Affirm Holdings, Inc. (AFRM) 2026 Q3 法說會逐字稿

Good afternoon. Welcome to the Affirm Holdings, Inc. third quarter fiscal 2026 earnings call. Following the speaker's remarks, we will open the lines for your questions. (Operator Instructions)

下午好。歡迎參加 Affirm Holdings, Inc. 2026 財政年度第三季財報電話會議。在講者致詞後，我們將開放電話線供各位提問。(接線員指示)

I'd now like to turn the call over to Zane Keller, Head of Investor Relations. Thank you. You may begin.

現在我想把電話交給投資人關係主管 Zane Keller。謝謝。您可以開始了。

Thank you, operator. Before we begin, I would like to remind everyone listening that today's call may contain forward-looking statements. These forward-looking statements are subject to numerous risks and uncertainties, including those set forth in our filings with the SEC, which are available on our Investor Relations website.

謝謝，接線員。在開始之前，我想提醒各位，今天的電話會議可能包含前瞻性陳述。這些前瞻性陳述受到多項風險與不確定性影響，包括我們向美國證券交易委員會（SEC）提交之文件中所載者；相關文件可於我們的投資人關係網站查閱。

The actual results may differ materially from any forward-looking statements that we make today. These forward-looking statements speak only as of today and the company does not assume any obligation or intent to update them except as required by law. In addition, today's call may include non-GAAP financial measures. These measures should be considered as a supplement to and not a substitute for GAAP financial measures.

實際結果可能與我們今天所作的任何前瞻性陳述有重大差異。這些前瞻性陳述僅截至今日為止，公司除法律要求外，並無任何義務或意圖更新。此外，今天的電話會議可能包含非 GAAP 財務衡量指標。這些指標應視為 GAAP 財務衡量指標之補充，而非替代。

For historical non-GAAP financial measures, reconciliations to the most directly comparable GAAP measures can be found in our earnings supplement slide deck, which is available on our Investor Relations website.

就歷史非 GAAP 財務衡量指標而言，與最直接可比之 GAAP 指標的調節表可見於我們的財報補充簡報投影片，該資料可於我們的投資人關係網站取得。

Hosting today's call with me are Max Levchin, the firm's Founder and Chief Executive Officer. Michael Levchin, Affirm's Chief Operating Officer and Rob O'Hare, Affirm's Chief Financial Officer. In line with our practice in prior quarters, we will begin with brief opening remarks from Max before proceeding immediately into questions and answers.

今天與我一同主持電話會議的有公司創辦人暨執行長 Max Levchin。Michael Levchin，Affirm 營運長，以及 Rob O'Hare，Affirm 財務長。依照我們過去幾季的慣例，我們將先由 Max 做簡短開場致詞，接著立即進入問答。

Before we begin the call, as a reminder, we will be hosting our 2026 Affirm Investor Forum next week on Tuesday, May 12 from 2 until approximately 5:00 PM Eastern Time. The event will be available to the public via livecast on our Investor Relations website.

在開始電話會議之前，提醒各位，我們將於下週二（5 月 12 日）美東時間下午 2 點至約 5 點舉辦 2026 Affirm 投資人論壇。活動將透過我們投資人關係網站的線上直播對外公開。

We will also publish a replay on our website after the event ends. With that, I'll turn the call over to Max to begin.

活動結束後，我們也會在網站上發布重播。接下來，我把電話交給 Max 開始。

Thank you, Zane. Fiscal Q3 was another one for the record books. Given this streak, one could be forgiven if one thought this is actually pretty easy. That's all because the fantastic Affirm team is starting to make it look that way. It is not, in fact, easy, and we're very proud of this particular quarter.

謝謝你，Zane。財政年度第三季又是寫下紀錄的一季。在這樣的連勝之下，如果有人覺得這其實很容易，也情有可原。這一切都要歸功於出色的 Affirm 團隊，開始讓它看起來就是如此。但事實上並不容易，而我們對這一季的表現感到非常自豪。

Zane said, I look forward to seeing many of you in person at the Industry Forum next week. On that note, back to you, Zane .

如同 Zane 所說，我期待下週在產業論壇上與各位中的許多人面對面交流。話說到這裡，交還給你，Zane。

Thanks, Max. Okay, let's get to your questions. Operator, please begin the Q&A session.

謝謝，Max。好，讓我們開始回答各位的問題。接線員，請開始問答環節。

Jason Kupferberg, Bank of America.

美國銀行（Bank of America）的 Jason Kupferberg。

Hey, guys, this is Kathy Chan on for Jason.

嗨，各位好，我是代替 Jason 的 Kathy Chan。

Great quarter. I just wanted to ask, I guess, first on the private credit side and in general on credit, it seems like the delinquencies were pretty stable this quarter. I guess, is there anything that you're seeing in terms of changes or slowdowns in credit? And, obviously unease in private credit seems to be a theme, but are you guys seeing any issues or changes on the funding side of the business?

很棒的一季。我想先問一下私募信貸那邊，以及更廣泛的信貸狀況：看起來本季逾期率相當穩定。請問你們在信貸方面是否看到任何變化或放緩的跡象？另外，私募信貸市場的不安似乎是一個主題，但你們在業務的資金端是否看到任何問題或變化？

Thank you.

謝謝。

I'll start with the credit side, Michael, we'll pick up the funding side, no, we are not.

我先從信貸面回答，Michael 會接著談資金面；沒有，我們沒有看到。

At this point, I think we've earned the right to say the affirmed consumer, and so these are not common on the universe or even North America or United States consumer, but people that we choose to underwrite and lend to, we are not seeing deterioration, we're not seeing any disturbances in the force.

目前我想我們已經有資格說，「Affirm 的消費者」——也就是我們選擇承作核貸並放款的對象——並不代表整體市場，甚至不代表北美或美國的整體消費者；我們沒有看到惡化，也沒有看到任何擾動。

Which naturally translated to a very stable and pleasant funding environment for us, Michael, can I tell you more?

這也自然轉化為對我們而言非常穩定且令人愉快的資金環境。Michael，你能再補充嗎？

Yes, the funding market broadly remains exceptionally constructive for us. We're kind of out of adjectives to describe just how great the execution has been.

是的，整體資金市場對我們而言仍然非常有利。我們幾乎已經找不到形容詞來描述執行成果到底有多好。

I know a lot of ink is being spilled elsewhere about what's going on in the capital markets, but.

我知道外界對資本市場正在發生的事情有很多討論，但。

From our perspective, we see a market that's very deep, we see sustained and reducing spreads and we see deals with significant oversubscription, along with foreclose partners who are, if anything, still clamoring for a bigger allocation of our portfolio, so we see the market as being very constructive to us and a key part of the reason why our we feel there's so much tailwind in the business.

從我們的角度來看，我們看到的是一個非常深厚的市場、利差持續且進一步收斂，並且交易出現顯著超額認購；同時，我們的融資合作夥伴若說有什麼變化，就是仍然更積極地希望獲得我們投資組合更大的配置。因此我們認為市場對我們非常有利，這也是我們之所以覺得業務有強勁順風的重要原因之一。

Thank you.

謝謝。

Nate Svensson, Deutsche Bank.

德意志銀行（Deutsche Bank）的 Nate Svensson。

Hey, guys. Thanks for the question, and congrats on the record number of Lebowski references in the letter. I did want to add -- yeah. Anyway, sorry, I wanted to ask on the upcoming big nothing was going through the transcript last quarter, Max, you were obviously pretty effusive about all the first order and derivative benefits from that event, things like cardholder sign-offs.

嗨，各位。謝謝讓我提問，也恭喜你們在股東信裡創下《大保齡離奇綁架案》（Lebowski）引用次數的新高。我確實想補充——對，沒錯。總之抱歉，我想問一下即將到來的「big nothing」（我在看上一季逐字稿時注意到的），Max，你上季對那個活動帶來的一階與衍生效益顯然非常熱情，例如持卡人註冊等。

I assume directionally you're expecting a lot of the same things, but on the call last quarter, you also talked about getting better and smarter as you do more of these. So I guess the question is around what ways you think you got better and smarter and maybe what are some of the incremental changes or initiatives we should be on the lookout for for the event next week?

我想方向上你們應該也期待看到很多相同的效果，但你上季也提到，隨著你們做更多這類活動，會變得更好、更聰明。所以我的問題是：你們覺得自己在哪些方面變得更好、更聰明？以及下週活動有哪些增量變化或新措施是我們應該留意的？

I don't really want to reveal all the surprises to be completely H1st, but I appreciate the kind words and we did get smarter.

老實說我不太想把所有驚喜都揭露，但謝謝你的好話，我們確實變得更聰明了。

I think probably if you want to sort of look for breadcrumbs, we got smarter about targeting, that certainly.

我想如果你要找一些線索，我們在目標鎖定上變得更聰明了，這當然。

And it's less about sort of sitting down in a lab somewhere and trying to come up with ideas, much more about looking at the data we gathered and the last but nothing and just using all the same ML/AI techniques we have here to ask the question, what's the least costly highest probability of conversion for anyone SKU, anyone consumer, anyone merchant, et cetera, so it'll get more efficient, that's certainly the case.

這與其說是坐在某個實驗室裡想點子，不如說是看我們從上一次「but nothing」活動蒐集到的資料，並運用我們在這裡同樣的 ML/AI 技術去問：對於任何 SKU、任何消費者、任何商戶等等，什麼是成本最低、轉換機率最高的做法；所以效率會更高，這是肯定的。

We got smarter, kind of qualitatively, I think we really underplayed the event itself in the early hours and kind of had to play a little bit of a catch-up on the marketing side of things, and this won't happen this time. So we will hit the ground running with just promoting it correctly to older IP people, again, maximizing the.

我們在質性層面也變得更聰明：我認為我們在活動剛開始的前幾個小時，對活動本身的宣傳其實有點低估，行銷端不得不稍微追趕一下；這次不會再發生。因此我們會一開始就把宣傳做對，面向既有的 IP 人群，並再次把。

Effective per dollar yield for our merchant partners, so we expect to be even more satisfying to those who are paying for these deals.

為我們商戶夥伴帶來的每一美元投入的有效產出最大化，因此我們預期能讓那些為這些優惠買單的人更加滿意。

Thanks excited to look out for it.

謝謝，期待看到。

Brian Keene with Citi, please proceed with your question.

花旗（Citi）的 Brian Keene，請提出您的問題。

Yeah guys can you give us. Some insights on the ABS market, the deal in March and then the recent deal, what's going on with spreads and demand for you guys?

是的，各位能否給我們。一些關於 ABS 市場的觀察：3 月那筆交易以及最近那筆交易，你們的利差與需求狀況如何？

Yes, thanks for the question. I think we've executed three deals so far this year, two revolving deals in the quarter and then we just priced a static deal. We haven't yet closed on and the trend really across all three is.

好的，謝謝你的問題。我想我們今年迄今已執行三筆交易，本季有兩筆循環型交易，然後我們剛完成一筆靜態交易的定價。我們尚未完成交割，而這三筆交易的整體趨勢確實是。

Incredible depth, lots of oversubscription in these deals and continued and sustained tightening of spreads and a key part of the reason why you see funding costs down on the order of 125 basis points year-on-year. Obviously, benchmark rates are down as part of that, but you're also seeing spreads coming in at the same time.

需求深度驚人，這些交易出現大量超額認購，利差也持續且穩定地收窄；這也是你們看到資金成本年對年下降約125個基點的關鍵原因之一。當然，基準利率下行也是其中一部分，但同時你也看到利差在收斂。

It's just really a reflection of the capital markets demand for our asset and our team's ability to execute.

這其實就是反映資本市場對我們資產的需求，以及我們團隊的執行能力。

Despite quite a bit of economic volatility and headlines out there, we feel like the market is just extremely constructive for earning.

儘管外部經濟波動不小、新聞標題也很多，我們仍覺得市場對獲利表現的態度非常建設性。

Yeah, and it just feels like they're starting to recognize maybe the differences between what you guys, your credit versus others in the short duration and obviously the quick turn.

是的，而且感覺他們開始認知到你們的信用狀況相較於其他人在短天期、以及明顯更快的周轉方面可能存在差異。

But it looks like the market is starting to recognize that, so it's good to see.

但看起來市場正在開始認可這一點，所以很高興看到這樣的發展。

The short duration of our asset is a huge advantage, and it's taken us a long time to earn that. We've also done a really good job, I think, in engaging the investor base and bringing on board over the years a wider set of investors, that's really important for the depth of the market that we play in, it's a key channel for our long-term growth, but also the more broad of the investor base you bring on, the better you could get on pricing.

我們資產的短天期是一個巨大的優勢，而我們花了很長時間才建立起這個優勢。我認為我們在與投資人基礎互動方面也做得非常好，多年來引入了更廣泛的投資人群體；這對我們所參與市場的深度非常重要，是我們長期成長的關鍵管道之一，而且投資人基礎越廣，定價就越有利。

Yeah. Okay, great. Congrats on the results.

是的。好，太好了。恭喜你們的成果。

Thank you.

謝謝。

Our next question comes from Rob Wildhack with Autonomous Research.

下一個問題來自 Autonomous Research 的 Rob Wildhack。

Please proceed with your question. Looks like we don't have them. Let's go to the next question. Our next question comes from Moisha Orenbach with TD Cohen. Please proceed with your question.

請提出您的問題。看起來我們沒有接上他。我們進到下一題。下一個問題來自 TD Cohen 的 Moisha Orenbach。請提出您的問題。

Great, thanks. I noticed that growth in the PayNX is your fastest growing segment now. Are there different either programs or kind of merchant partnerships or anything that is kind of driving that and do we think that's going to continue into fiscal Q4?

很好，謝謝。我注意到 PayNX 的成長現在是你們成長最快的分部。是否有不同的方案、或某種商戶合作夥伴關係、或其他因素在推動這個成長？以及我們是否認為這個趨勢會延續到財年第四季？

Hi, Mosha, this is Rob. We do expect that trend to continue in the fiscal Q4 and I think the answer was largely in your question, we did have one large one one very large program move to having an evergreen 0% via pay-in-for offer, so that definitely drove a bit of the uptick and then we also continue to see most of our pay-in-for, pay-in-ex volume coming from the Shopify program. Which continues to grow nicely. So a bit of sort of business as usual there, and then we did have one large program make a change to their financing program, which we think is real positive.

嗨，Mosha，我是 Rob。我們確實預期這個趨勢會延續到財年第四季。我想答案很大程度就在你的問題裡：我們有一個大型、非常大型的方案改為提供常態性的 0%「pay-in-four」方案，因此確實推升了一部分增幅；此外，我們也持續看到大部分的 pay-in-four、pay-in-ex 交易量來自 Shopify 方案。該方案仍在穩健成長。所以某種程度上算是照常推進；另外我們也有一個大型方案調整了其融資方案，我們認為這是非常正面的。

Ramsey Ellisal, Candor Fitzgerald.

Candor Fitzgerald 的 Ramsey Ellisal。

Hi, this is Ryan on for Ramsey. Thanks for taking your question today. I wanted to ask about the active merchant count, which went up by 44%, accelerating beyond a strong Q2.

嗨，我是 Ryan，代 Ramsey 發問。謝謝今天讓我提問。我想問一下活躍商戶數，成長了 44%，在強勁的第二季之後還加速。

Where or what is the largest opportunity to add more merchants? And how penetrated is the market, not in terms of consumer usage, but in terms of merchant presentment for BNPL?

在哪裡、或是什麼方面，是新增更多商戶的最大機會？以及就 BNPL 的商戶端呈現（不是消費者使用率，而是商戶端是否展示/提供 BNPL）而言，市場滲透率到什麼程度？

I mean, I think in terms of merchant count, we're still looking at some of our largest platform partners as the biggest accelerant to growing our current merchant base. So some of the big PSPs where we have relationships as well as large merchant platforms like Shopify, those have been really additive to our merchant base overall.

我的意思是，就商戶數而言，我們仍把一些最大的平臺合作夥伴視為推動現有商戶基礎成長的最大加速器。因此，一些大型 PSP（支付服務供應商），我們也與其建立合作關係；以及像 Shopify 這樣的大型商戶平臺，這些都對我們整體商戶基礎帶來很大的增量。

I think in terms of presentment. We still feel like it's really early innings in terms of presentment, obviously, we have a brand new program with Intuit and there's lots of optimizations to do within that merchant base, that's an enormous universe of merchants that we're just scratching the surface on and it's very early days in that program.

我想就呈現（presentment）而言。我們仍覺得在呈現方面還在非常早期的階段。很明顯地，我們與 Intuit 有一個全新的方案，而在那個商戶基礎內還有很多可以優化的地方；那是一個龐大的商戶宇宙，我們才剛開始觸及表層，而且那個方案仍在非常早期。

There's countless other examples across our portfolio, but I think in terms of the partnerships that drive some of these big merchant counts, I think we're still we we still have plenty of room to optimize how we show up on the ad merchant site.

在我們的投資組合中還有無數其他例子，但我想就那些能帶來大量商戶數的合作關係而言，我們仍然——我們仍有很大的空間去優化我們在商戶網站上的呈現方式。

Great. Thank you. Looking forward to hearing more at the Investor Forum.

很好。謝謝。期待在投資人論壇聽到更多。

Harry Bartlett, Rothschild & Co Redburn.

Rothschild & Co Redburn 的 Harry Bartlett。

Hey guys, thanks for the question. I just wanted to touch on the kind of agentic code development point in the letter. You cited the kind of noticeable ramp in agentically written code, and it looks like you're kind of doing double the amount of requests that you were doing previously. So I guess, could you try to talk about this broadly in terms of.

各位好，謝謝讓我提問。我想談一下你們在信中提到的「代理式（agentic）程式碼開發」這點。你們提到代理式撰寫的程式碼有明顯加速成長，看起來你們的請求量大概是先前的兩倍。所以我想，能否從更宏觀的角度談談這件事，例如……

How you're thinking maybe costs will develop versus how they developed historically, or whether you kind of see this more as a vehicle for more rapid product development, I guess. Thank you.

你們如何看待成本的演變，相較於歷史上的發展方式？或者你們是否更把它視為加速產品開發的工具？謝謝。

So this is very tempting to turn this into a 15-minute answer in the finer point of software development, which I am personally invested and involved in.

這個問題很容易讓我把回答變成 15 分鐘，深入到軟體開發的細節，而我個人也確實投入並參與其中。

So the shorthand, first of all. And I will rely on Rob in a second to maybe try to even quantify it, but it's unequivocally accretive to the bottom line to use AI the way we are. So this is a net strongly positive, the fact that we are increasing our development velocity is just incredibly strong for our bottom line and then some.

先用簡短版本回答。我等一下也會請 Rob 來補充、甚至嘗試量化，但毫無疑問，以我們目前的方式使用 AI，對損益表是增益的（accretive）。因此這是淨額上非常正面的；我們提升開發速度這件事，對我們的損益表帶來極其強勁的助益，而且不只如此。

The actual mechanics of development using Agentic. The processes and et cetera, we're pretty unique and we feel pretty great about where we are and where we're headed.

至於使用 Agentic 進行開發的實際機制。流程等等，我們相當獨特，我們對目前的位置以及未來方向都感到很有信心。

If you ever read the fine print of the likes of ChatGPT or Gemini, there's a little thing at the bottom that says, AI makes mistakes, basically, you're on your own. We don't really have the luxury of putting that in our code. If we make an underwriting mistake, if our engine somehow treats some consumer unfairly.

如果你讀過像 ChatGPT 或 Gemini 這類產品的細則，底下會有一行小字寫著：AI 會犯錯，基本上風險自負。我們在程式碼上沒有這種奢侈空間。如果我們在核貸/承保上犯錯，如果我們的引擎不小心對某些消費者造成不公平對待。

Or if we're off by any here and there, like none of that is okay. And so as much as we can and do use these tools, there are still many unique to affirm checks and balances and processes that ensure that what we ship is of the same or higher quality than what we did before these tools came around. And we spent quite a lot of time getting there, getting the confidence, testing it. And so the reason we have this sort of a.

或者我們在任何地方有一點點偏差，這些都不被允許。因此，即使我們能、也確實大量使用這些工具，我們仍有許多 Affirm 特有的檢核與制衡與流程，確保我們交付的成果品質與這些工具出現之前相同或更高。我們花了相當多時間才走到這一步：建立信心、進行測試。所以我們之所以會在今年較早的時候出現這種「把瓶塞拔開」的時刻。

Uncork it moments early in the years because we felt that we were ready to mass deploy it internally and have so far been very pleased with what's transpired and we'll definitely do more, I think, I'm sure our engineering leadership is listening/reading these letters and.

是因為我們覺得已經準備好在內部大規模部署；到目前為止，我們對結果非常滿意，而且我想我們一定會做得更多。我相信工程領導團隊正在聽/讀這些信件，而且……

I don't think anybody is begrudging me the right to say, we think we can 10x this productivity further. So it's very early days, we're very excited about it.

我想沒有人會反對我說：我們認為還能把這個生產力再提升 10 倍。所以這仍是非常早期的階段，我們非常興奮。

We have no shortage of things we want to build and therefore humans that are both the creators of ideas, the arbiters of good taste and the ultimate responsibility carriers for this no errors, no.

我們想打造的東西多到做不完，因此人類——既是創意的提出者、品味的裁決者，也是對「零錯誤、零……」負最終責任的人——仍然非常必要。

Fine print, no bugs are still very necessary. So we don't anticipate any sort of a decimation of the engineering team, but we are certainly very excited to give each one of our engineers basically superpowers and I know Rob has any additional cost points to make.

沒有免責細則、沒有 bug，這些仍然非常重要。因此我們不預期工程團隊會出現任何所謂的大幅裁減，但我們確實非常期待能讓每一位工程師基本上都擁有超能力；我知道 Rob 若有任何額外的成本面觀點要補充。

In terms of the cost, I mean, they did obviously show up in the P&L this quarter, they'll continue into Q4 as well. I wouldn't say it's a material impact to the P&L overall, it's sort of.

就成本而言，我是說，它們這一季顯然已經反映在損益表（P&L）上，第四季也會持續。我不會說這對整體損益表有重大影響，大概是這樣。

Very low single-digit millions per quarter in terms of spend. So to Max's point, I mean, we're seeing a lot of efficiency, spending money on developer tools is something we've always done. So we're just thinking about ways to make sure that on a holistic basis, that all-in budget makes sense for us and that we're seeing efficiency and lift from the entire portfolio of tools that we're employing.

以支出來看，每季大約是非常低的個位數百萬美元。所以呼應 Max 的觀點，我是說，我們看到很多效率；在開發者工具上花錢是我們一直以來都會做的事。因此我們正在思考如何確保從整體角度來看，這個全包式預算對我們是合理的，而且我們能從正在採用的整套工具組合中看到效率提升與效益拉升。

Super helpful. Thank you.

非常有幫助。謝謝。

Rob Wildtack, Autonomous Research.

Rob Wildtack，Autonomous Research。

Hey, can you guys hear me this time?

嘿，這次你們聽得到我嗎？

Yeah.

聽得到。

Great.

太好了。

Cool. I wanted to ask about the different Affirm surfaces, namely the app. You've highlighted in the past, the GMV lift there, that's intuitive. I'm curious, though, where.

酷。我想問一下 Affirm 不同的觸點（surfaces），尤其是 App。你們過去強調過那裡的 GMV 拉升，這很直觀。不過我好奇的是，

Consumer awarenesses on that? Like, are consumers still opening up there for that to make a payment and going, Oh, lucky me, here's this great offer, or have they become more attuned to the fact that this is a place where they can start looking for products and shopping via the app first?

消費者對那個的認知程度如何？例如，消費者是不是仍然只是為了付款而打開 App，然後才發現「喔，真幸運，這裡有個很棒的優惠」；還是他們已經更習慣把這裡當成一個可以先開始找商品、並透過 App 先行購物的地方？

Alright, you're front-running, like, half of my speech next week, so I'm not going to answer it. I'm not going to gratify this one with it. No, I'm kidding.

好吧，你這等於是提前問了我下週演講大概一半的內容，所以我不回答。我不會滿足你這題。沒有啦，我開玩笑的。

The short answer is. It is trending in the very direction you described. So Affirm app was deliberately designed to make sure that there is more value to be had there than just sort of in passing, setting up or checking up on your auto pay. So all the different components in the first three and the fifth tabs of the app are all designed to create engagement, to expose consumers to.

簡短的答案是：確實正朝你描述的方向發展。所以 Affirm App 是刻意設計成，讓使用者在那裡能獲得的價值不只是順手設定或查看自動扣款（auto pay）。因此 App 的前三個分頁以及第五個分頁中的各個組件，都是為了創造互動、讓消費者接觸到

Various merchant promotions. It is not an accident that the Big Nothing days are basically organized around the app. We're trying to teach consumers that this is where you go. There's always 0% offers in the app. The B&D is just a nexus of many of them concurrently, but at any given time, there's a lot to begin with. We have a really nice number. I won't spoil the eventual report on that one, number of searches that consumers run in our app. We're watching that grow, and it's all in the service of teaching consumers that the best experience of a firm is the app plus the card. So I'm deliberately obscuring some of the maybe more interesting portions that you'll have to wait six more days before we start really.

各種商家促銷。Big Nothing Days 基本上是圍繞著 App 來組織，這並非偶然。我們在嘗試教會消費者：這就是你該來的地方。App 裡一直都有 0% 的優惠。B&D 只是把其中很多優惠在同一時間集中呈現的樞紐，但其實任何時點本來就有很多優惠。我們有一個很不錯的數字。我先不劇透之後會公布的報告：消費者在我們 App 裡進行的搜尋次數。我們看到這個數字在成長，而這一切都是為了教會消費者：Affirm 的最佳體驗是 App 加上卡片（card）。所以我刻意把一些可能更有趣的部分講得比較含糊，你得再等六天，我們才會真的開始

Doing some fun reveals. But directionally, you're exactly right. We are motivated to make the app experience excellent both as a product and as a financial service to our consumers. And there's a bunch of things to show and many more that we're probably not going to show up necessarily next week. But it shapes the roadmap for years in our minds.

做一些有趣的揭露。但方向上，你完全說對了。我們的動機是把 App 體驗做得很棒——不論作為一個產品，或作為提供給消費者的金融服務。我們有一堆東西可以展示，還有更多可能不一定會在下週展示。但在我們心中，這會形塑未來多年的產品路線圖。

Got it. And then a quick one for Rob, if I may.

了解。然後如果可以的話，我想再快速問 Rob 一題。

It was up quarter over quarter. Could you just call out the drivers there?

它季對季是上升的。你能不能點出背後的驅動因素？

I'm sorry, could you repeat the question?

不好意思，你可以重複一下問題嗎？

The allowance was up quarter over quarter. Just wondering the drivers there.

備抵（allowance）季對季上升。想了解一下背後的驅動因素。

Yeah, I mean, it's, of course, partly a function of just seasonality.

對，我是說，當然部分原因就是季節性。

The allowance rate typically is elevated in Q3, just given we have... The sequential downtick from holiday volumes in Q2 down to a lower base in fiscal Q3. That's part of it.

備抵率通常在第三季會偏高，主要是因為我們有……第二季假期交易量較高，到了會計年度第三季會依序回落到較低的基期。這是其中一部分。

Another driver, which I think Max called out in his portion of the letter, which is that we did see elevated prepayments on the platform. It's a little bit of a counterintuitive point, but that's a really positive credit signal, and it has the effect of reducing.

另一個驅動因素，我想 Max 在信函的部分也有提到，就是我們確實看到平台上的提前還款（prepayments）偏高。這點有點反直覺，但其實是非常正面的信用訊號，而且它會帶來降低

The overall loan balance, which obviously, the good loans are being paid off early, and so you're left with more delinquencies off of a lower base. So that was that that contributed to a higher allowance rate all in, but we think it's a really positive credit signal across our users at large. So those are sort of the two biggest drivers, seasonality and then a bit of favorable prepayments from tax season.

整體貸款餘額的效果；顯然，好的貸款被提前還清了，因此你會在較低的基期上留下更多逾期（delinquencies）。所以這也推升了整體備抵率，但我們認為這對我們整體用戶而言是非常正面的信用訊號。因此最大的兩個驅動因素大概就是：季節性，以及稅季帶來的一些有利的提前還款。

This particular tax season was understandably refund maxed, is that what the kids say?

這個稅季顯然是「退稅拉滿」（refund maxed），孩子們是這樣說的嗎？

Dan Dolev, Mizuho.

Dan Dolev，瑞穗（Mizuho）。

Hey, guys. his bill is very impressive results. I just wanted to ask you, Max, can you hear me well?

嘿，各位，這些真的是非常亮眼的成果。我只是想問一下，Max，你聽得清楚嗎？

Yeah, sorry, we're just, we're silent waiting.

聽得到，抱歉，我們只是安靜地在等。

Sorry, I just wanted to ask really quickly, some of your competitors have done some significant layoffs because of AI. I just wanted to know what, the official affirmed stance is on this topic.

抱歉，我想很快問一下，你們的一些競爭對手因為 AI 做了相當大規模的裁員。我想知道，Affirm 對這個議題的官方立場是什麼。

Thank you so much.

非常感謝。

We are not planning AI-related layoffs. Full stop.

我們不打算進行與 AI 相關的裁員。就這樣。

I don't mean to belittle anyone out there making the right or what they believe to be the right decisions for their companies.

我無意貶低任何正在做出他們認為對公司正確決策的人。

So strictly a firm-centric view of the world from us.

所以我們只是從 Affirm 自身為中心的角度來看這件事。

If you look at our revenue per employee, it is already.

如果你看我們的人均營收（revenue per employee），它已經

Hanging out in like NVIDIA territory. I don't remember the last time I looked at it, but it's very high number of dollars per employee. So we today operate as a very lean machine. If you look at our overall headcount, it hasn't grown very much. If you look at the revenue per employee, you'll see that we're just highly efficient. If you look at overall operating leverage, it's done really well.

差不多在 NVIDIA 那個等級了。我不記得我上次看是什麼時候，但每位員工對應的美元數非常高。所以我們現在的營運本來就是一台非常精實的機器。如果你看我們的總人數，成長其實不大。如果你看人均營收，你會看到我們就是高度有效率。如果你看整體營運槓桿，也表現得非常好。

So long before AI tools came along, we had tooled ourselves up to be. Very efficient. These tools are giving us rocket boosters, wings, whatever metaphor you want, and we're very happy for it, but at least for now, and as far as the eye can see anyway, it is just a thing we're going to keep using to ship more. The list of things we want to ship is very long, and until very recently.

所以在 AI 工具出現之前很久，我們就已經把自己工具化到能夠非常高效。這些工具給了我們火箭推進器、翅膀——隨便你想用什麼比喻——我們也很樂見其成；但至少目前，以及可預見的未來，它只是讓我們能持續用來交付更多成果的工具。我們想要交付的清單非常長，而直到不久之前，

A lot of our conversations were, well, we don't know when we're going to prioritize this thing that you want back because we have so much more to build.

我們很多對話都是：嗯，我們不知道什麼時候才能把你想要的那個東西排進優先順序，因為我們還有太多要做的。

And blissfully, these conversations are now like, well, we can just have a hackathon and 40 hours later, we'll have a working prototype.

而令人欣慰的是，現在這些對話變成：嗯，我們可以辦個黑客松（hackathon），40 小時後就能有一個可運作的原型。

We just wrapped up one here where our product team literally delivered dozens of shippable features, which is just impossible to imagine 12 months ago. So we need all the people we got.

我們剛在這裡結束一場，我們的產品團隊真的交付了數十個可上線（shippable）的功能，這在 12 個月前根本難以想像。所以我們需要我們現有的所有人。

We think we have fantastic people and we like them all.

我們認為我們擁有非常棒的人才，而且我們都很喜歡他們。

Dan Perlin, RBC Capital Markets.

Dan Perlin，RBC Capital Markets。

Thanks. Good evening. I'm wondering, can you just speak, I think, maybe holistically to your expansion plans outside of Growth America? I know you talk about it a little bit embedded in the guidance here for the product and go-to-market initiatives and not being material in '26, but I'm just trying to think.

謝謝。晚安。我想請問，你們能否從更整體的角度談談在 Growth America 以外的擴張計畫？我知道你們在這份指引中，針對產品與 go-to-market（上市推廣）計畫有稍微提到，而且在 '26 年不會是重大影響，但我只是想釐清。

Completing in terms of investments as we start to think about next year.

當我們開始思考明年的投資時，如何把它串起來理解。

And also, I guess, in the context, although it's a little bit of a different driver, but the RLTC margins continue to run above long-term targets. I'm just wondering, as you go into the international markets, how that might impact it. Thank you.

另外，我想在這個脈絡下，雖然驅動因素有點不同，但 RLTC 利潤率仍持續高於長期目標。我想請問，當你們進入國際市場時，這可能會如何影響它。謝謝。

Sure, I'll take him in order. I think we're going to spend a bit of time talking about our expansion plans with a bit more specificity in terms of markets. So I'll leave the deep dive on the international markets for next week's investor forum, if that's okay.

好的，我會依序回答。我想我們會花一些時間，以更具體的方式（就市場而言）來談我們的擴張計畫。所以如果可以的話，關於國際市場的深入說明，我先留到下週的投資人論壇。

In terms of the investment portfolio for those launches, some of that work is already underway today. So that's definitely been an area that we've been testing in ahead of those markets coming live.

就這些上線（launch）所需的投資組合而言，其中一些工作今天其實已經在進行。因此，這確實是我們在這些市場正式上線之前就已經在測試的領域。

And as you've seen from the results, we've been able to drive really nice operating leverage despite that investment. So I think we would expect to do more of the same in fiscal '27, but I'll stop short of giving any sort of outlook or guidance for '27 today.

而且從你們看到的結果來看，即使有這些投資，我們仍能帶動相當不錯的營運槓桿。所以我想我們預期在 2027 財年會延續同樣的作法，但我今天不會進一步提供任何對 '27 的展望或指引。

In terms of unit level economics, I mean, I think as we ramp in new countries, we would expect potentially that there is a bit of an investment period where we're meeting new consumers and coming down the curve in terms of underwriting prowess. So there could be a small drag on revenue less transaction costs as we enter these new markets, but given the size of the US and Canadian businesses today, we think any. There would be pretty minimal.

就單位經濟（unit level economics）而言，我的意思是，我想當我們在新國家加速擴張時，我們可能會有一段投資期：我們在接觸新消費者，同時在承保能力（underwriting prowess）的學習曲線上逐步改善。因此，當我們進入這些新市場時，收入扣除交易成本（revenue less transaction costs）可能會有些微拖累，但考量到目前美國與加拿大業務的規模，我們認為任何影響。都會相當有限。

That's great. Thank you so much.

太好了。非常感謝。

Andrew Bosch, BMO Capital Markets.

Andrew Bosch，BMO Capital Markets。

Hey, thanks for taking the question. I wanted to talk about a firm card and the level of ads you can see to stack up here.

嗨，謝謝讓我提問。我想談談 Affirm 卡，以及你們在這裡看到的新增量級。

The 700,000 users is pretty impressive, especially coming off of the 900,000 last quarter. So is there anything that's working differently or stronger than it has been in the past as far as. Card customer acquisition goes? And then my follow-up would be, now that we're doubled the base at 4.4, are you starting to see more and more benefits of scale coming through the Pike?

70 萬名用戶相當令人印象深刻，尤其是在上季 90 萬的基礎之後。所以就。卡片客戶獲取而言，是否有任何做法與過去相比更有效或更強？接著我的追問是，現在我們的基礎已經翻倍到 4.4（百萬），你們是否開始看到更多規模效益逐步顯現？

I think the first part, there's a long list of things we have done and continue to do to just.

我想第一部分，我們已經做了、也持續在做的事情很多，目的就是。

Increase adoption. We've said it before and remains true that card are by far fastest growing and also our most profitable product. So there's absolutely no reason not to try to grow it. That said, we have not been in any way fueled choosing the growth. It's natural. There's not a.

提高採用率。我們之前說過，而且現在仍然如此：卡片是迄今成長最快、同時也是我們最有利潤的產品。所以絕對沒有理由不去努力把它做大。話雖如此，我們並沒有以任何方式去「硬推」成長。這是自然成長。並沒有。

Secret game somewhere being played or anything like that. So it's growing about as fast as we can make it grow without tilting anything in a weird direction. It still primarily remains a repeat product. We've never tried to advertise it or promote it outside of existing Affirm user base. It's still.

什麼秘密手段在運作或類似的事情。所以它的成長速度，大致就是在不把任何事情往奇怪方向傾斜的前提下，我們所能做到的最快速度。它仍主要是一個回購／重複使用的產品。我們從未嘗試在既有 Affirm 用戶群之外做廣告或推廣。它仍然。

Roughly in the 20% of the actives plus or minus. So we have a lot of road to cover before we start asking where can we get more cardholders. They are our favorite users in a sense that they transact most frequently, they tend to be least lossy just because we get to know them a lot quicker, a lot more frequently and so on. So it's all goodness, nothing sort of nothing hidden or regrettable there.

大約占活躍用戶的 20% 左右（上下浮動）。所以在我們開始問「還能從哪裡獲得更多持卡人」之前，還有很長的路要走。某種意義上，他們是我們最喜歡的用戶：交易最頻繁，而且通常損失率最低，因為我們能更快、更頻繁地了解他們，等等。所以這都是好事，沒有任何隱藏或令人後悔的地方。

And the economies of scale, I think to be completely H1st, I haven't thought it through very carefully if we're finding benefits of scale that are sort of truly unique. The one thing that is true in a software development context, which is a little maybe a glimpse at the resource allocation.

至於規模經濟，我想完全坦白說，我沒有非常仔細地想過，我們是否正在獲得那種真正獨特的規模效益。但有一點在軟體開發的情境下確實成立，這也許能讓大家稍微看到資源配置的邏輯。

Your fastest growing product is typically your smallest product. And so no matter how much you love your youngest child, you can't really allocate the greatest number of resources towards it because it's just too small. The card is now in the billions of dollars of volume. It is no longer a small product, which means that it deserves and gets the software engineering attention and the risk attention. These are all the various pieces that we would perhaps.

你成長最快的產品通常也是你最小的產品。所以不管你多愛你最小的孩子，你也不可能把最多的資源都分配給他，因為它實在太小。而卡片現在的交易量已達數十億美元。它不再是小產品，這意味著它值得、也確實獲得更多軟體工程與風險管理方面的關注。這些都是我們可能會。

Wonder if they're worth allocating from other parts. And so you can expect it to get more features sooner, more, maybe even more growth sooner, although that is not a forecast or forward-looking statement, is just sort of hitting its stride in every dimension, including internal resource allocation.

思考是否值得從其他部分挪用資源來投入的各個面向。因此你可以預期它會更快獲得更多功能、更快得到更多（資源），甚至可能更快帶來更多成長——但這不是預測或前瞻性陳述——而是說它在各個維度上都正在進入成熟期，包括內部資源配置。

Got it. Thank you.

了解。謝謝。

Matt O'Neill, Bank of America.

Matt O'Neill，Bank of America。

Yeah, hi, thanks for the question. Being cognizant of the upcoming forum, I'll try not to get too long-term focused, but Maybe we could talk a little bit about the card and what that sort of pretends to the longer-term, banking idea. Obviously, there's application put in this past quarter.

是的，嗨，謝謝讓我提問。考量到即將到來的論壇，我會盡量不把問題拉得太長期，但也許我們可以談談卡片，以及它對較長期的「銀行」構想意味著什麼。顯然，你們在上一季已經提交了申請。

Respecting that, I expect a lot more of this next week. Are there any sort of points you can kind of hint at as far as the focus around things like, sort of pay now, direct deposit, the dynamics to contemplate as you guys proceed down that path?

在尊重這點的前提下，我預期下週會有更多內容。就你們接下來推進的方向而言，是否能先透露一些重點，例如 pay now（即時付款）、直接存款（direct deposit）之類的聚焦方向，以及你們在走這條路時需要考量的動態？

It's definitely worth separating the bank application. In the product roadmap, like the movement completely different time horizons. We are excited to continue the conversation with our regulatory friends and it may take a little time, it may take a long time. We don't know and that's part of the process and certainly have nothing to share on that front at the moment.

確實值得把銀行申請（bank application）和產品路線圖（product roadmap）分開來看，因為它們完全是不同的時間尺度。我們很期待能持續與監管單位的夥伴對話，而這可能需要一些時間，也可能需要很長時間。我們不知道，這也是流程的一部分；目前我們當然也沒有任何可以分享的內容。

On the feature set of where the product roadmap is headed.

至於產品路線圖將走向何種功能組合（feature set）。

We'll cover some of that next week, so I definitely don't want to take Vishal's talking points away from him.

我們下週會涵蓋其中一些內容，所以我也不想把 Vishal 的重點先講走。

But we definitely have aspirations in a variety of consumer financial services. For the longest time, we've said we see ourselves our mission is pretty clearly we're trying to build H1st financial products to improve lives. We're not trying to build short-term loans at the point of sale to improve lives. And so there's plenty of opportunity, we think, to right the wrongs of some of the.

但我們確實在多種消費者金融服務上有抱負。長久以來我們一直說，我們的使命很清楚：我們想打造能改善生活的 H1st 金融產品。我們不是想打造在結帳點提供的短期貸款來改善生活。因此，我們認為有很多機會去矯正這個產業中一些。

Poorly made products in this industry and also just invent our owns and do interesting things there. So I'm giving you a little bit of a word salad here, but we have aspirations in just about everything that you can possibly imagine in Consumer Financial Services.

設計不佳的產品，也能發明我們自己的產品並做一些有趣的事情。所以我這裡可能講得有點雜，但我們在幾乎所有你能想像得到的消費者金融服務領域都有抱負。

More to come.

後續還會有更多。

I'm also cognizant that sometimes we announce things and take three years to get them to the point where they're good enough to launch. And so I'm extra cautious not to say, oh, yeah, here's something we're going to do. And we'll definitely do it, but it may take us a year or two.

我也意識到，有時候我們宣布一些事情，卻要花三年才把它做到足夠好、可以正式推出。所以我會格外謹慎，不會說，喔對，這是我們要做的事。而且我們一定會做，但可能需要一到兩年的時間。

Thanks. I appreciate that and the delineation between the regulatory process and the business build out. I'll hop back in.

謝謝。我很感謝你這麼說，也感謝你把監管流程與業務建置之間的界線講清楚。我先回到隊列裡。

Darren Peller, Wolfe Research.

Darren Peller，Wolfe Research。

Hey, guys, thanks. Can we just touch base again on the strength of the GMV side for a moment and the sustainability? Number one is just making sure there was nothing unusual or unsustainable about the quarter. But.

嗨，各位，謝謝。我們能不能再回頭談一下 GMV 端的強勁表現，以及其可持續性？第一點是想確認，本季是否沒有任何不尋常或不可持續的因素。不過。

Which I'm sure you'll say probably not. I guess I'm trying to figure out what's to stop this type of growth rate from being sustainable from your perspective. And then more importantly, I mean and on that note, we've heard a lot from competitors about trying to do more in the space, but it seems to have very little impact on your growth potential. So I mean anything you're seeing from a competitive landscape that's changed worth sharing over the past quarter or two would be great.

我相信你們大概會說沒有。我想弄清楚的是，從你們的角度看，是什麼會阻止這種成長率持續下去。然後更重要的是，就這點而言，我們聽到很多競爭對手說想在這個領域做更多，但看起來對你們的成長潛力影響很小。所以，過去一兩季你們在競爭格局上看到有哪些值得分享的變化嗎？

I'll start, I suspect, Michael, who is doing a small victory dance right next to me, will have something to add. But you're totally right, we don't see a reason and again, I don't want to front run our promises and storytelling next week, but no, there's nothing unnatural about this one. We move up and down with the economy, but.

我先開始。我猜 Michael（就在我旁邊小小地跳著勝利之舞）也會補充一些。但你說得完全沒錯，我們看不出有什麼理由——而且我也不想在下週我們的承諾與敘事之前搶先透露——但不，這一季沒有任何不自然的地方。我們會隨著經濟上下波動，但。

We are, we've hit a product market fit quite some time ago, but we're still tiny relative to the massive payment volume in the US Alone and e-commerce alone. We're really small, so taking share, it's not that hard yet. In terms of competitive, and I really will let my colleagues speak, it's hard to tell.

我們很久以前就達到產品市場契合（product-market fit）了，但相對於僅美國、僅電商的龐大支付量，我們仍然非常小。我們真的很小，所以要拿市占，現在還不算太難。至於競爭，我真的會讓我的同事來說，這很難判斷。

One of our long-tenured executives here has this line, they're never retreating, they're just reloading, so it's a fantastic space, the NPL overall is just a very compelling product, we don't have a monopoly in the idea and so it's always going to be a competitive space, there are really no monopolies in payments to begin with, so it's just. Not a thing we can expect to eventually own entirely to ourselves. We do, in my anyway, very biased view of the world, are the best at it.

我們這裡有位資深主管常說一句話：他們從不撤退，只是在重新裝填彈藥。這是一個很棒的領域，整體 NPL（先買後付）是一個非常有吸引力的產品，我們並不壟斷這個概念，所以它永遠都會是競爭激烈的市場；而且支付本來就幾乎沒有壟斷可言，所以這就只是。不是我們可以期待最終完全由我們獨占的事情。但以我（非常偏頗）的世界觀來看，我們確實是做得最好的。

We do have some really great economies of scale. Capital markets are now very familiar with our product. They understand exactly what we manufacture. They understand that we are entirely non-compromising in our view of what is and isn't fit to sell into forward flow or securitization. And so we have. A lot of trust with our counterparties, and we tend to take that very seriously.

我們確實有一些非常好的規模經濟。資本市場現在對我們的產品非常熟悉。他們很清楚我們「製造」的是什麼。他們也理解，我們對於哪些資產適合、哪些不適合拿去做 forward flow 或證券化的判斷，完全不妥協。因此我們擁有。與交易對手之間很高的信任，而我們也非常認真地看待這件事。

On the consumer side, we're not really heavy advertisers, certainly not heavy brand advertisers, and yet we do have a brand. We just ran a bunch of studies that show that we're really well recognized, people trust us, they understand after 15 years that when we say no late fees, we mean it, never charged a penny, don't have a plan to ever charge a penny of late fees. So that's been slowly but surely building up in our favor.

在消費者端，我們並不是重度投放廣告的公司，當然也不是重度品牌廣告投放者，但我們確實有品牌。我們剛做了一系列研究，顯示我們的辨識度很高，人們信任我們；15 年來大家也理解，當我們說不收逾期費（late fees），就是說到做到——我們從未收過一分錢，也沒有任何計畫要收任何逾期費。所以這些優勢一直在緩慢但確實地累積到我們這邊。

And then. Just in a pure sort of competitive front, I think speaking of maybe the most important and least understood advantage that we have been at it for a very long time, we have built some very sophisticated underwriting capabilities, we'll definitely talk a lot about that next week, so I'm going to bite my tongue right there, but we have some very cool stuff that we've done, not just recently, but over the years in underwriting.

然後。就純粹競爭面來說，我認為也許最重要、但最不被理解的一個優勢是：我們做這件事做了很久。我們建立了非常成熟的承保（underwriting）能力；下週我們一定會談很多，所以我先點到為止，但我們在承保上做了很多很酷的事情，不只是最近，而是多年來累積的成果。

Great percentage, maybe the totality of our competitors that have raised their hand and said, sure, underwriting is not that hard, we can do it, one by one found out that it is. It is actually very difficult.

很大一部分——也許是我們競爭對手的整體——曾經舉手說，當然，承保沒那麼難，我們也做得到；但他們一個接一個發現，確實是。它其實非常困難。

And by showing a result, we may fool the world by just print a good result quarter after quarter after quarter and we get yelled at, gosh, why don't you guys already admit that it's always going to be over 4%? It is a difficult balance to strike to print these unit economics day in and day out and. All of that or a lot of that comes from our AI team and the research that they do, and it's hard work. So I think we make it very easy to believe that just isn't that hard, and it really is. And the longer the show goes, the more it becomes obvious that we are pretty good at math and are very serious about it, and the rest of the competitors are not.

而且我們透過持續交出成果，可能會讓外界誤以為：只要一季又一季又一季地印出漂亮的結果就好。然後大家還會罵我們：天啊，你們為什麼不乾脆承認它永遠都會超過 4%？要在日復一日地維持這些單位經濟（unit economics）並。做到這一切或其中很大一部分，都來自我們的 AI 團隊以及他們所做的研究，而那是辛苦的工作。所以我覺得我們把事情做得太像很容易，讓人相信這其實不難，但它真的很難。而且時間越久，越能看出我們在數學上很強、也非常認真；其他競爭對手則不是。

And then, Darren, just to your question on the growth rate. Obviously, we're really happy with the growth rate that we posted in Q3 and we're incrementally more positive on the Q4 growth rate in the updated guide, I will just remind everyone that we did sunset a top three merchant in Q1 of this year, so we are comping against there there is a difficult comp. In the prior year period, and that comp did step up a little bit from Q3 to Q4. So it's a little bit more of a headwind to growth.

然後，Darren，回到你關於成長率的問題。顯然，我們對第三季公布的成長率非常滿意；在更新後的指引中，我們對第四季成長率也更偏正面一些。我只提醒大家，我們在今年第一季終止了一家前三大商戶的合作，所以我們在做同比時，會對上。去年同期一個比較困難的比較基期，而這個比較基期從第三季到第四季又稍微變得更嚴峻一些。因此對成長會有一點更大的逆風。

We're talking sort of a few points of growth in terms of headwind. And as we get into fiscal '27.

我們說的逆風大概是幾個百分點的成長影響。而當我們進入 2027 財年。

The comps get a lot easier for us. It's more of a same-store comp for us. So we do -- we don't think that the Q4 growth rate will necessarily be a ceiling as we look ahead into fiscal '27.

對我們來說同比會容易很多。會更像是同店（same-store）的比較。所以我們——我們不認為，展望 2027 財年時，第四季的成長率必然會是天花板。

Okay. Very helpful, guys. Thank you.

好的。非常有幫助，各位。謝謝。

Connor Allen, JPMorgan.

Connor Allen，摩根大通（JPMorgan）。

Hi. Thanks for taking my question. I wanted to ask about transactions per active. It's been growing above 20% for quite a while. And I was curious, maybe this quarter or somewhat recently, you could just kind of decompose that a little bit for us, and maybe it's a bit duplicative with some of your other comments about just broader engagement. But I don't know, anything you could share around cohorts and their behaviors around this engagement or how broad versus targeted the improving engagement is, maybe just a double-click, deeper dive on the engagement side.

嗨。謝謝讓我提問。我想問一下每位活躍用戶的交易次數（transactions per active）。它已經有一段時間維持在 20% 以上的成長。我很好奇，也許就本季或最近，你們能否幫我們把它稍微拆解一下；可能跟你們其他關於整體參與度的評論有點重複。但我想知道，你們能否分享一些關於不同 cohort 及其參與行為的觀察，或是參與度提升是更廣泛的、還是更偏向特定族群的——也就是在參與度這一塊再 double-click、更深入一點。

All of the above. It's really good. There's definitely a few good lines on that one next week, so I won't steal that thunderer. But this is actually a really good example of network effect, so I'll give you like a super brief preview. So even if we did absolutely nothing to improve products usability and just.

以上皆是。真的很不錯。下週在這點上肯定會有幾句很精彩的說法，所以我就不搶那個風頭了。但這其實是一個網路效應（network effect）的很好例子，我先給你一個非常簡短的預告。所以即使我們完全不做任何事情來改善產品可用性，只是。

Converted more and more consumers to cardholders, you would see increase of transactions per user with absolutely no effort on our part beyond that. But we don't just do that. We also sign new merchants, which means that we are visible with our logo, at the very least at checkout, but also in other forms of merchant communications, including but not limited to their own advertising.

把越來越多的消費者轉換成持卡人，你也會看到每位用戶的交易次數上升，而我們除了那件事之外幾乎不需要額外努力。但我們不只做那個。我們也會簽下新商戶，這代表至少在結帳頁面會看到我們的標誌；而且也會出現在商戶其他形式的溝通中，包括但不限於他們自己的廣告。

So that creates. And another push towards the flywheel where more consumers are aware of us, more consumers know that we are in fact real, that our promises of no late fees, et cetera, are shown up in more and more places, that pushes consumer flywheel along, more consumers sign up, more consumer trust is available, consumers get to their second or third loan quicker just because of more checkout.

因此就形成了。而且也進一步推動飛輪效應：更多消費者認識我們、更多消費者知道我們確實是真實可信的，我們「不收逾期費」等承諾在越來越多的地方被看見，這會推動消費者飛輪運轉，更多消費者註冊，消費者信任度提升；也因為有更多結帳場景，消費者更快就能拿到第二筆或第三筆貸款。

Counters available, which makes them eligible for the card, which we, of course, let them know as soon as they qualify, which drives the flywheel of cards. And so the acceleration across the usage, aka transactions per user in the business, is a function of both the merchant side of the network increasing through sales and the consumer side of the network increasing through.

可用的結帳櫃檯（checkout counters）增加，讓他們符合申請卡片的資格；當然，一旦他們達到資格，我們就會立即通知他們，進而帶動卡片的飛輪。因此，整體使用加速（也就是業務中每位用戶的交易筆數）同時取決於：商戶端網路透過銷售擴張，以及消費者端網路透過……的擴張。

Sign-ups because of the increased merchant reach, but also sign-ups from the occasional use of the card, which is much more frequently used. So those are just the two factors at Investor Forum will really break it down into all the various drivers.

由於商戶觸及面提升而帶來的註冊成長，同時也包括因偶爾使用卡片而帶來的註冊，而卡片的使用頻率要高得多。所以這就是兩個因素；在投資人論壇上我們會把各種驅動因素都拆解得更清楚。

James Faucette, Morgan Stanley.

James Faucette，摩根士丹利。

Hey, good afternoon, everybody. Just wanted to ask on -- this goes back a little bit to RLTC, and Max, I appreciate it's hard to -- it may seem easy to stay above where your targets are, but it's really hard. But along those lines, just trying to think about how the 0% APR mix ceiling can affect that, and just how you're thinking about how high that can go. You call out that that typically has lower RLTC margin.

各位下午好。我想問一個問題——這要稍微回到 RLTC；Max，我理解要——看起來維持在目標之上好像很容易，但其實很難。沿著這個思路，我想了解 0% APR 的組合占比上限會如何影響這件事，以及你們怎麼看它最高可能到哪裡。你們提到這通常會帶來較低的 RLTC 利潤率。

And along those lines, I guess I wonder if as merchants become more informed and see the benefits of working with the firm for 0%, if you can actually close that 0% RLTC margin gap with the rest of the business?

同樣沿著這個思路，我在想，隨著商戶更了解、也看到與公司合作提供 0% 的好處，你們是否能把 0% 的 RLTC 利潤率差距縮小到與其他業務一致？

It's a great question, actually. In reverse order. I think it's another example of the network effects playing out. So to answer it directly, I think yes, I think overtime time more merchants and part of why we staged these big nothing events and we'll do more.

這其實是個很好的問題。我先倒過來回答。我認為這是網路效應發揮作用的另一個例子。所以直接回答：我認為可以；我想隨著時間推移會有更多商戶——這也是我們為什麼舉辦這些大型「Nothing」活動，而且之後還會辦更多。

Is because they act as teaching aids, if you will, sort of the white papers write themselves, if you fund these 0% deals, you will sell more and you'll sell more predictably and there will not be a pull forward, these are actual sales events that work. And so all of that adds up to a product that we think is increasing in value in part because the size of our consumer audience is increasing as well and we're able to sort of.

因為它們可以作為教學輔助；可以說白皮書會自己寫出來：如果你資助這些 0% 的交易，你會賣得更多、而且銷售更可預期，也不會出現「提前透支需求（pull forward）」；這些是真正有效的促銷銷售活動。因此，這一切加總起來，讓我們認為這個產品的價值正在提升，部分原因也在於我們的消費者受眾規模也在增加，而我們也能夠某種程度上……

Shine a concentrated spotlight onto a merchant that wants to fund these deals, etc. And we have a lot of really interesting stuff in works for that.

把聚光燈集中打在想要資助這些方案的商戶身上等等。我們也正在推進一些非常有意思的事情。

I've been monopolizing the airwaves, so I'll let Rob or Michael answer the economic breakdown. But it does remain true that 0% is slightly lighter on the RLTC basis. We are not fussed by that.

我一直佔用發言時間，所以我讓 Rob 或 Michael 來回答經濟拆解。但確實仍然是：在 RLTC 基礎上，0% 會稍微輕一些。我們並不為此困擾。

Yeah, no, I mean, I think we love all our loan products equally. There's a lot to like about our interest-bearing loans, but to your point, James, I mean, there is slightly less revenue content today and I think as we look ahead as well within the 0% program, but the good news is there's less in terms of credit costs typically as well. So we really like that trade and we think it's a really good complement to the strong and profitable and high-growth interest-bearing book that we have as well.

對，我的意思是，我們同樣喜歡我們所有的貸款產品。我們的計息貸款有很多優點，但如你所說，James，目前在收入含量上確實略低；而且我們往前看，在 0% 計畫內也是如此。不過好消息是，通常在信用成本方面也會更低。所以我們很喜歡這種取捨，我們也認為它能很好地補強我們既強勁、又具獲利性、且高成長的計息貸款組合。

Yeah, I mean, again, we're really here to drive conversion for merchants, and we think 0% should be an ingredient in every merchant financing program. And as we look at the portfolio today, our largest programs are all utilizing 0%, which we think is a really good sign.

對，我的意思是，我們在這裡的核心是為商戶提升轉換率，而我們認為 0% 應該是每個商戶融資方案中的一個要素。而且就目前的投資組合來看，我們最大的幾個方案都在使用 0%，我們認為這是非常好的訊號。

We're definitely leaning into it within the Affirm Card as well on our own surfaces. So we're doing everything we can to get as much out there and to continue to push that product.

我們也確實在 Affirm Card 上、在我們自有的觸點上更積極推進。所以我們正在盡一切努力把這個產品推得更廣、並持續推動它。

David Sharp, Citizens Capital Markets.

David Sharp，Citizens Capital Markets。

Hey, good afternoon. This is Jack on for David. Congratulations on another great service. I wanted to dig in a little bit on the card side of the stuff.

各位下午好。我是 Jack，代 David 提問。恭喜又交出一季很棒的表現。我想更深入問一下卡片端的部分。

Sorry, I don't know if you guys can hear me a little bit of an echo. There. But, yeah, I wanted to kind of see what the profile of the average card user is. Obviously, I think there was a kind of a medium term target of $10 billion of GMV and about 7.5 million active card users. And, we're kind of approaching that level at about 60% of the card user level. So, yeah, kind of wondering if we can get an update on kind of what the profile is and kind of what the use cases are for those card customers?

抱歉，我不確定你們是否聽得到，我這邊好像有點回音。好了。不過，是的，我想了解一下平均卡片用戶的輪廓是什麼。顯然，我記得中期目標大概是 100 億美元的 GMV，以及約 750 萬名活躍卡片用戶。而我們目前卡片用戶數大概已接近那個水準的 60%。所以，是的，想請你們更新一下用戶輪廓，以及這些卡片客戶的使用情境（use cases）是什麼？

Super general terms, it skews a little bit higher credit quality than the average Affirm consumer for no other reason than we make it that way. We're still at the limit slightly more conservative as to who gets the card offers and approvals. It's really converging towards this is just the average Affirm consumer. But right now, I think the credit quality is slightly better on the card or somewhat better on the card.

非常概括地說，它的信用品質（credit quality）會比一般 Affirm 消費者略高，原因沒有別的，就是我們刻意這麼做。我們在誰能收到卡片邀請與核准方面，仍然稍微保守一些。它正在逐步收斂到「這就是一般的 Affirm 消費者」。但就目前而言，卡片的信用品質確實略好，或說明顯好一些。

The usage patterns are broader, more frequent, obviously, than sort of the more casual affirm consumer that uses us four or five times a year, six times a year now. Card customers start out, I think it's like a 40% higher and goes up from there. Maybe most useful thing is the category usage is more even.

使用型態更廣、更頻繁，顯然比那種較為偶爾使用的 Affirm 消費者更常用；後者一年用我們四、五次，現在大概六次。卡片客戶一開始的使用頻率，我記得大概高出 40%，而且之後還會再往上。也許最有用的一點是：品類使用更平均。

Like typically, it takes a little while for an affirm consumer to realize that if they found us or got exposed to us in category X, it takes them some number of months to rediscover us again at another retailer and say, oh, wait a second, it also works for fashion, not just travel. When you get the card, you have a muscle memory for it. This is a general-purpose tool. It works everywhere. And so the category dispersion begins a little bit sooner and just stays fairly wide. It still skews more considered purchases than kind of your typical low AOV spend, which is fine with us.

通常，Affirm 消費者需要一點時間才會意識到：如果他們是在品類 X 認識我們或接觸到我們，可能要過幾個月才會在另一家零售商再次發現我們，然後說「等等，原來它也適用於時尚，不只適用於旅遊」。但當你拿到卡片後，你會形成肌肉記憶。這是一個通用工具。到處都能用。因此，品類分散會更早開始，並且一直維持相當廣。它仍然更偏向需要審慎考量的購買，而不是典型的低 AOV 消費；我們也能接受。

That is a much easier value point to drive to merchants. They understand that they wouldn't have sold a $700 thing or a $500 thing unless a firm was involved for this particular consumer given their preferences, and the card highlights that that much better. And so sort of a quick sketch, I think at the Investor Forum, we'll say a lot about the card as well. We have some nice little surprises there.

這對商戶而言是更容易傳達的價值主張。他們明白：以這位消費者的偏好來看，如果沒有公司（a firm）介入，他們不可能賣出一個 700 美元或 500 美元的商品；而卡片把這一點呈現得更清楚。所以先快速勾勒一下；我想在投資人論壇上，我們也會談很多關於卡片的內容。我們那邊會有一些不錯的小驚喜。

Got it. Thank you very much.

了解。非常感謝。

Jacob Haggerty, Baird.

Jacob Haggerty，Baird。

Hey, guys. Thanks. So I was just looking at the loan loss on purchase commitment, and it looks like that came down as a percentage of.

嗨，各位。謝謝。我剛在看「購買承諾（purchase commitment）」上的貸款損失（loan loss），看起來它占比下降了，作為……的百分比。

Like lower than it's been in the last few quarters. Is there anything to that why you're getting maybe better economics from your purchasing partners or something along those lines?

比過去幾季都還低。這背後有什麼原因嗎？例如你們從購買合作夥伴那邊拿到更好的經濟條件之類的？

Yeah, that's really driven by the 0% volume in the business. It's not necessarily due to the economics with a single vendor or originating bank or anything in that ecosystem. It's just a function of. The sort of discount rate that we apply to 0% loans. So, yeah, no economic changes there. It's really a function of mix in term length.

是的，那主要是由業務中 0% 方案的量所驅動。不一定是因為單一供應商或發卡／放款銀行，或該生態系中的任何經濟因素。這只是……我們對 0% 貸款所套用的折現率的一種函數。所以，是的，那裡沒有經濟層面的變化。這真的只是期限長短組合（mix）的結果。

Gotcha. Thank you.

了解。謝謝。

Kyle Peterson, Needham & Company.

Kyle Peterson，Needham & Company。

Great. Good afternoon, and thank you for taking my question. I wanted to go back on funding, strictly on the forward flow side, see if you guys could give us whether it's a rank order, kind of relative sizing of some of these.

很好。下午好，謝謝讓我提問。我想回到資金來源的問題，嚴格聚焦在 forward flow 這一端，看看你們是否能提供一些排序（rank order）或相對規模的資訊，讓我們理解其中一些項目的大小關係。

Forward flow buyers, kind of as to what they look like under the hood. Understand everyone.

Forward flow 的買方，大概在底層結構上是什麼樣子。我理解每一家都……

Can be a lot different here, but I think some of the stress seems to be worse than some of these kind of semi-liquid retail vehicles versus, kind of larger, more permanent forms of capital. So, I guess, like, if you could just give us any relative sizing or color on what the forward flow channel looks like, that would be extremely helpful.

可能差異很大，但我覺得某些壓力似乎比其他更嚴重，像是一些半流動性的零售載體（semi-liquid retail vehicles），相較於更大型、更永久性的資本形式。所以我想，如果你們能提供任何相對規模或一些補充說明，描述 forward flow 通路的樣貌，會非常有幫助。

Without getting too specific.

在不講得太具體的前提下。

We're heavily weighted away from things that are very liquid and subject to those kind of volatility that you're referring to.

我們的配置很大程度上避開了那些非常流動、且容易受到你所提到那種波動影響的類型。

Our largest affordable counterparties, our joint venture with 6th Street, we have large pension funds and large insurance complexes, which obviously don't fit that description and even. Among the funds who do participate in our program, they tended to be, again, overwhelmingly not of the kind that I think people are talking about.

我們最大的、可負擔（affordable）的對手方，是我們與 6th Street 的合資企業；我們也有大型退休金基金以及大型保險集團，這些顯然不符合你描述的那種情況，甚至……在參與我們計畫的基金之中，他們也同樣傾向於——再一次地——絕大多數不是大家所談論的那一類。

And that's why we see such a strong renewal and repeat rate, while demand continues to be very high for the asset amongst whole loan buyers.

這也是為什麼我們看到非常強的續約與重複合作率，同時在整筆貸款（whole loan）買方之間，對這項資產的需求仍然非常高。

They really do like the ability for a firm to generate consistent credit.

他們確實很喜歡一家機構能夠持續產出穩定的信用表現。

Outcomes that they can underwrite to and generate returns for their funds. And we like the capital efficiency of those partnerships, and so we grow together and have done a really good job of that over the past three years.

這些結果是他們可以據以承作（underwrite）的，並為其基金創造報酬。而我們也喜歡這些合作關係的資本效率，因此我們一起成長，且在過去三年確實把這件事做得很好。

Great.

很好。

Thank you very much.

非常感謝。

Jamie Friedman, Susquehanna.

Jamie Friedman，Susquehanna。

Hi, thanks for taking my question. I wanted to ask about adaptive checkout. I don't mean to front-run the conversation next week, but if you might have any perspective on how that's evolving, it seems like a real opportunity to serve merchants and consumers alike. So any perspective on adaptive checkout would be helpful.

嗨，謝謝讓我提問。我想問關於 adaptive checkout。我不是要搶先下週的討論，但如果你能分享一些它如何演進的觀點，看起來這是同時服務商戶與消費者的真正機會。所以任何關於 adaptive checkout 的看法都會很有幫助。

Definitely don't mistake my lack of name checking it in this particular letter for any sort of backing away from the strategy. Quite the opposite. We are -- at this point, we're basically selling Adapt and Boost together as a single thing. One of the -- we have a tendency to be overly littered.

請不要把我在這封信裡沒有特別點名它，誤解為我們在策略上有所退卻。恰恰相反。我們——到目前為止——基本上是把 Adapt 和 Boost 當作一個整體一起銷售。其中一個——我們有一個傾向，就是描述得過於……到處都是。

Literal? Literal. It's been a long day. We're going to be overly literal in our description of our products, and so we're trying to learn how to package and market better. And so very soon, you'll just hear strictly about affirm checkouts, something like that.

字面？字面。今天很長。我們在描述產品時會過於字面化，所以我們正在學著如何更好地打包與行銷。因此很快你們就只會聽到「Affirm checkout」之類的說法。

But it's doing really well. It's becoming more and more.

但它表現得非常好。它正在變得越來越……

Understood by our merchant base, and that is what you want. I think hopefully very soon it'll just be the thing you turn on and you don't try to play with the knobs yourself. Our AI will find the optimal setting in real time for every new incremental consumer.

被我們的商戶群更充分地理解，而這正是你想要的。我希望很快它就會變成你只要打開就好，而不需要自己去調那些旋鈕。我們的 AI 會為每一位新增的消費者，即時找到最佳設定。

So I think we do have a bunch of content on it at the investor forum, so I'd rather not drop any stats on that here.

所以我想我們在投資人論壇上有不少相關內容，因此我寧可不在這裡拋出任何統計數字。

Understood.

了解。

John Hecht, Jefferies.

John Hecht，Jefferies。

Good afternoon and thanks for taking the question. A lot of questions have been asked and answered, but yeah, I'm wondering, Max, like, what do you -- yeah, this is obviously a competitive industry, not only with other buy now, pay later companies, but the general consumer credit spectrum.

下午好，謝謝讓我提問。很多問題都已經問過也回答過了，但我想問，Max，你怎麼看——是的，這顯然是一個競爭激烈的產業，不僅是與其他先買後付（buy now, pay later）公司競爭，也是在更廣泛的消費者信貸光譜中競爭。

Clearly, you guys are taking share in a competitive, maybe even increasingly competitive industry. And I'm wondering, obviously, size, scale, brand, all that stuff matters. That's stuff that's been around for a while for you guys. What do you think -- is there anything new or what do you think is changing in terms of competitive positioning as the industry, even though it's not mature, but as it generally matures?

很明顯，你們正在在競爭激烈、甚至可能愈來愈競爭的產業中取得市占。我想問，當然規模、體量、品牌等等都很重要。這些對你們來說已經存在一段時間了。你認為——有沒有什麼新的因素，或你覺得隨著產業逐步成熟（即便還不算成熟），在競爭定位上有什麼正在改變？

I will invite Michael and Rob to comment in a second because I'm going to scrape the bottom of my brain for something incrementally new. But we're very focused internally. I guess the reason I'm struggling to come up with something particularly clever is I can tell you a lot about what's changing here. It's really hard to see what the outside world is doing when you're that focused on internal effort.

我等一下會請 Michael 和 Rob 也補充，因為我得從腦袋深處再挖一點「新增量」的東西。但我們內部非常專注。我想我之所以很難想出什麼特別巧妙的說法，是因為我可以告訴你很多我們這裡正在改變什麼。當你如此專注於內部投入時，其實很難看清外部世界在做什麼。

It's a little bit easier to do what we do to be completely transparent. The consumer knows who we are. The one thing that is true, and we can see this when we do just consumer surveys as well as sort of more mechanical understanding of consumer preferences, there are people that have decided a firm is their jam and that's what they're going to use. And it's really compelling. We can tell consumers, hey. You should go and get yourself an Affirm card because brand X is not yet integrated with Affirm directly, but it's okay, it's accessible.

坦白說，做我們正在做的事情會稍微容易一些，因為我們是完全透明的。消費者知道我們是誰。有一件事是真的，而且我們從消費者調查以及更偏機械式的消費者偏好理解中都看得到：有些人已經決定 Affirm 就是他們的菜（their jam），他們就會用它。而這非常有吸引力。我們可以告訴消費者，嘿，你應該去辦一張 Affirm 卡，因為品牌 X 還沒有直接整合 Affirm，但沒關係，仍然可以使用。

And at some point in the past that felt like maybe they will, maybe they won't, we now have data that shows that they will. Consumers believe, some percentage of our consumers believe that Affirm is a general purpose tool that works anywhere, you just have to have the Affirm card or the Affirm virtual card on your app screen.

而在過去某個時點，這聽起來像是「也許會、也許不會」；但現在我們有數據顯示，他們會。我們的一部分消費者相信，Affirm 是一個可通用的工具，哪裡都能用——你只需要在 App 畫面上有 Affirm 實體卡或 Affirm 虛擬卡。

And. More and more of them understand how it's done, so long as we treat them right and we handle our post-transactional relationship as well as we do in the transaction, they come back. And that just makes repeats a little bit easier. We continue to maintain 90-plus percent of our transactions come from returning users to Affirm, which is great. It's a lot easier to get the second and third transaction than the first.

而且……越來越多人理解它是怎麼運作的；只要我們善待他們，並且把交易後的關係經營得跟交易當下一樣好，他們就會回來。這就讓重複交易變得更容易一些。我們仍然維持 90% 以上的交易來自回流用戶，這很棒。拿到第二筆、第三筆交易比拿到第一筆容易得多。

So all of that, it's a little bit easier to grow today than it was six months ago and 12 months ago and every passing quarter or year makes our growth actually feel a little bit easier. There's a great cycling expression, it doesn't get any easier, you just go faster and we still work just as hard as we've ever done, but the results are. Escalating, if you will.

所以綜合來看，今天的成長比六個月前、十二個月前更容易一些，而且每過一季或一年，我們的成長感覺確實都更容易一點。有一句很棒的自行車運動說法：它不會變得更容易，你只是騎得更快；我們仍然跟以往一樣努力，但成果正在……如果你願意這麼說的話，是在加速提升。

This is Michael. I think for complicated business with a lot of moving parts, I think about our position in the market a little is actually quite simple. This is what you get when you compound results like this over the course of many years without pivots, without changing your identity into who you are. We showed up to the capital market the same way we did when we first showed up to merchants the same way.

我是 Michael。我認為對於一個有很多可動部件的複雜業務，我對我們市場地位的看法其實相當簡單。這就是當你多年來持續複利式地累積這樣的成果、沒有轉向（pivot）、沒有改變你是誰的身分定位時，所得到的結果。我們進入資本市場的方式，跟我們當初第一次面向商戶時的方式是一樣的。

Offering to drive better conversion, better outcomes, the promises we make to consumers we've kept over the years and when you compound all that and stay really focused on doing the thing that matters, you end up building a pretty big lead. I think some of our other players in the space have changed who they are, want to try to enter new spaces and become something that they're not and it shows with their footballs and doubling on results.

為了帶動更好的轉換率、更好的成果，我們多年來對消費者做出的承諾也都兌現了；當你把這一切複利累積起來，並且始終高度專注在真正重要的事上，最後就會建立起相當大的領先優勢。我認為我們在這個領域的其他一些參與者已經改變了自己的定位，想要嘗試進入新的領域、變成他們原本不是的樣子，而這也反映在他們的「花樣」以及對結果的加倍押注上。

Ours is just the benefit of compounding the same awesome thing over and over and over again.

我們只是把同一件很棒的事情一再、一再、一再地複利累積，這就是我們的優勢。

That is very well said. All right, guys.

這說得非常好。好，各位。

Thank you for the perspective.

謝謝你提供的觀點。

Thank you.

謝謝。

Jeff Cantwell, Seaport Research.

Jeff Cantwell，Seaport Research。

Hey, thanks, guys. I wanted just to follow-up on that earlier question. On the Affirm card, you said there's a long list of things you've done and continue to do to increase adoption. I was just hoping to better understand what exactly is on that list of things you're doing. It's driving that increase in cardholders to 4.4 million. Can you maybe help us understand the work you're putting in to increase the number of cards in your customers' hands? And then as you look ahead.

嗨，謝謝各位。我想就剛才那個問題再追問一下。關於 Affirm 卡，你提到你們做了、也持續在做一長串事情來提高採用率。我希望能更清楚了解，那份清單上具體到底有哪些你們正在做的事。這些事情正在推動持卡人數增加到 440 萬。你能否幫我們理解，你們為了讓更多卡片到客戶手上所投入的工作？以及當你往前看。

What are going to be some of the biggest drivers to increase that number by even more? I'd imagine you would expect to see that 17% of cash rate increase further over time, but what would you say are going to be the biggest drivers to increase the number of Affirm cardholders? Is it the marketing of the product or opening new geographies, other new TAM opportunities? Just case we can help us understand the outlook for the Affirm card better. Thanks.

接下來有哪些最大的驅動因素，能讓這個數字進一步提升？我想你們應該預期那 17% 的現金使用率會隨時間進一步上升，但你認為推動 Affirm 持卡人數增加的最大因素會是什麼？是產品行銷，或是開拓新的地理市場、其他新的 TAM（總可服務市場）機會？只是希望你能幫我們更好地理解 Affirm 卡的展望。謝謝。

Yeah, we don't do performance marketing at Affirm. We don't have a business model to pay to acquire. I think maybe that's a bit of a misconception that some people who are less excited about the card than who we are have about it.

是的，我們在 Affirm 不做績效行銷（performance marketing）。我們的商業模式不是靠付費去獲客。我想這可能是一個誤解：一些對這張卡沒有我們這麼看好的人，可能會這樣看待它。

We're not out buying ads, not mailing cards to mailing advertisements to people in the mail. That's not how the business works. The reason why the card is such a compelling business for us is the best way to reengage consumers who we already know and have had successful transactions with.

我們不會去買廣告，也不會把卡片寄出去、或寄廣告郵件給人。這不是這門生意的運作方式。這張卡之所以對我們而言是非常有吸引力的業務，是因為它是重新喚起我們已經認識、且曾成功交易過的消費者的最佳方式。

And that's really the strategy. The strategy is to continue to scale the network and ensure that the. Consumers who get access to the card and that we build a card that they can understand and they can use it in as many transactions as possible to continue to take a big share of their spend from other payment devices. And that's the focus and it's really that simple.

而這就是策略的核心。策略就是持續擴大網路規模，並確保——讓能取得這張卡的消費者，拿到一張他們看得懂、也能在盡可能多的交易中使用的卡，從其他支付工具手上持續拿下他們支出的很大份額。這就是重點，而且真的就這麼簡單。

And just to give you some flavor of the things we do internally. Some of these will sound very unglamorous, but given our scale and our attention to numerical detail, I assure you these are very meaningful. So every pixel in the app is at any given time being A/B tested, by which I mean A, B, C, D, E, F, like multi-legged extremely high-density multivariate testing.

再補充一些我們內部在做的事情，讓你更有感覺。其中有些聽起來可能不太光鮮，但以我們的規模以及對數字細節的重視，我向你保證這些都非常有意義。所以，App 裡的每一個像素在任何時刻都在做 A/B 測試；我說的 A/B 是指 A、B、C、D、E、F 這種多分支、極高密度的多變量測試。

And the outcome of that is. We just shave down the friction. So if you, it's not super easy to replicate because we're fairly good at keeping our cohort separate, but if you've got enough people together, they open their app and none of them had the card, they would see a slightly different experience in very subtle ways.

而其結果是——我們只是把摩擦一點一點削掉。所以如果你——這不太容易複製，因為我們很擅長把不同 cohort（用戶群組）隔離開來——但如果你把足夠多的人聚在一起，他們打開 App，而他們原本都沒有這張卡，他們會在非常細微之處看到略有不同的體驗。

And in some number of weeks or days, we will know which one of them is most compelling when someone signs up for the card, but not just signed up for the card, but actually uses the card and sticks to it. And becomes a no more compelling or no less compelling credit risk. So there's a lot of downstream effects of any form of internal product change.

在幾週或幾天之內，我們就會知道哪一種體驗在某人申請這張卡時最有吸引力——但不只是申請，而是實際使用並持續使用。並且成為信用風險不更差、也不更好的客戶。所以任何形式的內部產品變更，都會帶來很多下游影響。

That we have to contend with. We can't just say, oh, go do that. It's not like you get a loan, everybody gets a loan. You get a loan, and then we have to make sure that the loan you got actually got paid off, and it was a good idea to give you the card based on whatever you in this thought experiment is. And so there's an incredible number of just optimization that happens on our own surfaces, and every time we think we've hit plateau, we find that there's another single or double-digit percentage gain to be had, and we're very far from running out of ideas.

這些都是我們必須面對的。我們不能只是說：喔，就去做那個。這不像你拿到一筆貸款、大家都拿到一筆貸款。你拿到一筆貸款後，我們還得確保你拿到的那筆貸款確實有還清，而且基於你（在這個思想實驗中）是什麼樣的人，給你這張卡是個好主意。因此，我們在自有介面上做了大量的優化；每次我們以為已經到平台期，總能發現還有另一個個位數或雙位數百分比的提升空間，而且我們離點子用盡還非常遠。

To give you a totally different flavor of what we might do at some point, there's painful little going on in store for any of the BNPL players and we think we're the best, we think we're the farthest ahead in terms of how to use our product inside of a physical retail, but boy, we have some. Really interesting ideas and we're getting on them as quickly as we can. And so that's another reason to use our card. As much as we love our online e-commerce domination, we definitely want the remaining 80% of commerce or 75% of commerce, whatever. And so there's just a lot to do with the product.

再給你一個完全不同面向的例子：未來某個時間點，對所有 BNPL（先買後付）玩家來說，線下門市端都會有一些「不好受」的事情正在醞釀；而我們認為自己是最強的，也認為在如何把產品用在實體零售場景上，我們走得最前面，但天啊，我們有一些——非常有趣的想法，而且我們正以最快速度推進。這也是使用我們卡片的另一個理由。儘管我們很喜歡在線上電商的主導地位，我們也絕對想要拿下剩下的 80% 或 75% 的商業交易量，不管是多少。所以在產品上還有很多事情可以做。

I'll end where Michael started. It is not a matter of external marketing, it's a matter of just making sure the product is as accessible, as easy to understand. We have a running tally of every possible declination when the card does not approve a transaction.

我最後回到 Michael 一開始說的：這不是外部行銷的問題，而是確保產品盡可能容易取得、盡可能容易理解。我們會持續統計所有可能的拒絕原因，也就是卡片未核准某筆交易時的每一種拒絕情況。

And every day, someone's job is to ask the question, was this decline intelligent, as in this was a bad credit decision, the consumer should not have been approved, or is it a mistake of the user, a mistake of a firm, mistake of our underwriting engine, et cetera. And so all of that is an enormous volume of work. It can move as quickly as my agents can code it, but then it still has to be tested in the real world and validated and made statistically significant.

而每天都有人專門負責問一個問題：這次拒絕是否「聰明」——也就是說，這是一個不好的信用決策、消費者本來就不該被核准；還是這是使用者的錯誤、Affirm 的錯誤、我們承保引擎的錯誤，等等。因此，這些都是非常龐大的工作量。它可以快到我的工程師能把程式寫出來的速度，但之後仍必須在真實世界中測試、驗證，並達到統計顯著性。

There is very little doubt in my mind that we will not run out of things to do there for years.

我幾乎可以肯定，我們在這方面多年都不會缺事情做。

Great. Thanks very much.

很好。非常感謝。

There are no further questions at this time, so I would like to turn the call back over to Zane Keller for closing comments.

目前沒有進一步的提問，因此我想把電話交回給 Zane Keller 作結語。

Okay. Thank you for joining the call this afternoon. We appreciate your time. We look forward to seeing many of you at the Investor Forum next week. See you there. Bye.

好的。感謝各位今天下午參加電話會議。我們很感謝你們撥冗。期待下週在投資人論壇見到各位中的許多人。到時見。再見。