← Behind the Numbers: an EMARKETER Podcast
Behind the Numbers: an EMARKETER Podcast
Where Should Meta Be Focusing Its AI Efforts? An AI Mark Zuckerberg? 3 Big Questions for Meta | Behind the Numbers
00:30:33
Loading summary
Transcript60 lines
- [00:00]
A
Foreign. Hey, gang. It's Monday, May 11th. Max, Emmy and listeners, welcome to behind the Numbers in the Marketer podcast. I'm Marcus. Joining me for today's conversation, we have Philly based principal social media analyst Max Willins.
- [00:16]
B
Yo.
- [00:17]
A
Hello there. We're also joined by New York Living analyst Emmy Ledemann.
- [00:24]
C
Hey.
- [00:25]
A
Hello. They were brief. Greetings. Today's fact. When you thoroughly shuffle a deck of cards, 52 cards, the exact order of cards you're holding has almost certainly never existed before in the history of the universe and will likely never exist again because after shuffling a deck of cards, there are more possible orders the cards could be in than there are atoms in the universe. The number is eight followed by 67 zeros, which we have a name for that. That's amazing. That might be the most interesting part of this. Someone bothered to name eight followed by 67 zeros.
- [01:15]
C
What's the name?
- [01:16]
A
80. Unvigintillion. Unvigantillion. Possible. Permutations. Permutations.
- [01:26]
C
It's a lot.
- [01:28]
A
That is a lot. That's what, 52 cards. They'll never be in the same order.
- [01:33]
C
I'm sorry for my silence. I'm just taking a second to digest.
- [01:36]
A
Because it's that good or you couldn't care less.
- [01:39]
C
No, no, honestly, it is. It is pretty crazy, but I feel like this is like a fun fact, like first day of like probability, like college class 100, you know, probability. College class, like, like taking probably like a intro to probability, like a math. Did you ever have to take probability?
- [02:00]
A
I think it was part of. It wasn't its own separate class. Should I be a teacher? Is that what you're. Is that what we're saying?
- [02:05]
C
No, that's exactly what I'm saying. This podcast now.
- [02:10]
A
All right, steady on today. Let me think about my new switch to academia. Anyway, today's real topic before I get let go, the big three questions surrounding meta. All right, let's set the table first. In Q1, 2026, Meta made $55 billion in total revenue. That's a lot of money. Okay, for context, we say numbers a lot of the time. What does that mean? It's a lot. An astonishing 33% more year over year. Double the growth. 33% is double last Q1's growth somehow. And it's the fastest growth in the last 18 quarters. So they're making more and more money and somehow they're growing it faster. At least the last quarter, North America is about just under half, 44% of the total part that grew 30% in Q1 year over year, it's close to double last year as well. So purpose of this episode and the series we have around kind of the big tech players to Netflix, Amazon, Google, Meta, et cetera, is to think about the big three questions for the company at the moment. And so I want to throw one into the ring first, gang, because mine kind of relates to the performance numbers that I just talked about. And so I think one of the questions is, I had, I think others had as well is are we happy with Meta's performance or. Or not? Because there seems to be some cognitive dissonance, understandably, once you dig into it. But it's interesting because by that measure you'd think, wow, the street was really. No, they weren't that happy with them. Megan Bobrowski of the Wall Street Journal was explaining that the market's negative reaction to Meta's Q1 results was in large part due to its upwards revision in its expected capital expenditures. Basically, they're spending more money on AI this year than they were planning to, going from 125 billion to 135 billion. And also they said that Q2 will probably be a bit slower than expected. Any thoughts here on this weird world where Meta can make $55 billion in a quarter and grow 33% and still that's frowned upon by the Street.
- [04:16]
B
I personally find it very telling that the street has wrapped Meta's knuckles after delivering that crazy performance. I think anytime you announce you're going to be spending $135 billion, even if it's over the course of the year, and even though to hear Meta tell it, the money is going to go and the compute is going to potentially go toward fueling products that are already working, I think it bespeaks a broader unease with where this gen AI craze is going. Right. I mean, to what you were saying. If you just kind of took the names out of it and said, we made $55 billion in a single quarter and even though we expect next month is going to be kind of slower, it's still probably going to be, you know, very healthy. Double digit growth on $55 billion and they have, you know, half the world's population using at least one of their products on a daily basis, that feels like a really positive story and positive company to hitch one's wagons to. But I think people are starting to get a little bit nervous wondering where all of this spending is leading from a, you know, a monetization standpoint. And so even, even with Meta a company that has, you know, basically been printing money for the last 10 years with its ad prop products. The fact that even Meta is kind of being looked at nervously is a sign of a broader story that's much bigger than this one company. That's sort of my read on it personally.
- [05:59]
A
It's $600 million a day. January through March is $55 billion. So it's a lot. But maybe it's quite a valid question and digging into what Max is talking about for a second Emmy in terms of the kind of nervousness surrounding Meta's AI investment, another question I had was where should matter be focusing its AI efforts? And there was an NPR piece by John Rowich. He was citing Arnau Dia Ratner, a research VP at tech consulting idc, saying I don't think Meta is going to be able to build a best in class generalist model because from a resource standpoint, both with respect to GPUs and human talent is just not going to be easy to do. He points to the incomprehensible sums of money that companies like OpenAI, Anthropic, Google, Nvidia are plowing into their own initiatives to hoover up GPUs that are crucial for AI data centers and to recruit top engineers. Mr. Dyer Ratner thinks that they are too far behind and thinks they might have better luck picking a niche like an AI video or image generation tool rather than trying to compete with an all purpose model like ChatGPT. Do we think that maybe that should be or could be the way that this goes, that they get a bit more specific and don't try to compete with these other guys for a generalist model?
- [07:23]
C
Well, I think this, we brought up this point with the whole like Big Tobacco moment conversation with Meta where it's like if they're already facing all this scrutiny just because of the endless scroll and the impact of those features on mental health, if you layer on an AI tool that only could potentially make things exponentially worse. So it just feels like kind of a risk from that standpoint to invest heavily in AI. But I think when you think about it from an advertiser's perspective, using Advantage plus in all of those AI tools and also Meta just announced that they were laying off 10% of their workforce, which is insane. I think people are going to feel very fatigued by that ad load and also if they can easily tell that those ads are AI generated. So I kind of just wonder as its ad business is growing at this rate and they're relying so much on AI, what is the cost when it comes to user sentiment?
- [08:34]
B
Well, it couldn't be much lower. I was doing some research for a separate thing and found that this is one of the many things that YouGov kind of tracks separately have this sort of giant list of public companies and figures who they keep kind of running tabs on the sentiments around them. And I think it's something like 32% of their panel of people have a favorable opinion of Meta. But as far as the NPR question and whether they should have a generalist model, it seems clear that they seem to understand this and are kind of pursuing several specialist agents already. So they have the business agents and they have kind of a general purpose, one that might help people do things like shop or get better at individual tasks or pursuing individual goals that they have, rather than an all purpose general knowledge bot, which is actually a good segue into one of the questions that I had if I can throw it on the table, which is will their business agents stick in the way that Workplace didn't? Anybody that's followed Meta for a long time knows that they made an attempt, I guess about 10 years ago to sort of turn themselves into sort of an enterprise AI, or excuse me, an enterprise business tool by launching something called Facebook Workplace, which was essentially meant to sort of be like a competitor to Slack, which I'm sure that they were hoping would turn into the tip of a spear that they could use to sort of burrow inside large enterprise business organizations. And that didn't really go anywhere. I think they did get themselves into a couple of large organizations, but metaphorically sunset it after a couple of years. But with their business agents, which they are taking every conceivable opportunity to talk up in public, they're going kind of in the opposite direction where Workplace was really definitely designed for larger organizations. The business agents I think are very much aimed at SMBs and sort of small businesses mostly not in the US it seems. It seems like that's sort of their attempt is to sort of like build out and become the kind of business infrastructure for digital native businesses around the world. And I would say that Meta as an ad business is about as sticky as you could possibly imagine. But if they can kind of get their claws dug even deeper into all these small businesses that use them to attract customers and monetize them, I think that's just going to sort of lock in a another, let's call it half decade at least, of very, very healthy growth. Because the sort of early case uses for this are not just helping Brands acquire new customers, which I'm sure is going to involve a lot of suggesting that they spend money on Meta's platforms, but also kind of agents that help these brands upsell and think about how to better monetize and engage their customers and users. Which again, lends itself very nicely to Meta's platforms, but also sort of solves a, you know, the top three kind of priorities of any small business owner. And so if they're able to get this to work, I think it would be we should basically get ourselves comfortable for many more successive quarters that look like this past one did, at least from a revenue standpoint, maybe not the capex one.
- [12:17]
A
Yeah, I like this one. Emmy, what have you got? Any big questions come to mind for Meta?
- [12:23]
C
Yeah, so in our time spent with social Networks report for 2026, we found that no platform can compete with TikTok in terms of user loyalty and engagement. So TikTok has the cohort of active users that doesn't compare to any of the other platforms. So going back to our conversation about how people actually feel about Meta and, you know, Facebook and Instagram as platforms, my question would be like, how are you protecting user sentiment and making sure that you're not just like another app that people open and then quickly close? Because it's just sort of like a force of habit type of thing. Like, I think people have a relationship with TikTok and see a lot of value in TikTok and Meta's apps are more of a habit to them. So I would just wonder how they're protecting that consumer relationship as they lean more heavily into AI fueled advertising.
- [13:22]
A
It's interesting because it does seem as though TikTok has escaped some of the scrutiny that Meta has faced with the kind of social media addiction trials. I know that they settled out of court for one of the cases and there are a lot of other cases to come. So we'll see if they're able to kind of dodge the bullet, so to speak, in some of the other cases. But do we think that part of it, like, why do you think it's struggling so much to, to kind of get rid of that kind of association compared to someone like a, like a tick tock? Because the social media addiction trials, one of the questions I had was Mesa lost a couple of quite prominent cases. What happens next? Just to recap, folks who haven't been paying attention to this, to, to this part of Mesa, Maria Curie of Axios explained that Meta's recent court losses adding urgency to lawmakers push to Pass legislation that could reshape how social media platforms are designed. After plaintiffs in New Mexico and LA won their cases arguing that social media products were addictive by design, there's thousands of other pending lawsuits which argue the same thing. Meaning monetary penalties could add up quite quickly as well as structural changes. I mean, what do you think TikTok will also be maybe dragged down by this? Maybe talked about in the same context as Meta? When it comes to the negative sentiments
- [14:41]
C
surrounding social Media, I think TikTok definitely has its fair share of flack, but I think even the people that decide to take a break from TikTok eventually end up coming back because it does have this community element, like grassroots type vibe that feels so silly to say because it's like this massive company, but I do think it has a certain positioning that Meta does not. I mean, I remember like when in 2020 when TikTok popped off, like people were calling it like the bright spot of the Internet. And we've clearly like come a long way since then, but I still do think it has a sort of like community type feel that in welcomes discourse and like peer to peer conversation in a way that Meta and I would say Instagram more specifically doesn't. Which kind of also brings me to the question of like how Meta is faring within LLMs, because I know that Reddit and is very like close. A lot of the content on Reddit is surfacing on LLMs and it's like something that is really beneficial for the platform is the fact that it's so based on that peer to peer engagement. And I just don't feel like Instagram specifically like really has that type of dynamic. It feels like it's just kind of a, a content farm for, for influencers and a lot of people just kind of use it passively.
- [16:06]
A
Yeah. Max, what are your thoughts here?
- [16:08]
B
Yeah, I mean, I think that there's a bunch to chew on here, but I think that the thing that's interesting about it's clear. So, you know, Meta and all the other kind of large platforms are going to have to play these very interesting games of judo with one another over the next couple of years where they decide, you know, what kinds of data they're going to allow into their competitors, you know, LLMs and how they do that. Like Instagram, I think sort of probably did this for, for the sake of kind of experimenting and to see what would happen when they announced shortly. I forget the exact timing, but it was basically like right when ChatGPT and Gemini were starting to begin competing in that first wave of user grabbing. And what Instagram or Meta did is they made public Instagram posts crawlable by Google, which at the time struck me as quite surprising, because on some level, a lot of these platforms, what they typically like to do is kind of hoard all the data that they have and all the engagement and all the information that they have inside their platforms. But my sense is that Meta probably did this just to sort of see whether, you know, Google or OpenAI would get really addicted to, you know, having insights from Instagram in its responses. And there are lots of tests that have been done by geo startups that. That show that Instagram is a domain that's cited quite, quite regularly. And a lot of that is down to the fact that it's indexed in Google now. And so I think that that's a really important piece to think about it. As far as TikTok and its kind of, you know, stickiness in people's lives, I was really struck by a piece of news that came out about two weeks ago where TikTok announced campus hubs, which to me smelled, you know, they basically framed it as this. Like, hey, this is a place where you can chat with all the, you know, you know, kids that you go to college with and, you know, stay connected and figure out what's going on and make new friends. But to me, it really just smelled like their attempt at building the infrastructure that. For Facebook groups, which has really proven to be one of the most unbelievably sticky and kind of impossible social surfaces to extricate oneself from. Like, I am somebody who basically does not use Facebook anymore, but as the. As a person who wants to sort of stay connected to his community and is also raising children, it is almost impossible for me to get out of Facebook groups. Like, I just, if I want to sort of connect with people in my community, sort of like, you know, setting up a group chat on my phone with texts, it's really turned into this sort of de facto place where people congregate around hyperlocal things and they. The campus hubs just feels like the earliest and easiest way for them to build a scaled base of people that they can lock in. And then as they grow up, that becomes an alumni group. And maybe you break it down into smaller pieces, whether it's by major or business school or where people go or whatever it happens to be. And so even though I think Emmy's point about the kind of cultural position that TikTok enjoys is right, TikTok is also aware of the fact that that stuff doesn't last forever. And so it's going to be really interesting to see how all that stuff evolves.
- [19:39]
C
Yeah. I wrote a story last year about Meta implementing a similar feature with Instagram, specifically where you can connect your profile to Unidays, which is like a student verification platform, and do the exact thing that now TikTok is now copying. So it's funny that Meta invented this sort of like, use social media to connect with your peers model. And we're seeing them realize that there's this new generation of students that never even had a Facebook profile and like, are shifting that over. So it seems like the whole like, offer some sort of niche community led element is what these platforms are realizing will make their offering sticky.
- [20:26]
B
Yeah, a hundred percent. I wanted to come back to another thing which, you know, Mark is kind of throwing on the table because it's very similar to one of the questions that I wrote down, or at least very, very much related to the question that Marcus put on the table. And my question is, what happens if New Mexico calls Meta's bluff? I'm sure you guys all saw this story a couple of weeks ago now where, you know, as mentioned, Meta lost phase one of its trial against that the state of New Mexico brought against it and they had to pay a big fine. Well, big for, you know, non Meta companies anyway. And phase two of this trial is about, is now underway. And one of the things that's going to happen in this phase is this judge in New Mexico is going to decide on what sorts of structural remedies Meta is going to have to undertake if it wants to not face even more penalties. And at one point, Meta's Meta issued a statement basically saying that the things that were on the table were unworkable and impractical and that if there was no sort of softening of the demands, that Meta was just going to have to leave New Mexico. And to me, this is a really going to be a really fascinating thing to watch because Metta has pulled this card before in different markets. Right? So like, for example, in Canada you still can't access news content on Facebook. In Australia, they tried something similar. In some ways, New Mexico is small potatoes, right? Like it's a state of, I think, two and a half million people, maybe not even that much. And so, you know, Meta is probably betting that they can put the screws to New Mexican legislators and, you know, Attorney General more effectively than, you know, the two and a half million people that are no longer using Instagram. Anymore is going to affect their balance sheets. I didn't have. You can't use Instagram in New Mexico on my bingo card. But it's. That is now like very much on the table, which I think is fascinating.
- [22:29]
A
Yeah, yeah. I. It's hard to know what the right strategy should be here because, yes, New Mexico you could probably, but other states you can't necessarily do that with. So is this the right way to go about it, or should they be getting ahead of this with saying, okay, look, here's our concessions, we'll change these things if you agree to not change these other things? It seems, yeah, combative is the right move here. Emmy, do you have anything else? Any other big questions?
- [22:58]
C
So I also saw last week that Meta announced the launch of its Ads AI Connectors tool. So basically, advertisers can bring in their third party AI tools and agencies and use those and like, apply them to their. Their Meta ad strategy. So my question is, like, what complications may arise from this? And considering that they laid off like 10% of their workforce and are working with smaller people internally, this just seems like a major change that perhaps can't just be supported with AI alone. So I'm just kind of curious, how are you going to adapt to these changes in your advertising offerings with fewer people?
- [23:49]
A
Yeah. Yeah. Max. Any others?
- [23:54]
B
The only one that I have is kind of on the bleeding edge speculative side, which is related to a scoop that Digiday got a couple weeks ago now about the possibility that Meta might kind of get itself back into the TV business. They're apparently having exploratory conversations with SSPs about kind of rebooting the Facebook audience network. But instead of it being, you know, for inventory spread across the web, it would be for inventory spread across CTV environments and programming. And this, to me is just really fascinating because, you know, Meta, particularly Instagram, has kind of made a big deal of the fact that they're kind of making a play for big screens and they want people to like, consume Instagram on their televisions. This, to me, is a sort of tacit admission that that is not going super great. Or maybe if not that it's not going super great, there's kind of an upper limit on it. But being able to monetize the intent data that their systems harvest is something that they're interested in exploring. Likely because Pinterest has sort of decided to make a go of this exact same thing with its acquisition of TV Scientific. But I'm just really, really interested to see where that goes, because it would represent a pretty significant break from the way that they have sought to sort of monetize in the past. But it is also potentially a chance for them to take a commanding position in a market that's right now filled with lots of little companies that are not as big or as strong as it is. Excited to see where that goes.
- [25:42]
A
Very nice. I have a bunch that I left on the table or on the shelf, I should say. Put them on the table. I'll rattle through them really quickly, just give you kind of the headline. But one is this kind of came up recently, so that's why it's open mind for me. It Can Meta maintain its new lead over Google in digital advertising? Our forecasting team thinks that they're going to overtake them in net digital advertising this year. We covered it in an episode April 24 called Meta to Dethrone Google as Top Ad Business. Here's why. So check that one out if you, if you want to listen to more of that one. So that was one of them. The other one is, is Meta quitting, maybe quietly quitting the Metaverse because it's laying off around 700 folks, many of them part of the Reality Labs division that runs Meta Metaverse products. The company said it was a right sizing in investment in Reality Labs, but they're still committed for the long haul. But I thought that was of note. Marissa had a piece Marissa Jones, one of our newsletter analysts, had a piece about Meta's ad disclosure update. Could trigger new regulatory pressure. So basically saying that the company is going to be switching its ads labels from on Instagram and Facebook from sponsored, saying sponsored on a post to ad. And she was saying that that tag's going to be more difficult for the average user to note whilst they're scrolling and could help ads feel more like organic content, which could benefit advertisers. But it also could blur the lines between ads and posts in users fees, which could get them in trouble with regulators. And then the last one, I think is the most wild one of all, which is Claudia Feminine of the Guardian writing that Meta is creating an AI version of Mark Zuckerberg so staff can talk to the boss. The digital clone being trained on his thoughts is being trained on his thoughts, tone and mannerisms to help workers feel connected.
- [27:27]
C
Wow.
- [27:28]
A
Yep. Anyway, we'll pick it. We'll pick a top three.
- [27:32]
C
Do you think that that's like a requirement or it's an optional activity, like talk to him?
- [27:38]
A
Yeah, hopefully optional.
- [27:41]
C
Hopefully it's optional. Let's, let's hope for Those. Those poor people.
- [27:47]
A
All right, so we are going to pick a top three. Emmy, you're a first to pick. We're gonna pick three. Our top three, one each. And that'll be our top three big questions for Meta right now. What are you keeping?
- [27:58]
C
The question I'm keeping is, how is Meta investing in the user experience, specifically competing against TikTok, which has much more engaged users?
- [28:09]
A
Okay, very good, Max.
- [28:11]
B
I'm going to be selfish and pick the question that I had about their business agents. It's something that Meta is super enthusiastic about. They announced recently that they. The. The AI was facilitating, you know, 1 million conversations a week between brands and their customers, and that's now up to 10 million a week. So I'm excited to see where that goes.
- [28:35]
A
Very nice. I'm going to keep the where should matter be focusing its AI efforts. And can they right the ship in terms of getting Wall street back on side? Because it seems like a lot of. Well, Google especially, obviously came up with their earnings and people were quite happy with how they were doing. Even those vast sums of money. People, investors seem much more comfortable with where these other companies are spending these AI dollars. Can Meta convince the street that it's spending its AI dollars? Well, there are top three big questions for Meta. Lots to choose from, lots going on, but there are top three for you. Thank you so much to my guests for helping me put those together. Thank you. First to Emmy.
- [29:14]
C
Thank you.
- [29:15]
B
And Max, Always a pleasure. Marcus, thank you.
- [29:17]
A
Yes, indeed. Thank you to the production crew, who is Lance in this case. Thanks to everyone for listening in to behind the Numbers in the Marketer podcast. And you can, of course, hang out with Susie David Canyon, host of the Reimagining retail show this Wednesday.
- [29:29]
C
And then my last one would be how are you competing for user favorability and specifically against TikTok, which has a more engaged user base.
- [29:39]
A
Wait, this is a second one.
- [29:42]
C
Third.
- [29:43]
A
Third one. No, you're going to pick one. Oh, wait, you don't pick three. It's one each, which makes three.
- [29:50]
C
So why didn't you stop me after the New Mexico one?
- [29:53]
A
Because I was. I thought they was related. I was like, oh, there's more.
- [29:56]
C
Okay, I can. I can choose that one. That's fine.
- [29:59]
B
Sorry, sorry.
- [30:00]
A
I'm not. It's my fault. Okay. Yeah, pick. You have to pick one. Max picks one. I pick one. That's the top three. Sorry, not three each. Sorry. So you go.
- [30:09]
B
You can start. I. I'm gonna be self.
- [30:12]
A
Emmy, you need to. You need to pick one.
- [30:14]
C
Oh, okay, wait. I'm. Now I'm. I know Max said he's going to be selfish. I'm going to choose mine then. Okay, let's. Let's restart.
- [30:23]
B
Okay.
- [30:23]
C
Okay.
- [30:24]
A
Emmy, what are you keeping.