Meta Muse’s First Numbers: Why Commissions Won’t Pay for It

Episode Transcript

00:00:00 - 00:00:23

Mike: Welcome to another episode of Growing Ecommerce. I’m one of your hosts, Mike, and with me as always, Chris. Hey, Chris.

Chris: Hey! Cheers.

Mike: Cheers. So on today’s episode, we’ve been musing about the first numbers on Meta Muse disclosed at Meta’s earnings call, and we’re going to give you a breakdown of that.

00:00:23 - 00:00:46

Chris: Is it enough to float the product without giving it away? Sorry, no, we’re giving ourselves away. You’re not literally asking me now, because…

Mike: No, no, let’s talk about it. At least if you can do some math.

Chris: Yeah.

Mike: Do you want to tell us anything about Meta earnings at large first, Chris? At large, were they good, bad? How did the market react?

00:00:46 - 00:01:10

Chris: Look, I followed the earnings call, and I just read the headlines. As always, the overall revenue literally obliterated the consensus of Wall Street by a significant margin, and the stock immediately reacted positively. But then something happened.

00:01:10 - 00:01:40

Mike: CapEx went through the roof.

Chris: Yeah, no big surprise, CapEx went through the roof—by the way, thanks to Muse, I guess. Part of it for sure. And I think what hurt the stock most: the stock is down roughly 11% in the last five days, although the top line is growing like crazy. Just to put it in perspective, 26% year-over-year at these revenue levels is completely outstanding.

Mike: Yes.

Chris: But CapEx surged, total costs surged massively because there was one big legal case which had to be accrued in that quarter, roughly $10 billion, and free cash flow was negative.

00:01:40 - 00:02:00

Chris: Free cash flow was negative, and I think we talked about that—that’s the theme right now: is there enough monetization power for all the things that CapEx goes into? Maybe that’s a great segue to Muse, because it should be one of the revenue drivers in the future, right?

Mike: Yeah, absolutely.

00:02:00 - 00:02:19

Mike: It’s definitely the direction of travel for them. This is increasingly looking like their next big project.

Chris: Yeah. And I mean about the amount of money that is spent here—I don’t know that number off the top of my head if they’ve disclosed it, but it’s a lot. What did you see? Did she say it publicly?

00:02:20 - 00:02:46

Mike: No, she didn’t say it publicly. They were at a retreat in Hawaii, as far as I know, and Alexandr Wang—he’s basically the product lead of this feature, he’s been pushing this—he talked about it publicly. She apparently told him multiple times while they were on the beach in Hawaii, “You know, this really needs to work. You understand that if this doesn’t work, it’s going to be bad for us, right?”

00:02:46 - 00:03:06

Chris: Oh, if this is coming from leadership, bro, yeah, you know the pressure is on. This is going to be real bad if this doesn’t work for us.

Mike: And we’ve talked about Meta Muse in a couple of previous episodes. There’s been a lot of hype for it, and I think the impression is that it’s working. It’s been the number one app almost every single day for week after week.

Chris: Great product. Great consumer product.

00:03:06 - 00:03:32

Mike: Yeah, but now we have some fun new numbers to play with where we can try to understand, like, okay, but is this actually working? Because it’s very expensive to operate and it’s being offered for free right now.

00:03:32 - 00:03:57

Chris: The bottom line is: does it? Of course not, not as of now. But do they have a trajectory to making Muse a substantial contributor to their top line? This is where it gets tricky, I assume. But maybe let me go through the numbers and then do some scenarios for where this could land.

Mike: For sure.

Chris: Could it be big enough at one point in time? Because they’ve got like 3.3 billion users in the world, and if you imagine rolling out Muse to even a fraction of their market…

Mike: Yeah, absolutely.

00:03:57 - 00:04:19

Mike: Even a fraction of the market. And we got some more details: at the time of earnings, there have been 10 million app installs.

00:04:19 - 00:04:43

Mike: And we learned that about 25% of those become daily active users—which I don’t have a benchmark for that, but 2.5 million implicitly…

Chris: Daily active users, yeah.

Mike: Daily active users. Because, you know, you’re right: the benchmark is typically—you often hear about monthly active users in software, but daily active users is a high level of engagement.

00:04:43 - 00:05:05

Chris: Yeah, that’s true. I think it’s good. They didn’t explicitly say 2.5 million, but let’s call it that. So impressive. And I don’t know how that compares to Threads at the same time, because I don’t know about Threads’ daily active users—I think their engagement was low, but their app installs were actually way more.

00:05:05 - 00:05:26

Mike: Yeah. So the question also is, Meta has this distribution effect, but I feel like the path from being on social media looking at Reels, and this distribution effect comes into play and says, “Hey, come check out this other social app. It’s called Threads, and you better reserve your handle before it’s gone,” stuff like that… I think that makes a lot of sense.

00:05:26 - 00:05:50

Mike: Whereas: “Okay, I’m on Instagram browsing Reels, and suddenly there’s this furry mascot saying, ‘Help me, let me help you organize your life. Do you know what an agent is?‘” I can sort of understand why distribution isn’t having quite the same lever. However, it’s a strong baseline. These are good numbers.

Chris: If this were a startup, it would be a unicorn, you know?

Mike: Absolutely, yeah.

00:05:50 - 00:06:19

Mike: So they also talked about shopping and some of this stuff. That’s where it gets interesting for us commerce people.

Chris: Yes, what do the numbers tell us?

Mike: Shopping is the biggest use case for Muse. By the way, is that a surprise to you?

00:06:19 - 00:06:41

Chris: No, I don’t think so. Personally, I wouldn’t trust Muse to do things like answer my emails or do stuff with my calendar, whereas with shopping, I would feel more comfortable. What about you?

Mike: Look, I haven’t tried it. I don’t know how good this whole shopping experience with Muse is. I would have thought that administrative stuff…

00:06:41 - 00:07:07

Mike: …would be the more dominant use case. But great for them, because I think this is where…

Chris: Shopping is probably the one thing where monetization is biggest.

Mike: Yeah, it’s the biggest. So good for them.

Chris: Yeah, for sure.

00:07:07 - 00:07:25

Mike: Because how do you… You know, probably Muse uses a lot of tokens to do this admin stuff, and what’s the path to monetization? By the way, I should have said admin stuff was easier. Lesson for next time.

Chris: Yes, let me take this note.

00:07:25 - 00:07:51

Mike: All right, so shopping is the most dominant use case. It’s at 25% among weekly active users. So we know their install base, we know their daily active users. We don’t know weekly, but it will be higher than the 25%.

Chris: Yeah, so it’s somewhere between 2.5 million and 10 million. It’s probably like 40% to 70%. It could be anywhere in that range. Call it half.

00:07:51 - 00:08:12

Chris: Just to paraphrase, because I think it’s important that we are on the same page with the numbers: there are these 10 million app installs—that’s their potential market right now. 2.5 million of them are daily active, highly engaged.

Mike: Highly engaged.

00:08:12 - 00:08:36

Chris: And we are now talking about the weekly active users. We don’t know how many of them there really are, but it will certainly be more than 2.5 million.

Mike: Exactly, so medium engagement people.

Chris: And there we know that shopping is the biggest use case. 25% of them are shopping.

00:08:36 - 00:09:04

Mike: Exactly. And then they dropped this pretty eye-watering number: $200 AOV.

Chris: Yeah, that would be huge for e-commerce. $200, right?

Mike: By the time this episode drops, these earnings were a week or so ago, maybe more, but the headlines were running with stuff like “just 25% of 10 million,” with no qualification, no nuance whatsoever. But actually, they did nuance this: it’s weekly active users, and this includes travel booking…

Chris: Yes, which obviously makes the order value way bigger.

00:09:04 - 00:09:32

Mike: Yeah. So the $200 AOV—given that we don’t know the share of travel here, I think it’s very hard to pin down. But when they say 25% shopping, they’re talking about commerce overall, not just product shopping.

Chris: Yeah.

00:09:32 - 00:09:54

Mike: The $200 AOV is… okay, but again, we don’t have the details here.

Chris: I think it’s good, but let me walk through the math now.

Mike: Here I want to give a shoutout to my online buddy Reed Jackson. Reed is a Googler based in Boston.

Chris: Yeah. I hope I’m not putting too much attention on you, Reed, because there are some Googlers who like to listen to this podcast.

00:09:54 - 00:10:15

Mike: Great people work at Google. What I really appreciate about Reed is that there are so few Googlers who really speak out online or share anything besides whatever was written by marketing.

Chris: Yeah, exactly.

00:10:38 - 00:11:02

Mike: Reed is funny, he’s smart, he shares opinions. Of course, it’s net positive towards Google—you have to bear that in mind—but he shared this interesting feedback on the Muse numbers. So let’s walk through his assumptions here. Now we answer the question…

Chris: Yes, we’ll find out.

00:11:02 - 00:11:26

Mike: In fairness, Zuckerberg and co. are talking about multiple revenue lines for Muse, and they definitely understand that this thing needs to make money.

Chris: They have said that ads will be one revenue stream.

Mike: Exactly. We’ve talked about ads in the past, but it was rather like, “Yes, there will be ads.” There was no certainty about how big this has to be, just that there will be ads.

00:11:26 - 00:11:54

Chris: Yeah, there’s going to be subscriptions, there’ll be ads, there’ll be B2B use cases. And what they mentioned in relation to shopping is that there’ll be a take rate commission.

Mike: Commission. All right, let’s run through this. There are 250 million monthly active users in the US and Canada for Meta’s family of apps.

Chris: 250 million. Okay, I’ve seen it might be 270, but it’s in that range.

00:11:54 - 00:12:22

Mike: Okay. Now let’s assume that the 25% holds there. It’s generous as it scales up, of course. And let’s assume that the $200 AOV holds as it scales up. These are generous assumptions because right now we’ve got the early adopters, the nerds—these people might be more affluent, and these numbers will probably get worse as time goes on, not better.

Chris: Hold on, my phone is ringing on my watch, and I don’t know how to hang up.

00:12:39 - 00:13:06

Mike: Professor Robert Bauer got caught.

Chris: Yeah, he wants me to give a lecture there.

Mike: Where were we?

Chris: Let’s start with the addressable market.

00:13:07 - 00:13:29

Mike: Let’s do the math now and show how big the potential actually is for Muse. So: 250 million monthly active users—it might be more, 270, but it’s in this ballpark—in the US and Canada for the family of apps. That is the regional target here for how big Muse could get.

Chris: It’s basically the serviceable addressable market for North America.

00:13:29 - 00:13:56

Mike: Exactly. Now, there are a couple of assumptions here. Let’s assume that the 25% shopping rate holds. I think it’s a generous assumption, because right now we have high-engagement early adopters, potentially more affluent, so likely these numbers will flatten. There could be counterfactuals here, like shopping becoming more mainstream, but let me assume 25% holds. Let’s assume $200 AOV holds, which is also very generous.

Chris: It’s generous.

00:13:56 - 00:14:18

Mike: And let’s multiply it by 12 months, which is also generous because we don’t know the exact purchase frequency, but let’s multiply it by 12 months.

Chris: Okay.

00:14:18 - 00:14:48

Mike: Then you take a 4% take rate, which is on the higher side, to be honest. It could vary per category—it might be more like 1% to 3%. After about 4%, people start to feel mediated, you know what I mean? But this would yield—if you walk through that whole funnel that Reed laid out for us—generously, 6 billion USD.

Chris: Which is a lot of money. But is it a lot of money if you’re Meta? No, it’s not.

00:14:48 - 00:15:11

Chris: Because in 2025—and honestly I don’t have to project for 2026—Meta crossed roughly 200 billion in revenue. By the way, ads’ share of this revenue was 98%. It’s still funny to think about it.

Mike: I know.

00:15:11 - 00:15:37

Chris: So, 200 billion overall revenue, created mostly by ads. 6 billion compared to that—and of course, given the fact that billions of CapEx ran in and still has to be invested down the line—is a very small piece, even in the generous case you made up here.

00:15:38 - 00:16:05

Mike: Reed estimates that it’s 3% of overall revenue.

Chris: Yes. And of course that’s only one revenue line there. Let’s assume subscriptions do really well, or whatever B2B stuff, but the bottom line is they need ads.

Mike: They need ads.

Chris: Ads are important. We have told you this years before.

00:16:05 - 00:16:26

Mike: You should be very grateful actually, because a couple of episodes back…

Chris: No, I’m just very happy that they listen to the podcast and realize what’s really important to their business. Mark routine-wise listens to the podcast!

Mike: In a couple of episodes back, we talked about Muse potentially becoming a major new advertising surface. And I was like, “Are we already wrong?” because I heard a lot of chatter about commissions. But ads are going to be there.

00:16:26 - 00:16:52

Mike: When you have ads in there, it will be this new GenAI surface.

Chris: And one thing we have to give to Meta: they have shown that they can build, manage, and develop an ad ecosystem that really prints money at high margins. So there will certainly be ad revenue—massive ad revenue, even. But the question is: how much is incremental?

00:16:52 - 00:17:19

Mike: Yeah.

Chris: Just to put it in perspective—because the smart listeners we have here will say, “Guys, hold on, chill out, it’s only North America.” Fair enough. How many active users does Meta have worldwide across the family? 3 billion, roughly.

Mike: Is that monthly active or total?

Chris: I think it’s total 3 billion. So even if you do this maxed-out scenario—which is by no means a realistic scenario—even if they get all 3 billion…

00:17:43 - 00:18:19

Chris: …into that, it’s 10 times the 6 billion, which is 60 billion. Compared to 200 billion overall. I mean, don’t get me wrong, a massive chunk of revenue, but still there has to be additional revenue.

Mike: This is the catch-22, because 6 billion in revenue would make them one of the largest retailers or marketplaces in the US. Amazon’s larger, but they would be massive at that scale. Which, by the way, you have to think that they have to compete against that whole market to get that big.

Chris: Yes.

00:18:19 - 00:18:38

Mike: And still, for a business of their size and a product with this kind of cost structure, it seems insufficient.

Chris: It seems insufficient. Based on this—really, shoutout to your Google colleague—it’s really a fun back-of-the-napkin calculation he’s doing here.

Mike: Yeah, it’s fun.

00:18:38 - 00:19:02

Chris: Of course there will be a lot of nuances to it, but as of today, if we believe in that scenario—which is a very bullish one—I think the global monetization strategy with commissions right now would land roughly around 30 to 35 billion. That would be my take on it.

Mike: Do you know what I think the global commission revenue is going to be, Chris? Hold onto your hat.

Chris: Now, okay…

00:19:02 - 00:19:23

Mike: Zero.

Chris: Zero? You don’t think that…

Mike: I don’t think that commissions are going to work for any of these AI services. ChatGPT was also saying that they were looking at a commission model. I could be wrong here, but the reason I say that is that websites and retailers are already blocking Muse and other agents. They’re already doing it, and it’s free—they don’t want it.

00:19:23 - 00:19:47

Mike: The only thing would be if the demand is so effectively captured that they don’t have the clout, they can’t afford to miss it, and they’re forced to be mediated. But otherwise, I just think commissions are not going to work. I don’t believe in the model, personally.

Chris: Then it has to be even more ads. That’s what it was always about—ads!

00:19:47 - 00:20:10

Mike: You had me at ads.

Chris: Very interesting. Look, from the perspective of the retailer, of course it’s a bad thing if you get an agent to do the shopping for your potential clients end-to-end, for sure. However, Mike—and maybe I don’t want to derail this discussion—what we know for sure is Google is going strongly in the direction of this AI purchasing journey.

00:20:37 - 00:21:05

Chris: Universal Checkout is live in the US—that’s one hell of a step in that direction. Is there any other possibility to create revenue for Google besides commissions? Still ads, of course. But don’t you think commission at some point has to be part of the revenue model, because ads in themselves will not carry it anymore? Because agents don’t need ads.

00:21:05 - 00:21:27

Mike: Yeah, I understand, but I still think Google is working on this hybrid usage model where there are humans in the loop and it’s AI-assisted. Everyone says Google is behind all this. I actually am of the opinion that Google is ahead and smartly conservative on so many things because they are embedding it into their cash cow, which is their ad business.

00:21:48 - 00:22:15

Mike: It’s clear Google is going to have the best model by the end of the year. Gemini came along, and we can debate if it’s best and how accessible it is. But let me get back to this topic: Google has tried this step already with Buy on Google, and it failed. The only reason—that’s why from day one they’ve looked at this as an ad-financed operation, because they know and they’ve been there before.

00:22:15 - 00:22:36

Mike: They’re not so stupid, and I think the others haven’t had that first-hand painful lesson yet. They’re going to, and they’ll find out that ads are the answer.

Chris: Fair enough. What I’m saying is that Google is the perfect example for that. I think there’s a future at some point we will reach, and Google is taking these little steps very smartly.

00:22:36 - 00:23:03

Chris: But if you look at the direction where it’s heading—also the universal cart thing—now I’m in the loop as a human being, but at the end of the day, UCP, universal cart, all the technical infrastructure points towards this option where I just tell my agent to buy A, B, and C. Once this is the case, Google will do this as well.

00:23:03 - 00:23:24

Chris: Then you have to have a revenue model, and I couldn’t think of any other than commissions. At the same time, I’m not a big fan of it, I’m just saying. However, for Muse, whether there will be commission or not, I think what we agree on is: yes, Mark, it’s Chris and Mike again—ads will be important for you.

00:23:24 - 00:23:53

Mike: It will be, also for Muse. I don’t think that they doubt that, but I don’t think they realize the extent of it. But I want to say one more thing about ads, which is that this does have a very positive effect or can have a positive effect on their ad networks at large, because it acts as an intent gateway.

00:23:53 - 00:24:21

Mike: So even if that channel is not as incremental, they’re going to be getting better and richer audience and intent data than ever, which they can then radiate to Instagram, to Facebook, whatever. So it can be a massive hub for them.

Chris: And at the end of the day, we all know that user growth rates are not the thing analysts look at anymore. It’s ad revenue, which is highly profitable for Meta.

00:24:21 - 00:24:47

Chris: But maybe Muse, if it’s really working well, might be this one feature which pulls new clients into the system again, reactivates clients, which again creates a growth loop for them. Whether they use Muse or are within the classic ads ecosystem, I think Muse strategically is a great product.

Mike: Absolutely.

00:24:47 - 00:25:13

Mike: But ads need to be in there.

Chris: Indeed, and we’re going to keep watching it.

Mike: On that note, Chris… by the way, do you want some stats also from the earnings call?

Chris: Let me hear it.

00:25:13 - 00:25:45

Mike: A short pause, please, because I had to think of you—it seems like they are really squeezing every ad to outpace impressions, right?

Chris: Tell me. Yeah, I know, I know. So… well, we can skip it, but sure, let me find it. It states: “core ad engine acceleration.” That’s what I call it. The growth was powered by a 10% increase in ad impressions.

00:25:45 - 00:26:07

Mike: So they just flood you with more.

Chris: Yeah, and a 14% jump in average price per ad, which, I know, man, that is unfortunately not the first quarter we’ve heard numbers like that.

00:26:27 - 00:27:10

Chris: So the one major reason why revenue growth was again massive this quarter was what they kind of called the core ad engine acceleration. The growth was powered by a 10% increase in ad impressions—so they throw more ads at you—and a 14% jump in average price per ad.

Mike: I was waiting for the other shoe to drop. That is unfortunately not the first quarter we’ve seen numbers like that.

Chris: Oh yes, it just keeps getting more expensive. So more ads and more expensive. I think we talked about that once—there has to be a cap at some point in time.

00:27:10 - 00:27:39

Chris: Anyhow, Muse: great product, expect ads to be shown there.

Mike: Yes, expect ads. A pleasure.

Chris: Likewise, always. Thank you, Mike.

Mike: Thanks everyone for listening. This has been another episode of Growing Ecommerce, brought to you, as always, by Smarter Ecommerce, also known as smec. You can learn more at smarter-ecommerce.com. And again, we really appreciate it if you leave a comment. Did you not like our math? Let us know in the YouTube comments. We really appreciate it. See you next time. Bye-bye.

Chris: Awesome.