Roblox Corp (RBLX) 2022 Q1 法說會逐字稿

Good morning, ladies and gentlemen. My name is Abby, and I will be your conference operator today. At this time, I would like to welcome everyone to the Roblox First Quarter 2022 Earnings Conference Call. Today's conference is being recorded. (Operator Instructions) Thank you.

早上好，女士們，先生們。我的名字是艾比，今天我將成為您的會議接線員。在這個時候，我想歡迎大家參加 Roblox 2022 年第一季度收益電話會議。今天的會議正在錄製中。 （操作員說明）謝謝。

And at this time, I would like to turn the conference over to Stefanie Notaney, Director of Financial Communications. Ms. Notaney, you may begin your conference.

在這個時候，我想把會議交給金融傳播總監 Stefanie Notaney。 Notaney 女士，你可以開始你的會議了。

Thank you, Abby. Good morning, everyone, and thank you for joining our Q&A session to discuss Roblox's Q1 2022 results. With me today is Roblox's CEO, David Baszucki; and CFO, Mike Guthrie.

謝謝你，艾比。大家早上好，感謝您參加我們的問答環節，討論 Roblox 2022 年第一季度的業績。今天和我在一起的是 Roblox 的首席執行官 David Baszucki；和首席財務官邁克·格思裡。

Before we start, I want to remind everyone that yesterday, after market close, we published a shareholder letter and earnings results on our Investor Relations website at ir.roblox.com. On this call, we will make some brief opening remarks and reserve the rest of the time for your questions. For our webcast participants, please note the question icon at the bottom of your screen where you can type in your questions. We'll do our best to take as many questions as possible in the time we have allotted today.

在開始之前，我想提醒大家，昨天收市後，我們在投資者關係網站 ir.roblox.com 上發布了股東信函和收益結果。在本次電話會議上，我們將做一些簡短的開場白，其餘時間留給您提問。對於我們的網絡直播參與者，請注意屏幕底部的問題圖標，您可以在其中輸入問題。我們將盡最大努力在今天分配的時間內回答盡可能多的問題。

On today's call, we may be making forward-looking statements, including, but not limited to, our expectations of our business, future financial results and strategy. Forward-looking statements are subject to risks and uncertainties that could cause actual results to differ materially from those described in our forward-looking statements, and such risks are described in our risk factors included in our SEC filings, including our Form 10-K filed for the fiscal year ended December 31, 2021. You should not rely on our forward-looking statements as predictions of future events. We disclaim any obligation to update forward-looking statements, except as required by law.

在今天的電話會議上，我們可能會做出前瞻性陳述，包括但不限於我們對業務、未來財務業績和戰略的預期。前瞻性陳述受風險和不確定性的影響，可能導致實際結果與我們的前瞻性陳述中描述的結果存在重大差異，此類風險在我們提交給美國證券交易委員會的文件中描述的風險因素，包括我們提交的 10-K 表格截至 2021 年 12 月 31 日的財政年度。您不應依賴我們的前瞻性陳述作為對未來事件的預測。除法律要求外，我們不承擔更新前瞻性陳述的任何義務。

During this call, we will also discuss certain non-GAAP financial measures. Reconciliations between GAAP and non-GAAP metrics for our reported results can be found in our press release issued yesterday as well as in our supplemental slides, copies of which can be found on our Investor Relations website.

在本次電話會議中，我們還將討論某些非公認會計原則的財務措施。我們報告的結果的 GAAP 和非 GAAP 指標之間的調節可以在我們昨天發布的新聞稿以及我們的補充幻燈片中找到，其副本可以在我們的投資者關係網站上找到。

Finally, this call is being webcast. And as a reminder, for those participants, you can enter your questions in the top right side of your screen. The webcast will be archived on our website shortly afterwards.

最後，本次電話會議正在進行網絡直播。作為提醒，對於這些參與者，您可以在屏幕的右上角輸入您的問題。網絡廣播將在不久之後在我們的網站上存檔。

With that, I'll turn the call over to Dave.

有了這個，我會把電話轉給戴夫。

Welcome, everyone. We welcome all of our investors in the Roblox community. I'm going to share a few notes before we dive in on the Q&A, especially for those who maybe didn't deeply read our shareholder letter.

歡迎大家。我們歡迎 Roblox 社區的所有投資者。在我們深入問答之前，我將分享一些筆記，特別是對於那些可能沒有深入閱讀我們的股東信的人。

Reiterating on our growth. Daily active users were at 54 million for Q1, which was up 28% year-over-year, and our highest ever. And our hours of engagement were up 22%. Bookings were at 631 million for Q1, which was 3% below Q1 of 2021. And I want to highlight, we generated over $150 million of net cash and $100 million of free cash in Q1.

重申我們的成長。第一季度的每日活躍用戶為 5400 萬，同比增長 28%，是我們有史以來的最高水平。我們的參與時間增加了 22%。第一季度的預訂量為 6.31 億美元，比 2021 年第一季度低 3%。我想強調的是，我們在第一季度產生了超過 1.5 億美元的淨現金和 1 億美元的自由現金。

March, we believe, was our most difficult month lapping COVID. And what is really exciting is all of the user gains generally we accrued during COVID, we've kept. That said, our -- as expected, really, our bookings are highly correlated with hours of engagement, and our hours have decreased in some cohorts as we've emerged from COVID.

我們相信，三月是我們應對 COVID 最困難的一個月。真正令人興奮的是，我們在 COVID 期間通常獲得的所有用戶收益，我們一直保持。也就是說，我們的 - 正如預期的那樣，實際上，我們的預訂與參與時間高度相關，並且隨著我們從 COVID 中出現，我們在某些群體中的工作時間有所減少。

I'll give an example, which is the USA 9 through 12 cohort, which is an extremely dynamic part of our numbers. And pre-COVID, we had 2 million DAUs in the 9 through 12 cohort spending less than 5 million hours per week. In the middle of COVID, that jumped to 3 million DAUs in that same 9 through 12 cohort spending 10 million hours per week. We emerged from COVID with that same 3 million DAUs, but now spending 8 million hours per week. And once again, this is highly correlated bookings and our hours of engagement. It partially explains our year-on-year bookings number.

我將舉一個例子，這是美國 9 到 12 歲的隊列，這是我們數據中極具活力的部分。在 COVID 之前，我們在 9 到 12 歲的隊列中有 200 萬個 DAU，每週花費的時間不到 500 萬小時。在 COVID 期間，在同樣的 9 到 12 歲隊列中，這一數字躍升至 300 萬個 DAU，每週花費 1000 萬小時。我們以同樣的 300 萬 DAU 從 COVID 中脫穎而出，但現在每週花費 800 萬小時。再一次，這是高度相關的預訂和我們的參與時間。它部分解釋了我們的同比預訂量。

For those of you that read our GAAP results, you'll notice that we're going to move to accruing revenue with an average user lifetime moving from 23 to 25 months. This will essentially increase the time we -- we near out our revenue. But I want to highlight, it could be interpreted as a sign of increased user retention, which is very positive for us.

對於那些閱讀了我們的 GAAP 結果的人，您會注意到我們將轉向累積收入，平均用戶生命週期從 23 個月變為 25 個月。這將從本質上增加我們接近收入的時間。但我想強調，這可以解釋為用戶留存率增加的標誌，這對我們來說是非常積極的。

We have enormous headroom even in the U.S. 9 through 12 cohort because we have a lot of room on our frequency there. We do not share our MAU numbers right now, but there's a lot of room there. I also want to highlight that in our 17 through 24 cohorts in the core markets, while lapping COVID, we are still seeing solid DAU growth numbers year-on-year. And in Q1, in core, we saw a 6% year-on-year.

即使在美國 9 到 12 歲的人群中，我們也有巨大的空間，因為我們在那裡的頻率有很大的空間。我們現在不分享我們的 MAU 數字，但那裡有很多空間。我還想強調，在我們在核心市場的 17 到 24 個隊列中，儘管與 COVID 相比，我們仍然看到 DAU 同比增長穩健。在第一季度，核心業務同比增長 6%。

Internationally, I want to highlight 2 countries. The first is India, which has enormous potential user activity. We saw in Q1 of '22, India grew 160% relative to Q1 of 2021 at under 1 million DAUs with a lot of headroom there. And then Japan, which is a potential for enormous economic activity. In Q1 of '22, we saw growth of 3x relative to Q1 of '21, with 183,000 DAUs. So that market is just in formation.

在國際上，我想強調兩個國家。首先是印度，它擁有巨大的潛在用戶活動。我們在 22 年第一季度看到，與 2021 年第一季度相比，印度增長了 160%，日活躍用戶數低於 100 萬，而且還有很大的空間。然後是具有巨大經濟活動潛力的日本。在 22 年第一季度，我們看到相對於 21 年第一季度增長了 3 倍，DAU 為 183,000。所以這個市場剛剛形成。

We continue to drive innovation up and down our stack. And a couple of innovations that I'll share with you that have emerged in Q1 that we talked about includes spatial voice, our layered clothing and fashion system, our age verification system that is working hand-in-hand with spatial voice.

我們繼續推動我們的堆棧上下創新。我將與您分享的一些創新在第一季度出現，我們談到了包括空間語音、我們的分層服裝和時尚系統、我們與空間語音攜手合作的年齡驗證系統。

We've introduced Roblox cloud and the ability for developers to access really the whole back end of their experience through their own APIs, if they so choose. We have migrated to our own internal Roblox translate system that has shown better long-term engagement growth than other translate providers we've used from the cloud.

我們引入了 Roblox 雲，讓開發人員能夠通過自己的 API 訪問他們體驗的整個後端，如果他們願意的話。我們已遷移到我們自己的內部 Roblox 翻譯系統，該系統顯示出比我們從雲中使用的其他翻譯提供商更好的長期參與度增長。

We've added a data center in India, which is part of the very rapid growth there, and has decreased latency by 50% in India. And we're in our first steps of our facial animation data in Roblox Studio now. We continue to believe that part of the long-term growth for our company is continuous innovation. And in addition to these, we have many more in the pipeline.

我們在印度增加了一個數據中心，這是該地區快速增長的一部分，並將印度的延遲降低了 50%。我們現在正處於 Roblox Studio 中面部動畫數據的第一步。我們仍然相信，我們公司長期增長的一部分是持續創新。除了這些，我們還有更多正在籌備中。

We have shared in our shareholder letter the enormous opportunity for us to lean in on the efficient frontier on the economy on Roblox. And we are taking steps up and down our economy stack to nudge towards optimizing long-term retention and engagement with the revenue flowing through our system and the ability for our developers to build larger and larger teams as they recognize more economic activity.

我們在股東信中分享了我們利用 Roblox 經濟有效前沿的巨大機會。我們正在逐步調整我們的經濟體系，以推動優化長期保留和參與，以利用流經我們系統的收入以及我們的開發人員在認識到更多經濟活動時建立越來越大的團隊的能力。

A few I'll highlight that are underway at Roblox include both on-game experience discovery and marketplace discovery. We're well into projects there to balance long-term engagement and retention with the monetization of the various experiences. We're adding validated accounts and including the ability for people to trust that the items they're buying are from the real Nike or the real Vans. We are in the midst of shortly releasing the ability for brands and developers to boost their experiences through discovery on our homepage.

我將重點介紹 Roblox 正在進行的一些活動，包括遊戲體驗發現和市場發現。我們在那裡開展項目，以平衡長期參與和保留與各種體驗的貨幣化。我們正在添加經過驗證的帳戶，並包括讓人們相信他們購買的商品來自真正的 Nike 或真正的 Vans 的能力。我們將很快發布品牌和開發者通過我們主頁上的發現來提升他們的體驗的能力。

And we started work on our immersive advertising system, which is really, long term, the vision we've had for really a parallel economy that supports brands, bringing traffic to their experiences. I want to highlight, anything we do with advertising will be consistent with our values, will be consistent with the ages on our platform, will be consistent with all laws and regulations around advertising. But there is enormous economic opportunity here.

我們開始研究我們的沉浸式廣告系統，從長遠來看，這是我們對真正支持品牌的平行經濟的願景，為他們的體驗帶來流量。我想強調，我們對廣告所做的任何事情都將符合我們的價值觀，符合我們平台上的年齡，符合有關廣告的所有法律法規。但這裡有巨大的經濟機會。

And then just highlighting, before we jump into the Q&A, and you'll have to excuse my voice as I go through this. We have a lot more great brand experiences, including the NFL. We launched NFL Tycoon, McLaren and Alo Yoga, American Eagle and Chipotle, who all launched experiences on our platform. And then on the music side, we hosted GRAMMY Week, we hosted the BRIT Awards, we hosted David Guetta, and with Sony, we did a 24kGoldn virtual concert.

然後只是強調一下，在我們進入問答環節之前，當我經歷這個時，你必須原諒我的聲音。我們有很多很棒的品牌體驗，包括 NFL。我們推出了 NFL Tycoon、McLaren 和 Alo Yoga、American Eagle 和 Chipotle，它們都在我們的平台上推出了體驗。然後在音樂方面，我們舉辦了格萊美週，我們舉辦了全英音樂獎，我們接待了 David Guetta，我們與索尼一起舉辦了 24kGoldn 虛擬音樂會。

As your CEO, I'm still -- just want to highlight, in closing, the bullishness we have within the company based on the ultimate size of this market. We have a very healthy amount of cash, even as we are in the midst of generating cash. We are unique in our developer community and the foundation of stability that we really rest our business on. And we have enormous network effects that drive our growth.

作為您的首席執行官，我仍然 - 最後只是想強調一下，基於這個市場的最終規模，我們在公司內部的樂觀情緒。即使我們正在產生現金，我們也有非常健康的現金。我們在我們的開發者社區中是獨一無二的，也是我們真正依靠的穩定基礎。我們擁有巨大的網絡效應來推動我們的增長。

So with that, thank you for joining us, and we'll move to Q&A.

因此，感謝您加入我們，我們將進入問答環節。

(Operator Instructions) And we will take our first question from Drew Crum with Stifel.

（操作員說明）我們將回答 Drew Crum 和 Stifel 的第一個問題。

So Dave, in your shareholder letter, you discussed innovations that are currently in beta testing, specifically custom materials and dynamic heads. What have you found in your testing in terms of how these initiatives have impacted your KPIs? And when do you expect these to go live? And then I have a follow-up.

所以 Dave，在你的股東信中，你討論了目前處於 beta 測試階段的創新，特別是定製材料和動態磁頭。就這些舉措如何影響您的 KPI 而言，您在測試中發現了什麼？您希望這些內容何時上線？然後我有一個跟進。

Yes. Custom materials, we're actually live with parts of that right now. It's a more subtle feature in that it affects the overall look and feel of our -- of everything on Roblox. One thing custom materials is going to allow is more diversity in the look and feel of experiences on Roblox because, right now, Roblox runs on a physically based rendering paradigm.

是的。定製材料，我們現在實際上已經使用其中的一部分。這是一個更微妙的功能，因為它會影響我們在 Roblox 上的所有內容的整體外觀和感覺。自定義材質將允許的一件事是在 Roblox 上體驗的外觀和感覺更加多樣化，因為目前，Roblox 運行在基於物理的渲染範例上。

So there'll be different types of brick and different types of grass, and developers will be able to customize that. I think one long-term Roblox developer, and I'm going to quote them for -- I'm just quoting the developer rather than making my own statement, said Roblox is doing God's work with custom materials. So I think our developer community is really accepting this.

所以會有不同類型的磚和不同類型的草，開發人員將能夠定制它們。我認為一位長期的 Roblox 開發人員，我將引用他們的話——我只是引用開發人員而不是發表我自己的聲明，他說 Roblox 正在使用定製材料完成上帝的工作。所以我認為我們的開發者社區真的接受了這一點。

On facial animation, we're launching the beta in Roblox Studio, and developers have access to that right now. We're not sharing a date on the longer-term launch. As you can imagine, we want that to be as soon as possible, to get that into every game so every avatar on Roblox is animated. So both of these will have more subtle long-term effects rather than direct KPI effects.

在面部動畫方面，我們正在 Roblox Studio 中推出測試版，開發者現在可以訪問它。我們沒有分享長期發布的日期。正如你可以想像的那樣，我們希望盡快實現這一點，將其融入到每款遊戲中，這樣 Roblox 上的每個頭像都是動畫的。因此，這兩者都會產生更微妙的長期影響，而不是直接的 KPI 影響。

Got it. Okay. Fair enough. And then, Mike, the April bookings were up high single digits month-to-month. Setting COVID aside, which would create some distortion in the numbers, can you remind us historically what the shape of 2Q bookings looks like and what you're anticipating as you progress through May and June this year?

知道了。好的。很公平。然後，邁克，4 月份的預訂量環比增長了個位數。撇開 COVID 不談，這會在數字上造成一些扭曲，您能否從歷史上提醒我們第二季度預訂的形式以及您在今年 5 月和 6 月進展時的預期？

Yes. Drew, I'll remind -- well, I'll comment on what the shape of the curve normally looks like. Normally, April is a very strong month for us. It was incredibly strong in '21. It was one of the last super high year-over-year growth rates as COVID last year were starting to wane and reopen. So we had expected year-over-year growth to bottom in April.

是的。德魯，我會提醒一下——好吧，我會評論一下曲線的形狀通常是什麼樣子的。通常，四月對我們來說是一個非常強勁的月份。它在 21 年非常強大。這是最後一個超高的同比增長率之一，因為去年的 COVID 開始減弱並重新開放。因此，我們曾預計 4 月份的同比增長將觸底。

Right now, it looks like it bottomed in March, which is good. So sequentially, our year-over-year growth rates in April were better than they were in March. And on an year-over-year basis, I expect that to be true in May and again in June.

現在，它看起來像是在三月份觸底，這很好。因此，我們 4 月份的同比增長率好於 3 月份。與去年同期相比，我預計 5 月和 6 月都會如此。

In terms of the overall shape of the curve, normally April is higher than May, and then June is higher than -- sorry, May is lower than April and then June is back up higher than May, and really, the opening of the summer season where normal seasonality starts to kick in. Last year, '21, we didn't see the normal seasonal curves because we were going from heavy COVID period early in the year, first 3 to 4 months, to much more of a reopen period, May, June, July through the summer.

就曲線的整體形狀而言，通常 4 月高於 5 月，然後 6 月高於 - 對不起，5 月低於 4 月，然後 6 月又高於 5 月，真的，夏天的開始正常季節性開始出現的季節。去年，'21，我們沒有看到正常的季節性曲線，因為我們正從年初的重度 COVID 時期，前 3 到 4 個月，到更多的重新開放時期、 5 月、 6 月、 7 月通過夏天。

This year, we're open. And so we expect more normal seasonality to take hold, meaning we do expect the summer to be strong vis-a-vis where we are in the spring. And so we'll see that in June, July, August. So we're glad to see the sequential improvement, but expect that to continue in May, June and into the third quarter.

今年，我們開張了。因此，我們預計會出現更正常的季節性，這意味著我們確實預計夏季會比春季強勁。所以我們將在六月、七月、八月看到這一點。因此，我們很高興看到連續改善，但預計這種情況將在 5 月、6 月和第三季度持續下去。

And we will take our next question from Bernie McTernan with Needham & Company.

我們將向 Needham & Company 的 Bernie McTernan 提出我們的下一個問題。

Maybe just to start, just high level, why is now the right time to be leaning into monetization and gives you the -- what gives you the confidence that consumers are willing to spend more on Roblox right now?

也許只是開始，只是高水平，為什麼現在是進行貨幣化的正確時機並給你 - 是什麼讓你相信消費者現在願意在 Roblox 上花費更多？

I'll answer this. And I want to highlight earlier what I said about values-consistent and efficient frontier. I'll give you a highlight of the opportunity we have, and I think this is more of an opportunity than consumers spending more money.

我會回答這個。我想在前面強調一下我所說的價值觀一致和有效的前沿。我將重點介紹我們擁有的機會，我認為這比消費者花更多的錢更像是一個機會。

We, right now, for example, in search and discovery, historically have made no real difference in how we feature experiences with the same engagement and the same long-term retention even if one of those experiences monetizes 10x, which the other one does. So I think this is a more of a consumer value of highlighting experiences that historically consumers have liked to spend Roblox in, rather than new kind of ways of trying to pull more money out of consumers.

例如，我們現在在搜索和發現方面，從歷史上看，我們在如何展示具有相同參與度和相同長期留存率的體驗方面並沒有真正的不同，即使其中一種體驗的貨幣化率是 10 倍，而另一種則是這樣。因此，我認為這更像是一種消費者價值，即突出消費者在歷史上喜歡使用 Roblox 的體驗，而不是試圖從消費者身上榨取更多錢的新方式。

I would say the same is true of our catalog experience, which there is a lot of room for optimization. One thing we're doing in our catalog that is -- I'll dive in a little bit more deeply on. We're moving our catalog to be 100% user-generated, and that includes clothing, bodies, heads and faces. But a second part of what we're doing in our catalog is enhancing the trading efficiency and really the dynamic nature of our catalog. So we have a range of free items and rare items.

我想說我們的目錄體驗也是如此，還有很大的優化空間。我們在目錄中正在做的一件事是——我將更深入地探討一下。我們正在將我們的目錄轉變為 100% 用戶生成的，其中包括服裝、身體、頭部和麵部。但是我們在目錄中所做的第二部分是提高交易效率和目錄的動態特性。所以我們有一系列免費物品和稀有物品。

This historically has been something Roblox has done very well with Roblox-related items, and we've been doing this for over 10 years. And the Roblox economy has gotten to the point where we have items that trade for USD 20,000 or 2 million Roblox as they become rare. We want to add that level of fun and dynamic nature to our catalog. This is actually something our community wants is more opportunity to really work in that kind of trading economy. So everything we do for our economy is consistent with engagement first. But it is, we believe, an opportunity to realize the efficient frontier.

這在歷史上一直是 Roblox 在與 Roblox 相關的物品方面做得非常好的事情，我們已經這樣做了 10 多年。 Roblox 經濟已經到了這樣的地步，我們的物品交易價格為 20,000 美元或 200 萬美元，因為它們變得稀有。我們希望在我們的目錄中添加這種級別的樂趣和動態性質。這實際上是我們社區想要的更多機會在那種貿易經濟中真正工作。因此，我們為經濟所做的一切都首先與參與一致。但我們相信，這是實現高效前沿的機會。

Understood. And then moving on to advertising, acknowledging that it's still early days in thinking through the opportunity. As you're thinking about what advertising is going to look like on Roblox, is it possible to have a performance marketing piece on advertising in addition to brand marketing? And then if there is performance, would that call to action only be based on experiences within the Roblox metaverse or goods as well? Or could it be -- translate to the real world as well?

明白了。然後繼續做廣告，承認現在考慮這個機會還為時過早。當您正在考慮 Roblox 上的廣告會是什麼樣子時，除了品牌營銷之外，是否有可能在廣告方面進行效果營銷？然後，如果有表現，那麼行動號召是否僅基於 Roblox 元宇宙或商品中的經驗？或者它可能是——也可以轉化為現實世界？

Yes. I will -- I don't want to speculate on any dates or any times. I do want to highlight that everything we do is values-consistent and within advertising guidelines and appropriate for the various ages. One can imagine advertising units that appear as in-game bilboards, in-game teleports that all developers on Roblox can choose to use. So this creates a dynamic ad server. Rather than with 2D image or video pre roll, these are in-game 3D units.

是的。我會——我不想推測任何日期或任何時間。我想強調的是，我們所做的一切都是價值觀一致的，符合廣告指導方針，適合各個年齡段。可以想像廣告單元顯示為遊戲中的廣告牌，Roblox 上的所有開發人員都可以選擇使用的遊戲中傳送器。因此，這將創建一個動態廣告服務器。這些是遊戲中的 3D 單元，而不是 2D 圖像或視頻預卷。

You can imagine brands using these units to bring traffic to their own brand experience. We've highlighted in the past brands like Vans and Nike have built brand experiences. We would never claim that they would use this form of advertising. There were other types of brands that might do it.

您可以想像品牌使用這些單位為自己的品牌體驗帶來流量。我們在過去強調過 Vans 和 Nike 等品牌已經建立了品牌體驗。我們絕不會聲稱他們會使用這種形式的廣告。還有其他類型的品牌可能會這樣做。

And then as fans are hanging out in brand experiences, you can imagine a wide range of activities within those experiences that includes acquiring virtual items just like in the physical world. That can include drops. That can include limited numbers of those items that our community would be interested in acquiring. That can include items related to stars, celebrities, athletes that our community would like to acquire virtually.

然後當粉絲在品牌體驗中閒逛時，您可以想像這些體驗中的各種活動，包括購買虛擬物品，就像在現實世界中一樣。這可以包括滴劑。這可能包括我們社區有興趣購買的有限數量的物品。這可能包括我們社區希望虛擬獲得的與明星、名人、運動員相關的物品。

And then I think longer term, that gets into -- the virtual item is tied with a physical item, either through a traditional shopping or, possibly, in the future, those virtual items are bought, those virtual items can go off our platform wrapped in an NFT. And that NFT is really the equivalent of owning and redeeming the physical item. So I think it goes way beyond performance into the whole life cycle of a fan with the brand and wanting to own part of that brand virtually or physically.

然後我認為從長遠來看，虛擬物品與物理物品綁定在一起，要么通過傳統購物，要么可能在未來，這些虛擬物品被購買，這些虛擬物品可以打包離開我們的平台在 NFT 中。而 NFT 實際上就相當於擁有和贖回實物。所以我認為它超越了性能，進入了品牌粉絲的整個生命週期，並希望在虛擬或物理上擁有該品牌的一部分。

And we will take our next question from Eric Sheridan with Goldman Sachs.

我們將向高盛的 Eric Sheridan 提出下一個問題。

I want to come back to the topic you raised in the shareholder letter around the innovation and user safety. Can we get a better sense of how we should be thinking about spend behind user safety in 2022 and a wider lens of how you think about spending on user safety beyond 2022?

我想回到你在股東信中提出的關於創新和用戶安全的話題。我們能否更好地了解我們應該如何考慮 2022 年在用戶安全方面的支出，以及您如何看待 2022 年以後在用戶安全方面的支出？

And then maybe one quick follow-up would be, you called out of examples of how user safety drives continued momentum on the platform. From our position, from the outside looking into the company, how should we think about user safety as driving sort of narratives and momentum around the broader platform you're trying to build for the longer term and measured in terms of business performance?

然後也許一個快速的後續行動會是，你提到了用戶安全如何在平台上推動持續發展的例子。從我們的立場來看，從公司外部來看，我們應該如何將用戶安全視為圍繞您試圖建立的更廣泛平台的敘述和動力，並以業務績效為衡量標準？

Yes. I think we should nuance this in that, on user safety, we would imagine as user growth continues, as engagement continues, as the various ways people use Roblox, whether it's expanding to learning, education, working, school, concerts, we believe safety and stability is a competitive advantage and something that consumers will just assume is there.

是的。我認為我們應該細化這一點，在用戶安全方面，我們可以想像隨著用戶增長的持續，隨著參與度的持續，隨著人們使用 Roblox 的各種方式，無論是擴展到學習、教育、工作、學校、音樂會，我們相信安全穩定性是一種競爭優勢，消費者只會假設它存在。

So in a way, it's a hidden function that we believe consumers will expect. And it's almost -- think of it in reverse as we set very, very high targets for all of our internal safety and stability metrics because we believe we then enforce the expanded growth on the platform.

所以在某種程度上，這是我們相信消費者會期待的隱藏功能。這幾乎是——反過來想，因為我們為我們所有的內部安全和穩定性指標設定了非常非常高的目標，因為我們相信我們隨後會在平台上實施擴大的增長。

We're doing a lot of work in ML. We're doing a lot of work in recognizing that behavior automatically, doing a lot of work in our ways of detecting both bad content and bad conversations. I think these are things that we just view as a foundation of the company and supporters of, really, our reputation and the fact that Roblox is a safe and civil environment. I think that's how we view them as accelerators of long-term growth, rather than maybe the more traditional metrics of what is our retention or what is our hours per user.

我們在 ML 方面做了大量工作。我們在自動識別這種行為方面做了大量工作，在檢測不良內容和不良對話方面做了大量工作。我認為這些是我們只是將其視為公司的基礎和支持者的東西，實際上，我們的聲譽以及 Roblox 是一個安全和文明的環境這一事實。我認為這就是我們將它們視為長期增長加速器的方式，而不是更傳統的衡量標準，即我們的留存率或每位用戶的小時數。

Yes. And Eric, it's Mike. In terms of modeling, I think in the short term, modeling it in a consistent percentage of bookings is probably a decent assumption. It's not an area where we are looking for great amounts of operating leverage. We're looking for high quality and safety and breakthroughs for the user base.

是的。還有埃里克，是邁克。在建模方面，我認為在短期內，以一致的預訂百分比對其進行建模可能是一個不錯的假設。這不是我們正在尋找大量經營槓桿的領域。我們正在為用戶群尋找高質量和安全性的突破。

Having said that, as Dave mentioned, they make the platform better. And you can argue they actually make monetization easier and better if people feel like they're in a safe and civil environment. So in the short run, I would say it'd be relatively similar as a percentage of bookings as what you've seen in the last few quarters. If there is leverage there, it's because top line just grew a little bit faster. But overall, as an area of focus, for aggregate dollars, this one is high on the list.

話雖如此，正如戴夫所說，他們讓平台變得更好。如果人們覺得他們處於安全和文明的環境中，你可以爭辯說他們實際上使貨幣化變得更容易和更好。所以在短期內，我會說它與您在過去幾個季度看到的預訂百分比相對相似。如果那裡有槓桿作用，那是因為頂線增長得更快了一點。但總體而言，作為一個重點領域，就總美元而言，這一項在名單上名列前茅。

Yes. And then I'm not going to say anything contradictory to what Mike said. But a big part of what our safety and stability team is doing is, as we automate tasks that human moderators and human reviewers do, without making any forward-looking prediction on that, one can imagine our safety and stability system actually becomes more efficient. And so we are both not claiming any long-term increases in efficiency, but simultaneously seeing great gains internally on our ability to automate tasks that human moderators do.

是的。然後我不會說任何與邁克所說的相矛盾的東西。但是，我們的安全和穩定性團隊正在做的很大一部分事情是，當我們自動執行人工審核員和人工審核員所做的任務時，無需對此做出任何前瞻性預測，可以想像我們的安全和穩定性系統實際上變得更加高效。所以我們都沒有聲稱效率有任何長期的提高，但同時在內部看到我們在自動化人類主持人所做的任務的能力方面取得了巨大的進步。

And we will take our next question from David Karnovsky with JPMorgan.

我們將向摩根大通的大衛卡諾夫斯基提出下一個問題。

David, in the shareholder letter, you noted some changes to product functionality to drive higher frequency and retention for that 9- to 12-year old cohort of U.S. users. Maybe just walk through what some of these updates are. And then are you already seeing the impact of that in your Q1 U.S., Canada DAU figure? Or is the increase there kind of more from aging up?

大衛，在股東信中，您注意到產品功能的一些變化，以提高 9 至 12 歲美國用戶群體的頻率和保留率。也許只是了解其中一些更新是什麼。然後您是否已經在您的第一季度美國、加拿大 DAU 數據中看到了這種影響？還是因為年齡增長而增加的更多？

Yes, this one is tricky because we do not share MAU numbers. And so you cannot discern our DAU to MAU frequency. We've historically, not just through COVID, but over the last 3 or 4 years, made gains in these frequency numbers. And at the same time, there's many more people on our platform monthly than there are daily. Frequency goes up and down the [snack].

是的，這很棘手，因為我們不共享 MAU 數字。所以你無法辨別我們的 DAU 到 MAU 頻率。從歷史上看，我們不僅通過 COVID，而且在過去的 3 或 4 年中，在這些頻率數字上都取得了進步。同時，我們平台上每月的人數比每天的人數還要多。頻率在[零食]中上下波動。

It goes all the way from what is the -- it might seem counterintuitive, but what is the raw launch time of our mobile app? What is the raw launch time of getting into an experience? That influences how quickly can someone get into the game. We've made games technically so that if you're in the middle of a Roblox experience and on your mobile phone patch, switch out to some other app and then come back in, you don't lose your context.

它一直從什麼是 - 這似乎違反直覺，但我們的移動應用程序的原始啟動時間是多少？進入體驗的原始啟動時間是多少？這會影響某人進入遊戲的速度。我們在技術上製作了遊戲，因此如果您正在體驗 Roblox 並使用手機補丁，切換到其他應用程序然後再回來，您不會丟失您的上下文。

Longer term, we believe anything around improving the usage of Roblox as a spontaneous communication medium for shorter conversations, maintaining context with friends will drive that frequency. And we're -- I guess, I'm personally optimistic, we have a lot of headroom there.

從長遠來看，我們相信任何關於改善 Roblox 作為短對話的自發溝通媒介的使用，與朋友保持上下文將推動這種頻率。而且我們 - 我想，我個人很樂觀，我們有很大的空間。

Yes. And then just, David, on the question about the breakout between age demos. In the U.S., in 9 to 12, when COVID started, the highest rate of penetration was in the 9 to 12 age demo, higher than 13 -- anything 13-plus. And so when COVID started, we had the highest number of users immediately increase their frequency really overnight, lots of hours and subsequently lots of associated bookings.

是的。然後，大衛，關於年齡演示之間突破的問題。在美國，當 COVID 開始時，9 到 12 歲的滲透率最高的是 9 到 12 歲的演示，高於 13 歲——13 歲以上的任何人。因此，當 COVID 開始時，我們有最多的用戶立即增加了他們的頻率，真的是一夜之間，很多小時，隨後還有很多相關的預訂。

In 13 to 16, 17 to 24, similar dynamics, but on a much lower rate of market penetration. And so in those markets, we had more consistent growth in frequency and bookings than we had in 9 to 12, where you jumped up very, very quickly. And then depending on where we were with lockdowns, we would see those numbers go back up. And as we reopen, users are still growing, but the frequency just changes as people go back to school.

在 13 至 16 日、17 至 24 日，情況類似，但市場滲透率要低得多。因此，在這些市場中，我們在頻率和預訂方面的增長比我們在 9 到 12 年間更加穩定，在這些市場中，你的增長非常非常快。然後根據我們在哪裡進行封鎖，我們會看到這些數字回升。隨著我們重新開放，用戶仍在增長，但隨著人們重返學校，頻率只會發生變化。

In the older age demos, though, again, we're getting a higher percentage of new users based on the base in which we started. So those age demos today are caught back up with peak, basically, whereas the 9 to 12 is still below. We assume, over time, as we, again, continue to add users in 9 to 12 and get higher levels of penetration and as frequency improves based on product changes and other things and just lapping these incredibly high periods of frequency in COVID, that will return to growth throughout U.S.

不過，在較早的演示中，我們再次獲得了更高比例的新用戶，這取決於我們開始的基礎。因此，今天的那些年齡演示基本上都趕上了高峰，而 9 到 12 歲仍然低於。我們假設，隨著時間的推移，隨著我們再次在 9 到 12 歲之間增加用戶並獲得更高水平的滲透率，並且隨著產品變化和其他事情的頻率提高，並且只是在 COVID 中重複這些令人難以置信的高頻率週期，這將在美國恢復增長

Okay. And then, Mike, maybe just one on margins. I wanted to see if there was any framework you'd be willing to provide for the balance of the year. And just given normal seasonality, should we assume some positive operating leverage as you kind of get into your heavier bookings quarter?

好的。然後，邁克，也許只是邊際上的一個。我想看看你是否願意為今年的餘額提供任何框架。考慮到正常的季節性，我們是否應該在您進入較重的預訂季度時假設一些積極的經營槓桿？

Yes, good question. Pre-COVID, and as a private company, we would routinely be in the low double digits of EBITDA margins and very high free cash flow margins. And it was a simple -- a company who is self-financing, and we generated a lot of free cash, just as a private company, over $1 billion. As we saw the step function in top line during COVID, obviously, you saw a lot of operating leverage really across all cost areas in Roblox.

是的，好問題。在 COVID 之前，作為一家私營公司，我們通常會處於 EBITDA 利潤率的低兩位數和非常高的自由現金流利潤率。這很簡單——一家自籌資金的公司，我們產生了大量的自由現金，就像一家私人公司一樣，超過 10 億美元。正如我們在 COVID 期間看到的最重要的階梯函數，很明顯，您在 Roblox 的所有成本領域都看到了很多運營槓桿。

And we went from that low double-digit, maybe mid-teens EBITDA rates to low 30s. And at that time, we felt a keen opportunity to keep investing in the business. We were not -- it was not a goal to maintain margins at 30%. So we continue to invest in hiring great people. We certainly continue to invest in our developer community, trust and safety and infrastructure. So we continue to make investments for the long run to build the platform.

我們從低兩位數，可能是十幾歲的 EBITDA 利率下降到了 30 多歲。那時，我們覺得這是一個繼續投資業務的絕佳機會。我們不是——將利潤率保持在 30% 並不是我們的目標。因此，我們繼續投資於招聘優秀人才。我們當然會繼續投資於我們的開發者社區、信任和安全以及基礎設施。因此，我們將繼續進行長期投資以構建平台。

Today, we come into the market with $3 billion of cash. So we're incredibly liquid. And the question is, should we continue to invest in the business for the long run? Or should we try to manage our margins? We are not going to test anyone's comfort levels on liquidity. This is a very liquid business. But in the short run, I don't think we're that focused on trying to maintain very high P&L margins. If it happens, that's fine.

今天，我們帶著 30 億美元的現金進入市場。所以我們的流動性非常好。問題是，我們是否應該繼續長期投資這項業務？還是我們應該嘗試管理我們的利潤？我們不會測試任何人對流動性的舒適度。這是一項流動性很強的業務。但在短期內，我認為我們不會那麼專注於試圖維持非常高的損益利潤率。如果發生了，那很好。

But right now, the investments that we're making, we see very high returns on in the long run, and we're very liquid. So we're going to be much more focused on doing the things that we think add value, and we'll let margins play out. Again, if bookings grow a little bit faster, that will be beneficial to margins.

但是現在，我們正在進行的投資，從長遠來看，我們看到了非常高的回報，而且我們的流動性非常好。因此，我們將更加專注於做我們認為增加價值的事情，我們會讓利潤發揮作用。同樣，如果預訂量增長得更快一點，那將有利於利潤率。

But I don't think we want to let that constrain investment, especially in hiring the dev community, trust and safety and infrastructure, because the unit economics of this business have always been very strong. And if we needed to dial back some investment, we certainly could do that. I don't think that's really the posture that we're taking right now. We see this as a great opportunity to keep investing.

但我認為我們不想讓這種限制投資，特別是在僱用開發社區、信任和安全以及基礎設施方面，因為該業務的單位經濟一直非常強勁。如果我們需要回撥一些投資，我們當然可以這樣做。我認為這不是我們現在真正採取的姿態。我們認為這是繼續投資的絕佳機會。

We will take our next question from Brian Nowak with Morgan Stanley.

我們將向摩根士丹利的 Brian Nowak 提出下一個問題。

I have 2. The first one, just around sort of the shape of the year into next year. I think in the fourth quarter letter, you sort of had some commentary about, as you go throughout the course of the year, the comps become more normal, the growth rates and bookings and users should get closer and that the exit rate for the year is sort of going to get over '23. Can you just talk to us about, is that still the way you're thinking about the shape of the year in '23? And sort of puts and takes of where you could overperform or perhaps face more challenges around bookings and user commentary from last quarter is the first one.

我有 2 個。第一個，大約是明年到明年的形狀。我認為在第四季度的信中，你有一些評論說，隨著你在一年中的整個過程，comps 變得更加正常，增長率、預訂量和用戶應該越來越接近，以及今年的退出率有點過了'23。你能和我們談談嗎，這仍然是你在考慮 23 年的形狀的方式嗎？第一個是關於上個季度的預訂和用戶評論方面你可能表現出色或可能面臨更多挑戰的看跌期權。

And the second one, the developer fees as a percentage of bookings did increase a bit in the quarter. And maybe, Mike, you just answered this question in the last one, but just talk to us about how you're thinking about developer fees philosophically. How is the mix changing? And how should we think about developer fee deleverage for the year?

第二個，開發商費用佔預訂的百分比在本季度確實有所增加。也許，邁克，你剛剛在上一個問題中回答了這個問題，但請與我們談談你是如何從哲學上考慮開發人員費用的。組合如何變化？我們應該如何看待今年的開發商費用去槓桿？

Yes. Thanks, Brian. Overall, what -- again, when we came into the year, we thought we would bottom in April in terms of year-over-year growth rate of bookings. Again, last year, bookings were very high in '21, and it was the last -- sort of the last big quarter before the economy started to reopen and people started to go outside again. So we saw in May, June and July actually declines, much less year-over-year growth a year ago and actual declines in our users in the U.S. and bookings coming from the U.S. even as we're growing internationally.

是的。謝謝，布賴恩。總體而言，當我們進入這一年時，我們認為我們將在 4 月的預訂量同比增長率方面觸底。同樣，去年，21 年的預訂量非常高，這是經濟開始重新開放和人們再次開始外出之前的最後一個重要季度。因此，我們在 5 月、6 月和 7 月看到實際下降，遠低於一年前的同比增長，我們在美國的用戶和來自美國的預訂量的實際下降，即使我們在國際上不斷增長。

So this year, I think we have the opposite impact. We're already open, and we think normal seasonality will return. And so what we had hoped to see is that we would have sequential improvements in the year-over-year growth rate of bookings. We thought April would be the bottom. It turns out, I think maybe March was the bottom, which is great.

所以今年，我認為我們產生了相反的影響。我們已經開放，我們認為正常的季節性會恢復。因此，我們希望看到的是，預訂量的同比增長率將連續提高。我們認為四月將是底部。事實證明，我認為三月可能是底部，這很棒。

April was a little bit better than March. We suspect that May will be a little bit better, and that will continue as we go through the summer. And that ought to shrink that gap between user growth numbers and the rate of bookings growth. And that's our expectation, especially as we go through the summer.

四月比三月好一點。我們懷疑五月會好一點，並且隨著我們整個夏天的到來，這種情況會持續下去。這應該會縮小用戶增長數量與預訂增長率之間的差距。這是我們的期望，尤其是在我們度過夏天的時候。

Now if user growth continues really, really strong, then we'll close the gap but we're not going to necessarily catch up with that number. But that's fine. As long as we're improving sequentially and the comparisons are, in fact, getting easier, that's really what we're looking for is that sign.

現在，如果用戶增長繼續非常非常強勁，那麼我們將縮小差距，但我們不一定會趕上這個數字。但這沒關係。只要我們按順序改進並且實際上比較變得更容易，這就是我們真正要尋找的標誌。

And again, this time last year, we actually had almost negative seasonality from what you would expect. This year, we expect much more normal seasonality. So that's a good setup for us for improved -- continued improved growth, year-over-year growth as we go from April, May, June, et cetera. And then before too long, we will have this sort of COVID -- hopefully, the COVID distortions out of the numbers on a year-over-year basis.

再一次，去年的這個時候，我們實際上幾乎沒有你所期望的負季節性。今年，我們預計會有更正常的季節性。因此，這對我們來說是一個很好的改進設置——從 4 月、5 月、6 月等開始，增長持續改善，同比增長。然後用不了多久，我們就會有這種新冠病毒——希望新冠病毒在數字上的年復一年失真。

On dev fees, we're making investments in the dev community. We have both our normal rates of take from transactions as well as engagement-based payouts. In the last year-plus, we've really leaned into engagement-based. We think there have been some real benefits from that. And again, I don't think of the dev community is a place where we see short-term operating leverage. We see an opportunity to keep investing in an incredible developer community.

在開發費用方面，我們正在對開發社區進行投資。我們既有正常的交易費率，也有基於參與度的支出。在過去一年多的時間裡，我們真的傾向於基於參與度。我們認為這帶來了一些真正的好處。再說一次，我不認為開發社區是我們看到短期運營槓桿的地方。我們看到了繼續投資於一個令人難以置信的開發者社區的機會。

What we want is developers feeling like they can continue to invest in Roblox and build their businesses on our platform. And so it's not an area in which we see we're going after leverage. We're still -- we have talked several quarters ago about moving into the 25% of bookings range. I think we were at 23% this past quarter. So we're in a very comfortable range right now. And so I don't -- I think that's a good way to model it out, if you will.

我們想要的是開發人員覺得他們可以繼續投資 Roblox 並在我們的平台上建立他們的業務。因此，我們認為這不是我們追求槓桿的領域。我們仍然 - 幾個季度前我們已經討論過進入 25% 的預訂範圍。我認為我們在上個季度達到了 23%。所以我們現在處於一個非常舒適的範圍內。所以我不——我認為這是一個很好的建模方式，如果你願意的話。

I just want to highlight that, with the current place our developer fees are at, we see more and more big organic brands coming to the platform. The Sonic experience on Roblox was highly organic and not an experience where we go out and really push for that to happen. The Spotify experience, once again, highly organic.

我只想強調一點，就目前我們的開發者收費而言，我們看到越來越多的大型有機品牌進入該平台。 Roblox 上的 Sonic 體驗是高度有機的，而不是我們走出去並真正推動它發生的體驗。 Spotify 的體驗再一次是高度有機的。

So this goes to our vision of having an economy that we migrate from the individual hobbyists to small companies. And so we will pass -- I believe we already have 100-person studio working on our platform to VC-funded studios now working on our platform, to brands creating experiences on our platform organically, which is kind of a validation of our developer fee structure.

因此，這符合我們的願景，即擁有一個我們從個人愛好者轉移到小公司的經濟。所以我們會通過——我相信我們已經有 100 人的工作室在我們的平台上工作，現在有風投資助的工作室在我們的平台上工作，還有品牌在我們的平台上有機地創造體驗，這是對我們開發人員費用的一種驗證結構體。

And we will take our next question from Omar Dessouky with Bank of America.

我們將向美國銀行的 Omar Dessouky 提出下一個問題。

I have 2 of them, one on layered clothing and one on your development strategy and philosophy. So in terms of layered clothing, since you released it in March and more fully in April, has adoption been ahead or behind your expectations? On your blog, you described layered clothing as a stepping-stone technology.

我有兩個，一個關於分層服裝，一個關於你的發展戰略和理念。那麼在分層服裝方面，自從您在 3 月發布並且在 4 月更全面發布以來，採用率是領先於還是落後於您的預期？在您的博客上，您將分層服裝描述為一種墊腳石技術。

How much time does that need to spend in production? Or what milestones does it need to reach before you feel confident about releasing UGC avatar bodies and dynamic heads? And have your learnings on layered clothing to date caused you to pull in or push out when you might release those 2 derivative technologies versus your thinking as of late last year? And I'll follow up with my second question after your answer.

這需要在生產上花費多少時間？或者，在您對發布 UGC 化身和動態頭部充滿信心之前，它需要達到哪些里程碑？到目前為止，你對分層服裝的了解是否讓你在可能發布這兩種衍生技術時加入或退出，而不是你去年年底的想法？在你回答之後，我會跟進我的第二個問題。

Yes. I'll share maybe a little bit of internally what you would hear at the company, and that is, we -- I personally believe that, as we move to user-created bodies and heads, the long-term impact of that could be bigger than layered clothing because, ultimately, who you are is your body and your head, and that's underway right now.

是的。我可能會在內部分享一些你在公司會聽到的內容，也就是說，我們——我個人認為，隨著我們轉向用戶創建的身體和頭部，其長期影響可能是比分層衣服更重要，因為最終，你是誰就是你的身體和你的頭，而這現在正在進行中。

We had really good adoption of layered clothing. I think roughly 30% of our daily actives are wearing it right now. The -- but bodies and heads, we believe, will be a much more -- I personally, I don't want to quote whether it's going to move our metrics. But my own personal thing is, it's just as big, if not bigger, and that is well underway. We've talked about it before. We want to get that out as quickly as possible.

我們很好地採用了分層服裝。我認為現在大約 30% 的日常活動人士都穿著它。 - 但是我們相信，身體和頭部會更多 - 我個人，我不想引用它是否會改變我們的指標。但我個人的事情是，它和它一樣大，如果不是更大的話，而且正在順利進行中。我們之前已經談過了。我們希望盡快解決這個問題。

Okay. Right. So then it's more of a strategic question, just to finish up. Could you tell me why your strategy of developing new technologies, like layered clothing, to be compatible with previous generations of digital matter is difficult for competitors to replicate? And how it might create barriers to entry as competition begins to -- competitors begin to launch their metaverse-like platforms geared towards your similar core demographic?

好的。對。所以這更像是一個戰略問題，只是為了完成。您能否告訴我為什麼您開發新技術（例如分層服裝）以與前幾代數字物質兼容的戰略很難讓競爭對手複製？隨著競爭的開始，它如何創造進入壁壘——競爭對手開始推出面向類似核心人口的類似虛擬世界的平台？

Yes. This is a great question. I want to highlight that, long term, when we view what makes Roblox competitive, it's a holistic collection of things. It includes the momentum and size of our very large creator community. It includes the quality of our development platform and tool set. It includes the foundation we have on safety and stability.

是的。這是一個很好的問題。我想強調的是，從長遠來看，當我們查看使 Roblox 具有競爭力的原因時，它是一個整體的集合。它包括我們非常龐大的創作者社區的勢頭和規模。它包括我們的開發平台和工具集的質量。它包括我們在安全性和穩定性方面的基礎。

It includes the innovations we're making right now, including UGC bodies and UGC layered clothing. But it also includes our ability to continuously innovate and release big innovations over the next 1 to 5 years, just as we have over the last 15 years. So these are all part of what I believe makes Roblox extremely competitive, but ultimately, it's our ability to innovate.

它包括我們現在正在進行的創新，包括 UGC 身體和 UGC 分層服裝。但它也包括我們在未來 1 到 5 年內不斷創新和發布重大創新的能力，就像我們在過去 15 年中所做的那樣。所以這些都是我認為讓 Roblox 極具競爭力的部分原因，但歸根結底，這是我們的創新能力。

We have a unique stack that we vertically control all the way from our game engine to our cloud, to our clients, to our developer tools that operates really as a consistent whole. So I think our long-term competitive advantage is much more a holistic mix of all of these. And one could think of any of those as moats.

我們有一個獨特的堆棧，從我們的遊戲引擎到我們的雲、到我們的客戶、到我們的開發人員工具，我們都可以垂直控制它，這些堆棧真正作為一個一致的整體運行。所以我認為我們的長期競爭優勢更像是所有這些的整體組合。人們可以將其中任何一個視為護城河。

We typically think more in terms of long-term innovation as opposed to relying or depending on anything we might think of a moat. Basically, innovation is moving forward. And the best way for Roblox to get to 1 billion monthly actives is not to simply defend what we might think of as moat. It's really to continue creating our vision of the metaverse, which is still very early, and we feel we have a unique handle on what that's going to be.

我們通常更多地考慮長期創新，而不是依賴或依賴任何我們可能認為的護城河。基本上，創新正在向前發展。 Roblox 獲得 10 億月活躍用戶的最佳方式不是簡單地捍衛我們可能認為的護城河。這真的是繼續創造我們對元宇宙的願景，這還很早，我們覺得我們對未來的發展有一個獨特的把握。

And we'll take our next question from Brandon Ross with LightShed Partners.

我們將向 LightShed Partners 的 Brandon Ross 提出下一個問題。

I have a couple. Earlier in answering I think it was David's question, you spoke to the high ROI on your investment spend. And I guess, throughout the call and the letter, you've spoken to platform features that have rolled out in the past several quarters. Is there any way for you to quantify or explain the impact of any of these features on engagements and/or monetization? And will we see some more tangible impact from the likes of voice and layered clothing as this year goes on? And then I have a follow-up.

我有一對。在回答之前，我認為這是大衛的問題，您談到了投資支出的高投資回報率。我想，在整個電話和信函中，您已經談到了過去幾個季度推出的平台功能。您有什麼方法可以量化或解釋這些功能對參與度和/或貨幣化的影響？隨著今年的繼續，我們會從聲音和分層服裝等方面看到更切實的影響嗎？然後我有一個跟進。

Yes. I want to share, what you're mentioning is something internally we try to do as much as possible. And I'll highlight also -- Mike, you can jump in if you want. In the past, through periods of rapid growth and just as we have rapid growth right now in many cohorts in many places around the world, we've tried to analyze the contributors to that and analyze the last year of product releases and tried to create statistical correlation. And we found that, that's difficult to do.

是的。我想分享一下，你提到的是我們在內部盡可能多地嘗試做的事情。我還要強調——邁克，如果你願意，你可以加入。過去，通過快速增長期，就像我們現在在世界各地的許多群體中快速增長一樣，我們試圖分析其貢獻者並分析產品發布的最後一年，並試圖創造統計相關性。我們發現，這很難做到。

I think with some of the innovations that we have coming right now, like spatial audio, we will be able to measure the time that those users are engaged in audio communication rather than text communication. And I think that will be a good signal going forward for that.

我認為通過我們現在即將推出的一些創新，比如空間音頻，我們將能夠測量這些用戶進行音頻通信而不是文本通信的時間。我認為這將是一個很好的信號。

So we will try, going forward, with some of these big innovations, whether it's spatial audio, our avatar system, our translate system, our developer cloud to share as much as we can about what makes those users possibly more engaged or possibly more retained. Generally, when we find users that are more engaged and more retained, that does contribute to the growth of the overall platform. So where we can, we will try to share those metrics with you.

因此，我們將在未來嘗試一些重大創新，無論是空間音頻、我們的頭像系統、我們的翻譯系統、我們的開發者云，以盡可能多地分享是什麼讓這些用戶可能更投入或可能更留存.一般來說，當我們發現用戶參與度更高、留存率更高時，這確實有助於整個平台的增長。因此，我們將盡可能與您分享這些指標。

Yes. And Brandon, having said that, take any point in time, a 3-, 4-, 5-year window of this business, and look at the users' engagement and the bookings growth of the business and the platform, the growth rates have been very significant. The investment areas have been very consistent, investment-grade engineers.

是的。而布蘭登，話雖如此，在任何一個時間點，這個業務的 3 年、4 年、5 年的窗口，看看用戶的參與度以及業務和平台的預訂增長，增長率都有非常重要。投資領域一直很一致，投資級別的工程師。

We have a product road map that we -- as we're ushering in a category, we make calls on what we think will make that better for our developers and our -- and for our users. And generally, the output of that has been very high growth, more users around the world, an older age demographic, more and more developers building great content, enormous amount of engagement and huge growth in the top line of the business.

我們有一個產品路線圖，當我們引入一個類別時，我們會呼籲我們認為會為我們的開發人員和我們的用戶以及我們的用戶帶來更好的產品路線圖。總的來說，其產出是非常高的增長，世界各地的用戶越來越多，人口老齡化，越來越多的開發人員構建了偉大的內容，大量的參與和業務收入的巨大增長。

So any specific decision is meant to continue to advance things that we have very good proof of generating high rates of return over time. And so, of course, every time we launch something, we want it to be high return. But generally, the business and the business model that we have has fantastic unit economics, and so we're always trying to improve those.

因此，任何具體決定都是為了繼續推進我們有很好的證據證明隨著時間的推移會產生高回報率的事情。所以，當然，每次我們推出一些東西時，我們都希望它是高回報的。但總的來說，我們擁有的業務和商業模式具有出色的單位經濟效益，因此我們一直在努力改進這些。

And over a very long period of time, the return dynamics here have been excellent. So we're going to continue to make those kinds of decisions based on a lot of historical good decision-making. And in the lack of saturation, which we don't believe we have anywhere, we're quite comfortable that those will generate high returns in the future.

在很長一段時間內，這裡的回報動態非常好。因此，我們將繼續根據許多歷史上的良好決策做出此類決策。在缺乏飽和度的情況下，我們認為我們在任何地方都沒有，我們對這些將在未來產生高回報感到非常滿意。

Yes. I'll share a few just tidbit metrics on layered clothing. As of the end of March, I believe, over 200 million items have already been acquired by the players and people on our platform. On voice, we're rolling this out in a very values-consistent and conservative way for validated users to start using our new validation system. And for those users that we've rolled this out to, they're averaging 20 minutes roughly per daily activity using spatial audio.

是的。我將分享一些關於分層服裝的花絮指標。截至3月底，我相信我們平台上的玩家和人們已經獲得了超過2億件物品。在語音方面，我們以非常一致且保守的方式推出此功能，讓經過驗證的用戶開始使用我們的新驗證系統。對於我們已經向其推出的用戶來說，他們平均每天使用空間音頻進行大約 20 分鐘的活動。

Great. And then just -- I was kind of comparing in my mind the approach that you've taken with outside IP holders with that of Epic. And it seems like you've really taken more of a hands-off approach outside of maybe music, while Epic has really aligned themselves closely with some key IPs such as Disney and then LEGO as a way to kind of widen the funnel. Why is your approach better? And does it make sense to closely align with key IP holders for you?

偉大的。然後只是 - 我在腦海中將您與外部 IP 持有人採取的方法與 Epic 的方法進行比較。似乎你真的在音樂之外採取了更多不干涉的方法，而 Epic 確實與迪士尼和樂高等一些關鍵 IP 緊密結合，以此作為拓寬渠道的一種方式。為什麼你的方法更好？與您的關鍵 IP 持有者密切合作是否有意義？

I think this is a really good question, and this goes to the core of our Roblox division and how we build a platform. Over 15 years ago, when we started our work on this category, what sometimes is referred to as user-generated content, or UGC, internally we refer to as self-service. And self-service is really difficult.

我認為這是一個非常好的問題，這涉及到我們 Roblox 部門的核心以及我們如何構建平台。 15 多年前，當我們開始研究這一類別時，有時被稱為用戶生成內容或 UGC，我們在內部將其稱為自助服務。而且自助服務真的很難。

It's hard to build a platform where everything from a small individual hobbyist all the way to, as I just referred to, Sonic the Hedgehog shows up on Roblox. It's hard to build a platform where those things happen without a custom compiler, a custom build, without a biz dev process, without a lot of close interworking.

很難建立一個平台，從小型個人愛好者一直到我剛才提到的刺猬索尼克出現在 Roblox 上。如果沒有自定義編譯器、自定義構建、沒有商務開發流程、沒有大量密切的交互工作，就很難構建一個平台，讓這些事情發生。

But we believe what we've seen in games and play will more and more move forward in brands, in music and, ultimately, in custom IP on our platform. So this is a vision for us. I don't want to comment on Epic or LEGO or what they're doing.

但我們相信，我們在遊戲和遊戲中所看到的將在品牌、音樂以及最終在我們平台上的定制 IP 中越來越多地向前發展。所以這對我們來說是一個願景。我不想評論 Epic 或 LEGO 或他們在做什麼。

We -- I personally believe there's enormous scalability in building a self-service platform in all of these dimensions, a single unified platform where, ultimately, a brand like LEGO would build a LEGO experience on top of Roblox and more and more be able to present that as their unique experience. So this is more the direction we're going. We believe it's very, very scalable. And at the same time, I think there's a lot of different ways that people have at looking at this.

我們——我個人認為，在所有這些維度上構建自助服務平台具有巨大的可擴展性，一個統一的平台，最終，像 LEGO 這樣的品牌將在 Roblox 之上構建 LEGO 體驗，並且越來越能夠將其作為他們獨特的體驗。所以這更像是我們要去的方向。我們相信它非常非常可擴展。同時，我認為人們看待這個問題的方式有很多。

And we have time for one more question, and that question will be from Clark Lampen with BTIG.

我們還有時間再提一個問題，這個問題將來自 BTIG 的 Clark Lampen。

I have 2. The first is on engagement. Dave, I wanted to come back to what you had said about U.S. 9 to 12 cohort performance. If we maybe go up a layer and think about the sort of 2.3, I think it is, hours that users are spending on platform right now, do you expect that to sort of stabilize at kind of an above pre-COVID rate going forward?

我有 2 個。第一個是關於訂婚的。戴夫，我想回到你所說的關於美國 9 到 12 歲隊列表現的內容。如果我們可能會再上一層，考慮一下 2.3，我認為是用戶現在在平台上花費的時間，你是否期望它會穩定在高於 COVID 之前的速率？

And as we think about sort of bookings and hours over time, is that -- does the incremental hour generation come from growth of the overall user base? Or do you see upside to the average user sort of time spent on platform? And then I got a follow-up on OpEx.

當我們考慮一段時間內的預訂量和小時數時，增加的小時數是否來自整體用戶群的增長？或者你認為平均用戶在平台上花費的時間有什麼好處嗎？然後我對 OpEx 進行了跟進。

Yes. I think I shared earlier, there's a lot of headroom in our -- one of our core cohorts, such as USA 9 through 12. What I would like to share is, internally, there's room on the monthly active. They're not a huge room. We have such a big portion of the 9- to 12-year olds in the U.S. already on the platform, but there's room there. There's a lot of room on frequency, which is our DAU to MAU ratio.

是的。我想我之前分享過，我們的核心團隊之一，例如美國 9 到 12 年級，有很多空間。我想分享的是，在內部，每月活躍度還有空間。他們不是一個很大的房間。在美國，我們有很大一部分 9 到 12 歲的孩子已經在這個平台上，但那裡還有空間。頻率有很大的空間，這是我們的 DAU 與 MAU 的比率。

And as I shared earlier, that cohort is spending slightly less time as we've emerged from COVID, as they were within COVID. I would say, our focus will be on more of the DAU to MAU ratio. We believe that's correlated with retention. We believe that focus will ultimately drive even bigger MAU ratios. And we believe that's where we're going to be heads down a bit more than hours per DAU, especially in that cohort.

正如我之前分享的那樣，隨著我們從 COVID 中出現，該隊列花費的時間略少，因為他們在 COVID 中。我想說，我們的重點將更多地放在 DAU 與 MAU 的比率上。我們認為這與留存率有關。我們相信，專注最終將推動更大的 MAU 比率。而且我們相信這就是我們將在每個 DAU 中降低超過幾個小時的時間，尤其是在那個隊列中。

Just on, Clark, on hours per DAU, we did see a trend. Obviously, pre-COVID, we were at a certain level, very high. Then during COVID, we went very, very high hours per DAU. During COVID, we've come back down a little bit as reopening has started. But -- so we're down below peak quoted hours per DAU, but we are higher than we were hours per DAU going into it.

就在克拉克，每 DAU 的小時數，我們確實看到了一種趨勢。顯然，在 COVID 之前，我們處於一定水平，非常高。然後在 COVID 期間，我們每個 DAU 的工作時間非常非常長。在 COVID 期間，隨著重新開放的開始，我們已經有所回落。但是 - 所以我們低於每個 DAU 的峰值報價小時數，但我們高於每個 DAU 的小時數。

And so as users have come into Roblox, spent a lot of time, even as we've been growing the user base, there is an even higher level of overall engagement. And that's really true around the world. So in all of our regions, we see the same behavior. So we do feel like we're starting at -- as we've reopened it, at an accelerated level vis-a-vis where we were coming into the pandemic.

因此，隨著用戶進入 Roblox，花費了大量時間，即使我們一直在擴大用戶群，整體參與度也會更高。這在世界各地都是真的。因此，在我們所有的地區，我們都看到了相同的行為。因此，我們確實覺得我們正在開始 - 當我們重新開放它時，相對於我們進入大流行的地方，它的速度更快。

That's helpful. And then just on hiring, Mike, most of your employees are sort of engineering talent. That hiring backdrop has gotten a lot more competitive. I know in the past, you've talked about wanting to increase headcount pretty significantly versus sort of late '21 levels. Are you having any trouble in this market either finding talent? Or is really the by-product of what we're seeing right now mainly higher cost per head inflation?

這很有幫助。然後就招聘而言，邁克，您的大多數員工都是工程人才。這種招聘背景變得更具競爭力。我知道在過去，你曾談到想要顯著增加員工人數，而不是 21 年後期的水平。您在這個市場上尋找人才有什麼困難嗎？還是我們現在看到的主要是更高的人均成本通脹的副產品？

Yes. What you're seeing is, it is a competitive market, and like everyone has seen a class of talented folks go up for sure. Right now, we're ahead of our internal plans, which is great. It was a good first quarter in terms of recruiting. We continue to feel like we're a fantastic place for talented engineers and product professionals and even those in the G&A functions to build their careers. And we expect to continue to grow our talented staff. And so far, this year, we're doing great against our targets.

是的。你所看到的是，這是一個競爭激烈的市場，就像每個人都看到了一群有才華的人肯定會上升。現在，我們領先於我們的內部計劃，這很棒。就招聘而言，這是一個不錯的第一季度。我們仍然覺得我們是一個非常適合有才華的工程師和產品專業人士，甚至是 G&A 職能部門的人來建立他們的職業生涯的地方。我們希望繼續培養我們有才華的員工。到目前為止，今年，我們在實現目標方面做得很好。

Thank you for joining us today. Abby, that's a wrap for us.

感謝您今天加入我們。艾比，這對我們來說是一個包裝。

Thank you. And ladies and gentlemen, this concludes today's conference call. We thank you for your participation, and you may now disconnect.

謝謝你。女士們，先生們，今天的電話會議到此結束。我們感謝您的參與，您現在可以斷開連接。