您提供分段指導。

I just wanted to understand with the App sales, is it going to be tranched out where you're going to roll off a few studios over time throughout the year?

我只是想了解應用程式銷售情況，它是否會分階段進行，你們會在全年內陸續推出幾個工作室嗎？

Or is it once it closes, all 10 studios are off, no more apps revenue, and that's kind of the way the mechanism will be, and that would happen when would that be at the end of the quarter or trying to understand the timing dynamic for Apps revenue modeling for this year.

或者一旦關閉，所有 10 個工作室都會關閉，不再有應用程式收入，這就是機制的方式，這將在本季度末發生，或者試圖了解今年應用程式收入模型的時間動態。

So we're going to be selling the entirety of the Apps business, Arsenije.

所以我們將出售整個應用程式業務，Arsenije。

So that would all come off of the P&L and the balance sheet all at once.

這樣，這些就會同時從損益表和資產負債表中剔除。

And then timing, as I mentioned, I think, in my talk back that we're targeting for that to close within Q2, obviously, we may be subject to regulatory approval.

然後是時間安排，正如我之前提到的，我想，我們的目標是在第二季度內完成，顯然，我們可能需要獲得監管部門的批准。

So that timing may change slightly, but that's kind of where we're targeting is within Q2.

因此時間可能會略有變化，但這正是我們在第二季的目標。

Eric Sheridan, Goldman Sachs.

高盛的艾瑞克‧謝裡丹 (Eric Sheridan)。

I wanted to go back to one of the five initiatives you talked about for 2025.

我想回到您談到的 2025 年五項舉措之一。

Personalizing ad experiences.

個性化廣告體驗。

How much of that initiative can be done with what's already been built today versus incremental investments you need to make against the goals of personalization over the medium to long term.

相對於為實現中長期個人化目標而需要進行的增量投資，有多少措施可以利用今天已經建構的成果來實現。

And from the outside looking in, what should we be watching in terms of how you might implement that and how it might be rolled out in the business more broadly in the years ahead?

從外部來看，我們應該專注於如何實現這一目標，以及未來幾年如何在業務中更廣泛地推廣這一目標？

Look, this is all research and development right now.

看，現在這都是研究和開發。

And just to break down what it means.

讓我們來分析一下它的意義。

Today, if you've got a game, let's use Candy Crush, again, as an example. a human being is uploading 20 ads the end user that sees ads for the Candy Crush campaign run through us sees one of those 20 ads.

今天，如果你有一款遊戲，我們又以 Candy Crush 為例。一個人上傳了 20 個廣告，最終用戶在看到我們進行的 Candy Crush 活動廣告時會看到這 20 個廣告中的一個。

In the middle, the AI will determine which one of those 20 to place.

在中間，人工智慧將決定放置這20個中的哪一個。

Now that's not very personalized.

這不太個性化。

It's very static and the users are seeing the exact same sort of format of ad and the exact same type of video as all the other users who are seeing that campaign.

它非常靜態，用戶看到的廣告格式和影片類型與看到活動的其他用戶看到的完全相同。

Now with generative AI and these large language models, you can imagine you can run those top creatives through a model and create many variations.

現在有了生成式人工智慧和這些大型語言模型，你可以想像你可以透過模型運行那些頂尖的創意並創造出許多變化。

Facebook and Google are talking about this, too, generative AI-enabled ad.

Facebook 和 Google 也在談論這個生成式 AI 廣告。

It's something that we're working on, we think it will be very impactful in driving consumer response to the ad unit.

這是我們正在努力的事情，我們認為它將對推動消費者對廣告單元的反應非常有效。

It's trickier than a text ad because we've got a full-screen video experience plus a lot of times, we've got HTML in the gaming realm, we've got even a mini game inside the HTML.

它比文字廣告更棘手，因為我們擁有全螢幕視訊體驗，而且很多時候，我們在遊戲領域擁有 HTML，我們甚至在 HTML 裡面有一個迷你遊戲。

So there's a lot of work to do to personalize that experience.

因此，為了使這種體驗個性化，還有很多工作要做。

And then there's a lot of work to do to build a model that can do it, do it automatically and then be able to process it within the constraint of the GPUs that we have.

然後還有很多工作要做來建立一個可以做到這一點的模型，可以自動完成，然後能夠在我們現有的 GPU 約束內處理它。

But it's something that we're working on and we think will be impactful when it rolls out, and it's something that's in focus for us this year.

但這是我們正在努力的事情，我們認為當它推出時將會產生影響，這也是我們今年的重點。

Bernie McTernan, Needham & Company.

伯尼‧麥克特南 (Bernie McTernan)，尼德漢姆公司。

Great.

偉大的。

Maybe just a follow-up on Eric's question on personalization.

也許只是對 Eric 關於個人化的問題的後續回答。

From some of the checks that we've done, it seems that e-commerce brands are largely using social ads within mobile gaming.

根據我們進行的一些檢查，電子商務品牌似乎在手機遊戲中大量使用社群廣告。

Is that an opportunity for you guys that you can help actually make games or ads that look more like mobile gaming ads for e-commerce and other verticals.

這對你們來說是一個機會嗎？

Yes, possibly.

是的，有可能。

I mean, look, we hope not to be the ones deciding that.

我的意思是，看，我們希望不要成為決定這一點的人。

We hope the AI and the machines are the ones deciding that.

我們希望人工智慧和機器能夠決定這一點。

And the beauty is once you're in a machine-created realm, you're going to be able to create way more variations than the human being can.

美妙之處在於，一旦進入機器創造的領域，你將能夠創造出比人類更多的變化。

So whether it's an e-commerce product somehow gamified in the advertisement or something that we can't imagine right now, in theory, the machine is going to be able to do that, and the advertiser and the consumer will gain the benefit of all those new ad variations that will be output out of our system.

因此，無論是以某種方式在廣告中遊戲化的電子商務產品，還是我們目前無法想像的東西，理論上，機器都能夠做到這一點，廣告商和消費者都將從我們的系統輸出的所有新廣告變體中受益。

Got it.

知道了。

And then just one last one for me.

對我來說，只剩下最後一個了。

Any feedback that you're hearing from the gaming companies as you're moving into e-commerce and other verticals?

當您進軍電子商務和其他垂直行業時，您聽到來自遊戲公司的任何回饋嗎？

I mean look, the gaming companies are seeing better performance on our platform every single quarter.

我的意思是，遊戲公司在我們的平台上每季的表現都越來越好。

So they're obviously happy.

所以他們顯然很高興。

We're a catalyst for their growth and really a requirement for their growth at this point.

我們是他們成長的催化劑，也是他們目前成長的必需品。

The publishers themselves, there's been some social commentary on this.

出版商本身對此也發表了一些社會評論。

When they start seeing more and more of their impressions shifting to e-commerce, they absolutely love it.

當他們看到越來越多的印象轉向電子商務時，他們非常高興。

If you're a game publisher, your worst nightmare would have been, I'm going to monetize my game with all of my competition ads.

如果你是遊戲發行商，你最可怕的惡夢就是，我要利用我所有的競爭對手的廣告來將我的遊戲貨幣化。

That just sucks as an end product, but that's all they had.

最終產品實在太差了，但這就是他們所擁有的一切。

And so we allow all of our MAX publishers to see the advertising in this run.

因此，我們允許所有 MAX 出版商看到本次發布的廣告。

Some of those publishers have commented about the number of impressions that are shifting to nongaming categories.

部分出版商對轉向非遊戲類別的展示次數進行了評論。

When you see that commentary, they're absolutely excited about it.

當你看到那條評論時，他們肯定非常興奮。

And again, this concludes the question-and-answer session and the webinar for this quarter.

再次，本季的問答環節和網路研討會到此結束。

We thank you all for joining us and look forward to seeing you again next quarter.

我們感謝大家的加入我們，並期待下個季度再次見到你們。

Take care until then.

在此之前請多保重。

Thanks, everyone.

謝謝大家。

Thank you.

謝謝。