TRANSCRIPT

Q4 2024 — Follow-up call transcript

Q4 2024PDF · 10 pages77 KB
1
Meta Platforms, Inc. (META)
Fourth Quarter 2024 Results Follow Up Call
January 29th, 2025
Operator: Good afternoon, my name is Krista, and I will be your conference operator
today. At this time I would like to welcome everyone to the Meta Fourth
Quarter and Full Year 2024 Results Follow up Q&A call. All lines have been
placed on mute to prevent any background noise.
After the speaker’s remarks, there will be a question and answer session. If you
would like to ask a question during this time, please press star followed by the
number one on your telephone keypad. To withdraw your question, again, press
star one. We ask that you limit yourself to one question and this call will be
recorded. Thank you very much. Kenneth Dorell, Meta’s Director of Investor
Relations, you may begin.
Kenneth Dorell: Thank you. Good afternoon and welcome to the Follow up Q&A call. With me
on today’s call are Susan Li, CFO, and Chad Heaton, VP of Finance. Before we
get started, I would like to take this opportunity to remind you that our remarks
today will include forward looking statements. Actual results may differ
materially from those contemplated by these forward looking statements.
Factors that could cause these results to differ materially are set forth in
today’s earnings press release and in our quarterly report on Form 10-Q filed
with the SEC. Any forward looking statements that we make on this call are
based on assumptions as of today and we undertake no obligation to update
these statements as a result of new information or future events.
During this call we will present both GAAP and certain non GAAP financial
measures. A reconciliation of GAAP to non-GAAP measures is included in
today’s earnings press release. The earnings press release and an
accompanying investor presentation are available on our website,
investor.atmeta.com. And now I’d like to turn the call back over to Krista for the
first question.
Operator: Thank you. We will now open the lines for a question and answer session. To
ask a question, please press star one on your touch tone phone. To withdraw
your question, again, press star one. We ask that you limit yourself to one
question. Please pick up your handset before asking your question to ensure
clarity. If you are streaming today’s call, please mute your computer
speakers. And your first question comes from the line of James Lee with
Mizuho Securities. Please go ahead.
James Lee: Great. Thanks for taking my question. Susan, I was wondering maybe you can
comment about WhatsApp a little bit here. Give us an update on where you are
in terms of AI agent for business and what are the key lessons learned. And can
you also talk about the product launch to non-English speaking countries going
forward? Thanks.

-- 1 of 10 --

2
Susan Li: Hi James. Thank you for the question. So, we are still in the early stages of
piloting business AIs, with a relatively small number of businesses, but we are
gradually expanding our tests. Our focus initially right now is on supporting
businesses that use click to message ads for WhatsApp and Messenger. With
WhatsApp, we have been rolling out the service in English to more small
businesses in India.
We have been expanding to several other countries in Asia-Pacific including
Indonesia and Malaysia. We’re also now slowly expanding business AIs in
Spanish across Latin America including Mexico. And on Messenger, we’re
testing in English across several Asia-Pac countries including the Philippines,
Singapore, Malaysia, and expanding to Spanish speaking Latin America also.
So, I would say we’re early. We’re continuing to hear encouraging feedback as
we expand our tests. Businesses are telling us that the AI saves them time,
while consumers note that they get more timely response times to their
inquiries. The new Business AI product group that we’ve stood up is really in
service of accelerating our efforts to scale business AIs. We’re making
investments across the board in technology and go to market efforts and our
partner ecosystem.
We are really focused on helping businesses use AIs to drive ad conversions, in
terms of the use cases that we want to expand into. So, we are looking at how
we can activate AI and chats to support commerce use cases, to help generate
or qualify leads. Our ultimate ambition is to build AI products that businesses
can use across a variety of use cases to deepen their customer relationships,
and to make the entire sort of transaction process more efficient.
But there is, you know, there’s a lot to do in this roadmap. It’s very ambitious.
We have a lot we have to execute on. We have to enable more access to
business data and systems so that the AI can respond and take action
appropriately.
We have to build richer in-thread experiences. So, I think it’s going to take time.
We’re working hard towards this. It’s a very big priority for us in 2025 across all
of these areas. And it’s another area too where we expect it to benefit from the
sort of parallel continued advances in Llama as well, which should further
increase the quality of responses and enable new capabilities.
Operator: Your next question comes from the line of Mark Mahaney with Evercore ISI.
Please go ahead.
Mark Mahaney: OK, thanks. Could I throw a couple of real quick ones by? Does the total
expense guidance for the year include severance costs related to the recent 5%
reduction and how material are those? When you talk about headcount growth
in fiscal ‘25, is it reasonable to assume that what we’ve seen the last two
quarters is kind of high single digit percent is what’s kind of implied in what
you’re talking about?

-- 2 of 10 --

3
And then, if I could ask you just again about DeepSeek, and I liked how Mark
addressed it. It’s hard to see how there isn’t some wonderful efficiency hacks
and cost hacks that have really come out of what we’ve learned about
DeepSeek in the last week and maybe in the last month or two.
And it’s kind of hard to see how you wouldn’t be able to draw on that at some
point and really get kind of either a much greater, well, much greater ROAI from
all of your investment or give you the toggle ability to reduce your AI. Am I
thinking about that incorrectly, Susan? It seemed I’m pretty bullish on what the
implications of this are for your cost structure going forward. But am I wrong?
Thank you.
Susan Li: Thank you. I’m going to try and get to all the components and you’ll have to let
me know if I missed anything. So, the first thing is, does the Total Expense
Guide include the effects of severance related to the announced performance
based layoffs? It does. So, that Total Expense Guide is inclusive of sort of
everything that’s in our current outlook. It also includes the savings related to
the extension of server useful life.
So, it would have been higher without that. I don’t think we have quantified
what the severance related costs are. So, I don’t think we’re breaking that out. I
will transition that into the headcount growth question. So, it’s not a 5%
reduction in force. Those roles are roles that can be backfilled. So, we’re not
expecting that the total size of the company is going to shrink by 5%.
And then we aren’t quantifying the growth rate of headcount in 2025. But
we’ve talked about what areas are growing and where are the areas that we are
being very efficient as far as we can. And then the final part of your question
was about DeepSeek?
Mark Mahaney: Yes.
Susan Li: And yes, sorry, thank you. Just trying to stay on track here. Look, so I think I’m
really mostly echoing what Mark has said. We both think that DeepSeek has
made good research strides on model design and efficiency. To your point, we
are looking at what we can learn from what they have released and where it
makes sense for us to make adaptations. I think we will. So again, we’re looking
forward to learning from them.
At the same time, you heard Mark say, we remain committed to investing in
infrastructure for training and inference. Both because, again, we don’t know
exactly what we’ll need, we don’t know exactly how broad our inference use
cases will be, and we think this is a real source of competitive advantage for us.
So, I both share your excitement that there will be room in the future to run
some of this work more efficiently. But at the same time, I would reiterate my
earlier comments that we are early in these stages of what the CapEx
investments could look like and we don’t really know yet how the entire
landscape will evolve.
Mark Mahaney: Thanks.

-- 3 of 10 --

4
Susan Li: Your next question comes from the line of Stephen Ju with UBS. Please go
ahead.
Stephen Ju: All right, thank you. So, I think -- I don’t want to put words in your mouth,
Susan, but I think throughout the course of last year, you might have indicated
some amount of frustration that your growth rate for advertising might have
been slowed down by capacity constraints. So, is it kind of fair to think that as
you put up this CapEx budget for this year, I guess the shackles kind of come
off a little bit in terms of what you’re going to be able to do, to grow your ad
franchise.
And also, I think what you’ve also spoken about how culturally in some parts of
the world there’s a greater propensity for users to be chatting with merchants
on WhatsApp and Messenger, et cetera. So, is that kind of behavior something
you need to drive in some of the more developed markets where we may not be
doing as much of that to make business AI more useful? Thanks.
Susan Li: Yes. Thank you, Stephen, for the question. So, on the capacity constraint side,
we expect that we are going to alleviate a lot of our capacity constraints in the
first half of this year as we are bringing more data center space online, and
we’re also working to increase the efficiency of our workloads to free up more
capacity. But we’re also just focused on other investments in improving ad
performance more generally, too.
On the second question, about sort of cultural differences and organic
commercial behavior and the degree to which there is an organic business
messaging ecosystem, I think you are right.
You probably noticed that the comments that I, in my earlier comments about
where business AIs have already launched, that reflects the markets where we
know that there is a lot of this behavior already, and we know that we’re going
to need to invest more aggressively to figure out how to go to market in
markets that have for various historical reasons not evolved down that path.
And so, that is one of the challenges, I think, in standing up business messaging
more broadly, but also one of the places where we think there’s opportunity for
AI to really bring down the cost of the business end of that conversation in a
way that makes those conversations and engagements scalable for businesses
in developed markets over time.
Operator: Your next question comes from the line of Youssef Squali with Truist
Securities. Please go ahead.
Youssef Squali: Thank you very much. Hi, Susan. In the weeks and days leading up to a TikTok
shutdown, even though I think it only lasted technically about 14 hours, did you
see any change in advertiser behavior and/or their spend on the platform?
And just a housekeeping question on gross margins for ‘25. I know you don’t
guide, but with the material growth in CapEx planned for the year and the

-- 4 of 10 --

5
extension of the useful life of servers to, I think, five and a half years, how do
you think that will impact gross margins for the year? Could they -- could those
two offset each other?
Susan Li: On the first question, as you noted, the actual TikTok outage was very short. So
the impact was not material, certainly, in our Q1, either revenue or engagement
metrics. And I don’t think we really have more specifics to share on that
front. Our focus really is just on building the best services for our community
and driving growth through that work.
And Mark articulated a quite ambitious roadmap across all of our GenAI core
technology and product bets this year. So I will – and then Chad is going to
take the second question.
Chad Heaton: Yeah. So in terms of the gross margins, we’re not sharing a gross margin
outlook for 2025. I will say the primary driver of cost of revenue growth in
2025 is going to be infrastructure costs, followed by Reality Labs.
Operator: Your next question comes from the line of Shweta Khajuria with Wolfe
Research. Please go ahead.
Shweta Khajuria: Thanks a lot for taking my questions. Actually, I have a question on Reality
Labs. So Susan, where are you in terms of spend? Well, first of all, what are the
major spend buckets, if you could articulate, that you expect for 2025, and
where are you in terms of the magnitude of spend for Reality Labs? Is it fair to
think that maybe the spend -- Reality Labs spend is peaking? Thanks a ton.
Susan Li: Thank you for the question. So, broadly speaking, we think our Reality Labs
investments are kind of spread across two categories; Metaverse, which
includes the VR, MR, and social platform initiatives. And Wearables, which
includes our AR and AI glasses work, and we expect in 2025 that roughly half of
our Reality Labs investment will be across the Wearables initiatives and the
remaining half on Metaverse.
And I would say in each of those areas, there are still -- there are big product
bets we’re making and there’s sort of a deep technological set of challenges
that we know that we need to unlock in order to be able to bring scalable
consumer products to market, which is ultimately what we’re really focused
on. So, it remains one of our core investment priorities. We do anticipate that
operating losses in Reality Labs will increase in 2025 as they did in 2024.
In terms of the increase, we expect our Wearables devices to be the primary
driver of Reality Labs operating losses growing in 2025 across both cost
revenue and operating expenses. And that really comes from our efforts to
further accelerate the adoption of our AI glasses products.
Operator: Your next question comes from the line of Dan Salmon with New Street
Research. Please go ahead.

-- 5 of 10 --

6
Dan Salmon: Great. Thanks. Good evening, everyone. So, Susan, you mentioned, obviously,
there’s some gray areas that can exist between different categories of CapEx -
sort of older generations of GPUs can be used for core functions. But if I come
back to that definition of, I think you called it sort of AI infrastructure, core AI
infrastructure, non AI infrastructure, can you give any color on how you define
those buckets a little bit more?
And then I may have missed the detail on the call, but does the five and a half
year useful life apply to servers in all of those groups?
Susan Li: Yes, so the way I would roughly think about it is Gen AI capacity is the capacity
that we are putting towards primarily right now training our next generations of
Llama models. So, it is mostly training capacity, but we anticipate and hope
that there will be inference, a lot of demand for inference capacity in the
future. Core AI really refers to the capacity that is powering our ranking and
recommendation engines that underly both the organic content experience.
You’ve probably heard us use the word discovery engine to describe that as
well as the ads experience. So, basically figuring out which of the many pieces
of ads and organic content any given person will most likely find interesting and
useful in their experience. And then non-AI refers to the things that are not in
those two buckets.
For example, the capacity that it takes to serve you a video stream, that sort of
thing would fall into non AI. So, those are the three buckets. And I think I
shared color on the call that we expect total infrastructure spend within each of
those buckets to increase in 2025 with the majority of our CapEx going
towards the core business.
But, again, as you alluded to, it is not a perfectly clear allocation. And the useful
life extension does apply across all of the categories.
Operator: Your next question comes from the line of Michael Nathanson with
MoffettNathanson. Please go ahead.
Michael Nathanson: Thanks. Hey, Susan. I have two for you. One is an answer to that question
of Dan’s about you defining some of the spend.
I wondered, has there been -- can you help us understand where the capital
intensity has rolled out by geography to help on the rankings and ad experience
side? Because I’m trying to look at the average pricing; I know there’s a lot
going on there.
But is there any that you can add of geographies that you’ve rolled out these AI
initiatives and where we’re seeing it? That would be very helpful. And
secondly, on Reality Labs, how are you measuring the ROI, right? I know there
are a lot of nascent ideas here that have never been done before. So how are
you as a CFO testing whether or not the spend is working or not
working? Thanks.

-- 6 of 10 --

7
Susan Li: Thanks, Michael, for the question. So on your first one, I would say -- I’m trying
to make sure I’m approaching the question correctly.
So in terms of where CapEx has rolled out or is being deployed by geography, I
would say in the core AI use cases, which is where it’s being used to do things
like improve the next generation of ads architecture models, where it’s being
used for something like Andromeda, which I described in the call, that’s not a
geographically isolated thing.
So generally, features that we release to improve ads performance, these sort
of ranking and recommendation systems, the delivery back-end, those sorts of
things -- outside of testing, tend to be rolled out broadly across the world.
The GenAI initiative specifically, the business AIs, which we talked about
earlier, the tests are actually being done in, again, these markets where
business messaging is already up and running and where we already have a
thriving click-to-messaging ads ecosystem.
So that tends to be more like Southeast Asia, Latin American countries. But a
lot of the sort of fundamental work that we’re doing to continue investing in
ads growth, things like making sure that we are rolling out different ads
features that might be vertical-specific to other verticals, things like
that. Those don’t have any particular geographic bounds on them.
On the second question, which is about Reality Labs, I think that this is a place
where I think Mark has kind of articulated that the vision for Reality Labs really
is about building the next computing platforms, right? And to the extent that
we succeed at doing that, both that will give us, I think, a really important
position in terms of shaping the next computing platform and building
ecosystems on top of it that we think will be very complementary to the way
the Family of Apps works today.
And I think building businesses on top of that will be also, I think, a natural
extension of both how our ad business works today, but you can imagine that
there will be also new businesses unlocked around digital items and things like
that.
But in the nearer term, I think one of the things that is driving the way we think
about the 2025 investment is we are really more bullish on the importance of
wearables specifically, right? And I think that has pulled our timetable on the
next computing platform up. So we’re really expecting to continue to invest
heavily in the wearables category to help accelerate that.
And Mark has outlined his goal of 5 to 10 million units on the next
generation. And then eventually that gets you hopefully on a path to hundreds
of millions of units and to really a scale consumer product. But it is a place
where obviously we are paving the way in the market, and that requires
investment on our part.

-- 7 of 10 --

8
Operator: Your next question comes from the line of Mark Kelley with Stifel. Please go
ahead.
Mark Kelley: Great. Thanks very much. I’ll just keep it to one. I was hoping maybe, with RFK
Jr. going through this confirmation hearings process, we’ve been asked about
pharma advertising across the digital companies. Is there any way for us to
think about your exposure to pharma overall? Certainly, I know there’s a lot we
don’t know yet. But any way for us to frame that, if things do change in terms
of how pharma companies are able to advertise? Thank you.
Susan Li: I don’t think that’s something that we have shared specifically. Now, I would
just say broadly, that we have a very diversified advertising business across a
very large range of both geographic and vertical exposure. So, I wouldn’t say
that this is something that’s weighing on our 2025 outlook.
Operator: Your next question comes from the line of Rich Greenfield with LightShed
Partners. Please go ahead.
Rich Greenfield: Hi. Thanks for taking the question. I guess this is sort of philosophical and
expands upon how Mark opened up the call. He talked about getting to the
billion users. And as you think about the AI landscape, Susan, so much of the
Internet and digital has been winner-take-most. How do you think this plays
out?
I mean, again, just in the context of what’s happened with DeepSeek over the
last five weeks. Alibaba announced today. What is the framework that we
should be -- or that you think about? Forget about what we think about. What
do you think? How do you think this landscape and AI plays out in terms of,
what share goes where? And is it different than past digital and Internet cycles?
Susan Li: That’s a really interesting question. I think the part that I feel most equipped to
speak about is I think we have a lot of plausible paths to build businesses on top
of the leading AI technology. And those businesses are in some cases
extensions of what we already do.
So, whether it is, the fact that GenAI will come up with a lot of fundamentally
new and easier ways to create compelling content for our ecosystem is clearly
sort of very accretive to the experience on the Family of Apps. I think our ability
to make our ads better, for example, using sort of these technologies to help
advertisers run extremely customized campaigns at scale. Whether it’s to build
a richer business messaging ecosystem, or whether it’s, Meta AI, which is our
consumer version of the experience.
I think there are a lot of avenues for us to really build both adjacent businesses
that intuitively make sense given where our existing businesses are, and also
where the size of the opportunity is really meaningful.
I think it’s probably harder to opine on how the rest of the landscape plays out.
But I feel very good about the opportunities that we have to really build great
businesses in this landscape.

-- 8 of 10 --

9
Operator: Your next question comes from the line of Deepak Mathivanan with Cantor
Fitzgerald. Please go ahead.
Deepak Mathivanan: Great. Susan, can I ask a couple of questions? So first, Clara, obviously, is a
huge addition to the team. Can you discuss some areas where Clara and her
team can have a big impact in the near to medium term on the enterprise and
SMB side? Do you think there’s opportunities to monetize Llama maybe as
APIs or subscriptions or more directly through some cloud services? So that’s
the first question.
And then wanted to also ask about the disclosure on Advantage+ Shopping
revenue run rate. It’s about 10% of the total revenues as you roll out the
streamlined campaign creation as the default option, to a broader set of
advertisers. Is that something that’s going to be relevant for all types of
advertisers or certain specific categories? Basically, just trying to figure out
how much performance gains that you saw with Advantage+ can scale further.
Thank you so much.
Susan Li: Thank you for the questions. So I’ll start with your second question first, on
Advantage+ shopping. I think you were asking about the new simplified
campaign setup flow that we began testing in Q4. That basically combined
Advantage+ shopping and Advantage+ app campaigns with the manual setup
flows.
So in the new design, there’s now one campaign creation flow. Advertisers get
the performance benefits of Advantage+, by default. Previously, they had to
choose between setting up a manual or automated solution. And they’ll still
have the access to features that were available previously if they prefer to
manually adjust settings.
But in general, we think both the new simplified flow is going to help
advertisers leverage the performance gains enabled by Advantage+, and then
we also plan to expand these tests to more advertisers in Q1 before we roll it
out more widely in the following months.
But again, this is an area where we’re generally excited about because we have
seen, at least in our studies, that advertisers see meaningful gains in return on
ad spend, after they adopt Advantage+ shopping campaigns. So, we’re
optimistic that helping them do that more easily will benefit their performance.
Your first question, which was about the addition of Clara to our team, we’re
very excited to have her. Her initial focus is on universal business agents, which
we believe will combine our strengths in AI with our messaging platforms like
WhatsApp and Messenger.
And, I think I alluded to this earlier, but this product group that we have stood
up that she is leading is, is really going to, we hope, accelerate our efforts to
scale business AIs. And it’s where we are, again, investing in both the

-- 9 of 10 --

10
technology, the go-to-market efforts, making sure that there’s a rich partner
ecosystem, et cetera. So we’re very excited to have her on board.
Kenneth Dorell: Krista, we have time for one last question.
Operator: And that question comes from the line of Aaron Kessler with Seaport Research.
Please go ahead.
Aaron Kessler: Great. Thank you. Maybe just on the Q1 revenue guidance, any more color
there? Seems like it’s maybe a little bit conservative, just given the strength
that we saw in Q4. It looks like you do have a little bit of a tougher comp from
last year, Q1, as well as FX. So is that the right way to think about maybe why
the guide wasn’t a bit higher? Thank you.
Susan Li: Yes, I mean, there are a range of outcomes embedded in the Q1 outlook.
Overall, it reflects our expectation for continued strong constant currency
revenue growth. But the largest factor driving the decel on a reported basis is
currency, so that we expect to be a three-point headwind in Q1 after being
approximately neutral to revenue in Q4, just with the Dollar strengthening, in
particular against the Euro.
Beyond currency, we are lapping a period of stronger demand in Q1 of last year,
which included a benefit from leap day in 2024. And political ad spending is
also not expected to be the same tailwind to year-over-year growth as it was in
Q4 since we’re coming off of the U.S. election.
So, our outlook reflects all of these impacts, but overall, ad demand remains
healthy. We continue to see opportunity to drive strong revenue growth
throughout 2025, and we’re excited about the investments that we’re making
in our core business. And I think with that, we’re going to wrap, so thank you,
everyone, and happy Lunar New Year.
Operator: This concludes today’s conference call. Thank you for joining us, and you may
now disconnect your lines.

-- 10 of 10 --