TRANSCRIPT

Q3 2024 — Follow-up call transcript

Q3 2024PDF · 10 pages78 KB
1
Meta Platforms, Inc. (META)
Third Quarter 2024 Results Follow Up Call
October 30th, 2024
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 Third
Quarter Results Follow-up Q&A Call. All lines have been placed on mute to
prevent any background noise.
After the speakers’ remarks, there will be a question-and-answer session. If
you would like to ask a question please press star 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 at
investor.fb.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 touchtone 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 Mark Shmulik with Bernstein.
Please go ahead.
Mark Shmulik: Yes. Hi. Thanks for taking the questions. Susan, a couple. Just I appreciate the
kind of color around the engagement traction with Meta AI against kind of the
500 million MAUs. Wanted to think about how the engagement is growing
relative to kind of the MAU number, something like a WAU/MAU.

-- 1 of 10 --

2
And then second question is a bit more mechanical. Just on the guidance
around the tax rate, how do we think about that just kind of going forward into
2025 and some of the puts and takes around that? Thanks.
Susan Li: Thanks, Mark. So your first question was around more color on engagement
relative to Meta AI.
So I don’t think we have shared the kind of retention metrics that you are
describing, sort of like daily or weekly actives relative to monthly actives. I
think we’ve been talking about it more qualitatively so far.
We're happy with the growth overall. We’re continuing to see particularly
promising signs on WhatsApp in terms of retention and engagement. India is
and remains our largest market for Meta AI usage.
And it’s interesting because we see that people engage with Meta AI from a
variety of different entry points. The deepest engagement occurs via
messaging, although we also are seeing some strong traction with Feed deep
dives. And people are also using it in group chats and creating images to share
as posts or stories.
Your second question was around tax. I don’t think we have given an outlook
beyond Q4 and we shared that our Q4 tax rate will also be in the low teens.
One of the things that is an interesting dynamic in our tax rate is our tax rate
tends to be reduced when our share price increases. That’s because the tax
deduction for RSUs that we issue as compensation is based on the vesting
value while GAAP recognizes the expense based on the grant value.
So you saw that our tax rate through 2024 has been lower relative to what it
was in 2023, and that, in part, reflects a higher share price than we had in ’23
and also, in part, due to some R&D tax credits this year.
Operator: Your next question comes from the line of Colin Sebastian with Baird. Please
go ahead.
Colin Sebastian: Thanks, guys, for taking the questions. I guess, Susan, first, on the decision not
to provide more details on 2025 spending, how much of that, I guess, if you
could elaborate, is really just digesting some of the near-term resources that
need to be put to work on all these initiatives? Or is more an effort to drive
more efficiency or find more ways to find efficiency I guess in the business?
Susan Li: Thanks, Colin. So I believe we previewed several quarters ago the decision to
switch when we give forward-year guidance from what was historically this
call, so the October or Q3 call, to the Q4 or January call. I think maybe we even
previewed that a year ago?
Kenneth Dorell: A year ago.

-- 2 of 10 --

3
Susan Li: It was a year ago. Yes. So that’s been a long awaited change. But the reason
we made that change back then is because our budgeting process is quite
dynamic and we found that we really just had a much higher fidelity guidance
by the time that we were done with the budget, which gets wrapped at the end
of the year rather than kind of in this October like midway point.
So, that’s really the sort of -- the change in timing was because of the timing of
our budgeting processes and we previewed that a year ago.
Having said that, you know, we have spoken a bit qualitatively about the shape
of 2025 planning. We have talked about the growth and CapEx that we expect,
you know, given the sort of capacity ambitions that we have and what we’re
building in order to fuel sort of our AI and non-AI work for next year.
And then we’ve also talked about -- Mark gave some color in his remarks and I
spoke a bit about this also on how we’re thinking about the headcount planning
process, where we think we want to make sure that we’re investing
appropriately to sort of maximize the opportunities for return across a portfolio
of projects.
So, and I think both he and alluded to where those priority areas are and how
there’s kind of an ROI bucket that we think about and what the strategic
bucket looks like. So, all of that will factor into the final 2025 budget, and then
correspondingly the guidance that we give in the January ’25 call.
Operator: Your next question comes from the line of James Lee with Mizuho. Please go
ahead.
James Lee: Great, thanks for taking my question. Susan, I was wondering if you can give us
an update maybe on AI agents of business. Any update for potential
commercial launches for Messenger or maybe WhatsApp? I think last quarter
you talked about testing business agent in Southeast Asia and India. And just
wondering, do you have anything new to call out? Thanks.
Susan Li: Yes, thanks, James. So, we’re continuing to progress with of tests of AI
agents. It remains very early but we’ve recently just expanded availability to
businesses using Click-to-Message ads on WhatsApp and Messenger in English
across a handful of markets, I think the focus has been still primarily in Asia, I
think India and Singapore on WhatsApp, and the Philippines, Singapore,
Malaysia and the UAE on Messenger.
Our focus has been to start to support shopping use cases such as responding
to people who are asking for more information about a product or about its
availability. We’re also expanding to support (lead gen) to help businesses
generate high-quality leads.
But we’re very, very early. This is a place where I would say we are really
focused on learning from these tests, in terms of what makes the business AI
agents most effective and what makes them -- how the businesses who deploy

-- 3 of 10 --

4
them are measuring success and how we can drive the most value with this use
case.
So, again, this is very early in our journey. There is a lot that we’re working on
going forward; the response quality we expect will continue to improve both
with our ongoing work along with the next generation of Llama models.
We expect the scaling to languages beyond English will be an important part of
the growth here going forward. And ultimately, also expanding the types of
responses that business AIs can support, improving the quality of those
responses and fine-tuning the responses based on what we learn from kind of
the current test data, and building richer in-thread experiences.
So, all this is to say there is a lot of work that is underway, but the long road of
the business AI journey is ahead of us.
Kenneth Dorell: Krista, can we go to the next question?
Operator: Your next question comes from the line of Michael Nathanson with Moffett
Nathanson. Please go ahead.
Michael Nathanson: Thank you. Susan, I have one hard one and one easier one. I wonder when
you talked about the near-term opportunities to improve recommendation and
then monetization, how quickly do you think you see the ROI on that CapEx
spend? Is there anything you’ve done on kind of when do we start seeing it,
perhaps the results or something there?
And then the easier one is think about the CapEx spend this year. Is there
anything you can help us with between maybe shorter cycle depreciation like
servers versus how much of this spend, percentage wise, will be on longer cycle
things like infrastructure? So anything there would be helpful. Thanks.
Susan Li: Thanks, Michael. So across the set of things that we’re investing in with our AI
investments, with our core AI work, we’ve talked a lot historically about how
we have a very ROI-based approach to our investments here.
We’re still seeing strong returns as improvements to both engagement and ad
performance translate to revenue gains. I think Mark mentioned that this year
the ranking improvements we’ve made have delivered an 8% increase in time
spent on Facebook, a 6% increase on Instagram.
And there are a number of other opportunities across our core AI work that we
see to continue delivering good ROI against those investments.
And the Gen AI work, obviously, is much earlier, but even then, we expect
they're going to open up avenues for us to accrue value to different parts of our
business, with those happening along different timeframes, too.
And so, in the Core Ads business, the Gen AI tools that we have built here that
will help us enable businesses to make ads significantly more customized at

-- 4 of 10 --

5
scale, which is going to accrue to ad performance, that’s a place where, again,
we’re already seeing promising results in both performance gains and adoption.
I think we shared that over a million advertisers use our Gen AI ad tools
specifically.
So there are a lot of opportunities even in the sort of nearer-term roadmap for
the deployment of Gen AI tools and technologies to make the Core Ads
business better, again, that we’re excited about in the nearer term.
So I would say there are a good number of opportunities of that nature core AI
ranking, recommendation, improvements and deployments of Gen AI within
the Core Ads business that we expect are going to deliver ROI well before some
of the longer-term Gen AI bets around business messaging, Meta AI, et cetera.
So that’s the first question. Your second question is…
Chad Heaton: Yes. I’ll pick up the second question about your infra CapEx. So I think Susan
mentioned here, we’re still working through our 2025 infrastructure plan, but
at this point we expect total infrastructure spend within each of non-Gen AI,
Gen AI, and core AI to increase in 2025.
We expect servers will be the largest driver of growth in 2025 and remain the
largest portion of our overall CapEx budget, though we also expect higher
spend in data centers and network equipment as we continue to just scale our
overall infra footprint.
As I mentioned, servers are going to be the largest driver of growth. We expect
growth in non-AI capacity as we invest in the core business, including to
support a higher base of engagement as well as refreshing existing servers, as
well as AI capacity as we scale Gen AI training and continue to invest
meaningfully in core AI.
We do anticipate data center spend to be higher in 2025, primarily due to us
entering the core construction phase of our smaller, higher-density data
centers that we’ve discussed previously. And we’re building higher capacity
networks to accommodate the growth in Gen AI and Core AI-related traffic in
2025, along with investing in fiber to handle future cross-region training traffic.
So, those are some of the drivers that we’ll be expecting on the infra side in
2025.
Susan Li: And, actually, just before we move on, I want to come back to my first answer.
I realize I kind of answered more with a focus on the monetization bit
specifically. You also asked about recommendations and I wanted to sort of
elaborate here and to just say, we think that we have actually a lot of
opportunity to improve recommendations, both in near-term and longer-term
projects. But some of the nearer-term opportunities are really around
relevance gains as we scale our models for content recommendation.
So the first is increasing model complexity. That’s going to enable us to learn
more from the same data with more compute. And the second is expanding
data sets so that we can learn from more data. And we’re also just making a lot

-- 5 of 10 --

6
of optimizations to the existing recommendation systems that help drive
incremental improvements and efficiency.
So, on Instagram in Q3, for example, we completed infra-efficiency projects
that consolidated some of our existing recommendation systems, that lets us
concentrate our future improvements on a smaller number of recommendation
models, it drives higher engineering efficiency going forward.
And then in the longer term, there are a lot of new model architectures we’re
working on. We have benefitted, frankly, over the past few years from
integrating new model architecture into the Facebook Reels recommendation
system. We talked about the watch time gains last year when we switched
from CPUs to GPUs for Facebook Reels. And we’re continuing to explore how
to, basically, integrate these types of models to the recommendation systems
for other surfaces.
And again, I think that is a place where we expect that we will be able to make
our recommendation systems deliver more relevant and interesting content.
And that, again, is something that is, for us, we are able to measure how
effectively we can do that and how much we can drive growth and engagement
over time.
Operator: Your next question comes from the line of Daniel Salmon with New Street
Research. Please go ahead.
Daniel Salmon: Good afternoon, everyone. Susan, on the call, you called out the growing use of
Advantage+ creative tools and a wider expansion for them coming next year.
How would you compare the revenue potential of that module or tactic to the
impact that you’ve seen from Advantage+ shopping campaigns, which you
noted is driving over $10 billion of revenue a few quarters ago? And then just
second, any quick update on the traction for Advantage+ app campaigns that
you’re seeing? Thanks.
Chad Heaton: Yes, I can share a little bit about Advantage+. We are seeing continued growth
with our single step and our end-to-end automated solutions, that includes
Advantage+ shopping, and as you mentioned, Advantage+ app campaigns. On
the Advantage+ shopping side, growth remains really strong.
We’ve seen increased adoption since we rolled out the extended list of
conversions businesses can optimize for in Q2. On the Advantage+ app ads,
we’re seeing growing adoption of app campaigns and particularly a success in
the gaming vertical, which is a stronger vertical for app objectives in general.
In Q3, we continued testing a new end-to-end automation solution that
supports lead-gen campaigns. So, Advantage+ lead campaigns are sort of
similar to Advantage+ shopping campaigns. It enables advertisers to apply AI
to multiple campaign levers at once for ads using the lead’s objective, including
audience-created placements and budget.

-- 6 of 10 --

7
So overall we continue to see a lot of opportunity in our Advantage+ suite of
solutions. We also expect the features that we’re introducing to drive results
more customized to each business objectives, like conversion value rules, will
enable us to better optimize AI-enabled campaigns. So yes, overall it’s going
really well and continues to be a driver of our business.
Operator: Your next question comes from the line of Shweta Khajuria with Wolfe
Research. Please go ahead.
Shweta Khajuria: Thank you so much for taking my question. Is it possible to talk about the DAP
result in this quarter, anything in particular, did it come in ahead, in line, or
below your expectations? And something -- because it also impacts
impressions growth, so any color on that would be great, thank you.
Susan Li: I mentioned this on the call, but we continue to see daily usage grow year over
year across Facebook, Instagram, and WhatsApp in Q3, both globally and in the
U.S., so we feel good about the engagement trends overall both from a usage
perspective, as well as deeper engagement metrics in the way people spend
time on the platform, and also our ability to continue to improve our ranking
systems and deliver good content and video recommendations that make
people’s experiences more interesting and valuable.
Chad Heaton: And then in terms of impressions, we had 7% year over year impression growth
that was driven by users, as Susan mentioned, but also video engagement
growth across Facebook and Instagram.
So I think we felt good about the impression growth. As we’re looking forward
into 2025, we continue to see impression growth opportunities through
engagement growth and ongoing ad load optimizations for our video products.
So yes, we’re feeling good about the opportunities here.
Operator: Your next question comes from the line of Rohit Kulkarni with Roth Capital.
Please go ahead.
Rohit Kulkarni: Hey, thank you, couple of questions, one on depreciation and the lifetime of
servers. I think you guys are the only ones who are depreciating it over five
years, most of the servers and networking equipment -- a few of your peers are
extending lives to six years. So maybe talk about the decision-making behind
that.
And second is on Reality Labs and the OpEx for Reality Labs. It has
consistently increased as a percentage of overall OpEx. With Orion out there,
the hypothesis is that perhaps the velocity of spend may start to slow down a
little bit. Would love to get your reaction to that as well as how do you think
about the decision-making around increasing OpEx from Reality Labs into the
next couple years?
Susan Li: Thank you, Rohit. So the first question around useful lives on servers, we don’t
currently have plans to extend the useful lives of our servers. That’s really
because, given the ongoing performance gains from new generations of GPU-

-- 7 of 10 --

8
based servers, we really want to make sure that we are making the best use of
the capacity available in our data centers by using the latest generation of
servers.
But we do periodically evaluate server useful life, and when we do, we’ll take
into consideration the historical data of how long we used them for and future
looking data on server retirement plans. But again, no current plans to extend
the useful lives of our servers. The second question you had was around…
Chad Heaton: Reality Labs.
Susan Li: Reality Labs and the way we think about the investment in Reality Labs. I
would say we are -- we have a really big investment portfolio here. It is
multifaceted across both the current and next generation of hardware devices
and software.
In the near-term, we’re really focused on growing adoption of our existing
products. We want to accelerate the adoption of products like the Ray-Ban
Metas that we’ve seen have very strong product market fit and that, frankly,
we think are accelerating the timeline on sort of bringing AR glasses to market
and are just an extremely natural form factor for a lot of our AI tools.
You kind of talked about Orion, and much of our investment, of course, is in
areas that are not currently reflected in products on the market. There’s still a
lot to do around conducting the fundamental R&D to invest and productize and
scale new technologies. And there is certainly a lot we need to do on Orion
before we think it’s ready to be a product that we would bring to market in
terms of making both the actual display quality better, the form factor more
comfortable, and making it more affordable to build at scale.
So there is -- it's just a big portfolio in Reality Labs. It’s a place where, while the
ambitions are significant and, again, span a lot of products, it’s also a place
where we’re focused on efficiency, too, and that is definitely a part of the
budget conversation, which is where are the places where we can be more
efficient in Reality Labs such that we have room to expand our ambitions
around things like smart glasses and accelerate bringing those to market and
trying to find tradeoffs as we think about the Reality Labs portfolio.
Operator: Your next question comes from the line of Jason Helfstein with Oppenheimer.
Please go ahead.
Jason Helfstein: Thanks for taking the questions. I just want to ask a bit about gross margin.
With the significant CapEx year-to-date and at the end of last year, you’ve
been able to offset that as far as impact on gross margin through just really
strong revenue. I mean, obviously you're not going to give us guidance, but
should we start to think about gross margin being negatively impacted by kind
of CapEx?
And I don’t know if it specifically starts in fourth quarter, but as we move
through next year, just maybe how should we think about gross margin I guess

-- 8 of 10 --

9
going forward? And then just quick housekeeping and you may have said it in
the main call and I missed it, but just why did G&A step down by like a little less
than $2 billion sequentially? Thanks.
Susan Li: Yes. I’ll get to your G&A question first. So the decrease in G&A, that really was
driven primarily by lower legal-related expenses. There’s kind of lumpiness in
G&A on a quarter-to-quarter basis due to the nature of legal accruals and
expenses, and in Q3 we just didn’t have the same level of legal-related
expenses that we had in the first two quarters of the year.
And then your first question about gross margins, obviously there are multiple
components. We’ve talked about the depreciation expenses that we expect to
accelerate significantly as the impact of the CapEx build to date and next year’s
incremental CapEx flow through.
The cost structure, we’re not breaking that out specifically between cost
revenue and R&D, but we’re generally investing across our infrastructure fleet
and the cost of revenue, infrastructure costs do capture all the depreciation
and OpEx of our infra operations outside of what’s used for AI research. So,
that’s the sort of piece that is specific to Gen AI training, Reality Labs, and
FAIR and that goes into the R&D bucket instead. And then, of course, how
successfully we’re able to deliver returns and drive revenue growth in 2025.
Kenneth Dorell: Krista, we have time for one last question.
Operator: Your last question comes from the line of Tom Champion with Piper Sandler.
Please go ahead.
Jim: Hi, thanks, this is Jim on for Tom. Appreciate you guys taking the question.
Gaming was called as a vertical of strength, not this quarter but for, I think, the
four quarters prior. Would be curious kind of what drove that and maybe why
that wasn’t a driver of strength in this quarter. And then, secondly, would just
be curious to kind of hear the status of the Messaging business in the U.S. and
the ambitions there. Thank you.
Chad Heaton: On verticals, we don’t typically call out -- we usually focus on which were our
drivers of strength. We called the online commerce vertical was the largest
contributor to year-over-year growth. That was followed by healthcare and
then entertainment and media verticals. I can share a little bit more about
those.
Online commerce, the growth remained healthy but slowed from Q2, primarily
due to lapping accelerating growth from China-based advertisers from a year
ago. On the healthcare side, it was one of our fastest growing verticals in Q3.
We saw broad-based strength across advertiser regions with particular
strength in North America.
And then on entertainment and media, again, saw strong year-over-year
growth there. It was driven by broad-based strength across North America and
Asia-based advertisers, particularly from China. So gaming still is an important

-- 9 of 10 --

10
vertical for us but it just wasn’t one of the biggest contributors in this most
recent quarter.
Susan Li: And on your first -- second question, sorry. Status of the Messaging business
in the U.S. I think we talked about this maybe a little bit on one of the earlier
questions. We are really focused on scaling Click-to-Message ads across the
two levers of both broadening adoption in more markets as well as facilitating
more down funnel conversions.
On that market piece, it is earlier in the U.S. We are seeing good growth in
Click-to-WhatsApp ads. And that is a place where we are continuing to invest
in scaling consumer adoption of WhatsApp and that’s going to create a bigger
opportunity for us with the Click-to-WhatsApp business in the U.S., also
focusing on down funnel objectives just continues to make those ads more
generally performant and is certainly something that also is, I think, especially
valuable to more sophisticated advertisers.
So it’s still early. But our ambitions there are that we are going to invest a lot
here. It’s a market that we care a lot about and we know that there is a strong
business Messaging ecosystem in other markets and we want to do everything
we can to replicate that success in developed markets like the U.S. and
Western Europe.
Kenneth Dorell: Great. Thank you, everyone, for joining us today. We appreciate your time and
we look forward to speaking with you again soon.
Operator: This concludes today’s conference call. Thank you for your participation and
you may now disconnect.

-- 10 of 10 --