TRANSCRIPT

Q1 2023 — Follow-up call transcript

Q1 2023PDF · 15 pages180 KB
1
Meta Platforms, Inc. (META)
First Quarter 2023 Results Follow Up Call
April 26th, 2023
Operator: Good afternoon, my name is Dave and I will be your conference operator
today. At this time, I would like to welcome everyone to Meta’s First Quarter
Results Follow-Up Q&A call. All lines have been placed on mute to prevent
any background noise. To ask a question, please press “1” then the number
“4” on your telephone keypad. This call will be recorded. Thank you very
much. Ms. Deborah Crawford, Meta’s Vice President of Investor Relations,
you may begin.
Deborah Crawford: 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 press release and in our annual report on
Form 10-K 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 may 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 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 the operator for the first question.
Operator: Thank you. We will now open the lines for a question and answer session.
To ask a question, press “1” then the number “4” on your touchtone phone.
Please pick up your handset before asking your question to ensure clarity.
And if you’re streaming today’s call, please mute your computer speakers.
Your first question comes from the line of Brent Thill with Jefferies. Your
line…

-- 1 of 15 --

2
Brent Thill: Thank you. We’ve heard really good things from advertisers on Advantage+,
and I’m just curious what you’re seeing – how important that rollout has
been and any other color you could provide on that. And I had a quick
follow-up.
Susan Li: Hi, thanks, Brent. Yes, I mean, we are hearing good things too. And we’re
also seeing good things in the adoption and sentiment that we are measuring
about Advantage+. Advantage+ Shopping campaigns, I mean, they’ve been
growing steadily since launch. I think Mark mentioned that, in the last six
months we’ve seen a 7x increase in daily revenue.
And in particular it’s such a good tool for ecommerce and retail advertisers,
which are the segments that we originally designed the solution for and are
such a big part of our advertiser mix.
We have been broadening availability to other verticals. We’re seeing good
traction in other verticals, including CPG, and we are, as with all of the sort of
new ad product launches, we try to be rigorous about measuring
performance. And from the studies that we run, we’ve seen Advantage+
Shopping drive lower cost per conversion than similar campaigns run without
it. And so, again, the feedback is positive.
At this point I would say there is like a portfolio of tools that are in the
Advantage+ suite, and we’re continuing to iterate on those, but there is
Advantage+ Shopping Campaigns, which allowed advertisers to no longer
need to manually test campaigns with different targeting, creative placement
and budget setups. We test combos for them.
We have Advantage+ app campaigns, which simplifies the app install
campaign creation and drives performance for app advertisers by using real-
time learnings to adjust ads across audiences, placements, and creative.
We have Advantage+ Placements, which is a tool that finds the most
effective placement for a given ad, shows ads across a handful of possible
placements, and gives our system more flexibility to control costs and drive
results for those advertisers. And now we have Advantage+ Creative, which

-- 2 of 15 --

3
automatically adjusts ad creative for each person that views your ad and
helps show the version that they're most likely to respond to.
So it’s a suite of products that we’re continuing to innovate, and we feel
good about the traction and adoption that we’re seeing there.
Brent Thill: Thanks, and Susan, just real quick on the Q2 guide. To get back to double-
digit growth, which is assumed at the high end of your range, are you seeing
a better macro backdrop, a similar backdrop? Like what needs to happen for
that to happen?
Susan Li: Yes. I mean, look, ultimately Q2 is an easier compare, right? So if we look at
Q2 of 2022, that was the first full quarter without Russia revenue after the
beginning of the Russia-Ukraine conflict, and that was sort of the beginning
of an accelerating period of weaker demand, and we have lesser currency
headwinds.
So there is just some math here that is making the comparison easier.
Having said that, there is a wide range of possible expectations, including
macro outcomes that are captured in the Q2 outlook, and the high end of
that reflects better expectations for what that demand landscape could look
like, but there’s a lot of uncertainty.
Brent Thill: Thank you.
Operator: Your next question comes from the line of Colin Sebastian with Baird. Your
line is open.
Colin Sebastian: Oh, great. Thank you. Deborah, good luck with the next move, and it’s been
great working with you at Meta.
Two questions on the expense side. I guess, first off, on Reality Labs just
contemplating kind of the operating losses and the long-term vision, was just
hoping for some maybe detail on how productivity efforts or the
restructuring efforts help reduce the intensity of annual investment required
in that particular group?

-- 3 of 15 --

4
And then secondly, I know this may be hard to answer, but trying to sort of
splice out the cost of the AI initiatives and particularly the generative AIs
since you’ve had AI for years, but this incremental AI, is there any way to
quantify the cost of that whether that’s from an infrastructure or OpEx
perspective? Thank you.
Susan Li: Yeah. First of all, I will echo your appreciation for Deborah, who we’re going
to miss very much.
Chad Heaton: Yes.
Susan Li: This is her very last day here -- or the last week here and her very last
earnings call today, so I am sad to see you go.
Deborah Crawford: Thank you.
Susan Li: Okay, so to your first question on Reality Labs, sort of contemplating the
operating loss outlook and also how our efficiency and restructuring efforts
apply. So we’ve talked about our efficiency efforts as having two broad
components. The first is kind of going through and reevaluating road maps,
deprioritizing projects. Effectively, I would think of that as areas where we
are shrinking scope.
And the second is sort of -- the second effort in the efficiency workstream
has really been around making the organization more efficient, and that has
been work, like spans and layers, making sure that we feel like we have sort
of the appropriate cross functional staffing and that we’re looking to
streamline processes; that’s not really about scope reduction, that is about
just making the organization more efficient and productive.
So when you look at those two efforts, what I would say is that the scope of
our Reality Labs ambitions, especially over the long term, has not changed. I
-- in the short term, there are places where we have made product changes, I
think sunsetting the Portal -- consumer Portal line last year was a good
example of that, but the magnitude of the long run vision and opportunity
and ambition for Reality Labs remains the same. And you’ll see that certainly
play through in the forward-looking investment in Reality Labs.

-- 4 of 15 --

5
Reality Labs did go through the same sort of organizational efficiency
exercise as all of the other teams across the Family of Apps, so they are
certainly becoming more efficient in that process as well, but I do think it’s
important to emphasize here that we think that the long run ambition of our
Reality Labs work is still great, we think we’re building a sort of a -- the future
of computing platforms, and we’re early in the process of putting together
our long-range plan, which helps us benchmark what some of that product
roadmap can look like. But we’ve said that we expect Reality Labs’ operating
losses to increase in 2023, and I’d expect that to be the case also in 2024.
So it’s a big opportunity, it’s very ambitious, it is becoming more efficient,
but the long-run scope of our ambitions I think has not changed.
The second part of your question is around the costs of our AI initiatives,
particularly Gen AI, so this is a really hard place to try and attribute costs.
There is so much AI and machine learning work that underpins already so
much of our core Family of Apps experience and product offerings.
What I would say in terms of the Gen AI piece specifically, that is a relatively
smaller headcount team right now, because it is a nascent effort, so that’s
not very significant in terms of the percentage of headcount at the company,
but it is a place where we are trying to deploy GPU capacity more quickly,
that is captured in the CapEx guide right now, but I do think it’s worth sort of
double-clicking on Gen AI capital intensity.
We are planning to stand up several large scale training clusters in our data
centers. It’s too early to tell exactly how much additional capacity we may
need to build in future years to support these use cases, but it is likely that
Gen AI workloads will be relatively more CapEx-intensive than other AI
workloads, given the large scale compute infrastructure that you need to
train Gen AI models and then what we think will be the relatively high cost of
inference for Gen AI applications once we start deploying those across
various products and services.
So we are well set up operationally to support the work here because of the
investments we’ve made in our infrastructure strategy and the future data

-- 5 of 15 --

6
centers we are building, which should all be setup to support our AI
workloads. But it’s a place where although 2023 CapEx is captured in our
guidance, I would say it’s a little early to know what the impact will be yet on
2024.
Colin Sebastian: Great. Thank you.
Operator: Your next question comes from the line of Tom Champion with Piper Sandler.
Your line is open.
Tom Champion: Great. Good afternoon. Susan, curious if you could just talk about the
linearity of growth in revenue through Q1? And then also just any additional
context on the strength of online commerce as a vertical? I know you
referenced China during the call, but any other comments and maybe the
connection with Advantage+ Shopping? Thank you.
Susan Li: Yes, in terms of Q1 revenue growth the stronger demand that we saw and
especially in e-commerce, which has been such a big vertical for us and was
particularly hard hit, I think, by the pull back in e-commerce spend kind of as
the world reopened post COVID.
So that is a vertical that returned to positive year-over-year growth for us in
Q1. It benefitted from growth among advertisers in China. And so, I think we
feel encouraged there and we are certainly also building a lot of tools and
product offerings to continue to really make it easy for e-commerce
advertisers to optimize on our platform.
The linearity of growth and revenue, I mean we have -- we have sort of a lot
of quarterly seasonality. So -- and I think part of that corresponds to the way
marketers sort of spend their budgets. So I think we saw many of our typical
season trends, I would say, in Q1.
It’s not perfectly linear through the quarter but it does grow relatively -- it
does grow relatively throughout the quarter.
Operator: Your next question comes from the line of Rohit Kulkarni with ROTH MKM.
Your line is open.

-- 6 of 15 --

7
Rohit Kulkarni: Hey, thanks for the question. And best of luck, Deborah. You’ll be missed. A
couple questions. First on CapEx, how quickly the cost of training complex AI
models is coming down and maybe you could comment on that as being one
of the forefront people to do research and AI and how that kind of informs
your kind of longer-term kind of structural CapEx kind of outlook.
And then referring to Mark’s comments on kind of mitigating the direct
impact of ATT changes and improving ad performance, I guess what I would
like to know is when you compare the performance that advertisers are
seeing right now today in Facebook, do you feel Facebook has kind of
completely caught up to what advertisers used to see on pre-ATT days as sort
of performance is concerned or do you see like a pathway towards
advertisers on a like for like seeing comparable results as such.
Susan Li: Hi, thanks for the question. Before I answer, I just wanted to add one thing. I
realize I was thinking about the linearity on revenue growth for Q1 from
Tom’s question earlier. I did want to add that our growth rate in Q1 exited a
bit higher because of lapping the onset of the Russia-Ukraine war. So, that is
part of the math that is factoring into the year-over-year growth rates,
especially going forward.
OK, sorry. Rohit, coming back to your question. So, the first question is
about the cost of complex AI models. Again, I think we’re relatively early
here, so we don’t have a lot of details to share.
But what I would say is, again, I’d emphasize that from what we know right
now, our Gen AI efforts I think we expect to be more capital-intensive than
the sort of core AI work we’ve been doing to date.
The – similar to many AI applications, the Gen AI work requires both training
AI models and then also inference when you use that AI model to serve user
requests. And the training infrastructure allows us to develop proprietary
algorithms and models which then enables the various Gen AI applications
that we build.

-- 7 of 15 --

8
On the other hand, the inference infrastructure is dedicated to the
deployment of models across various products and surfaces. And so, the
training infrastructure is going to scale, starting with the size and number
and complexity of the models that we’re developing, and the inference
infrastructure we expect to grow with the success of our products and the
numbers of users that end up supporting or engaging in these Gen AI
experiences.
So, we are in the process now, I think I mentioned this earlier, of standing up
several large-scale training clusters, that’s just training focused, and when we
begin to build applications and use cases that we hope will engage users at
broad scale then we’ll also need to build out the sort of inference
infrastructure.
So, this is part of why our early outlook is that the Gen AI work may be
relatively more CapEx intensive than some of the other AI workloads that
we’ve done. I’m going to turn over the second question to Chad.
Chad Heaton: Yes, so your second question was about ATT and specifically are we making
progress to getting basically back to a place where we were pre the ATT
changes? And I guess the way I would say is, I think that’s probably the
wrong way to think about it. I think ATT is just the new landscape that all
digital advertisers, including us, operate in. And so, we’re in this new space.
And so, from there our focus is on improving ad performance with
investments in onsite objectives and AI.
Susan talked before about CAPI, continuing to see good progress there. The
Gateway, we’ve had a new Gateway feature, the multi-account feature. So,
those are specific things we’re doing on ATT. And then more broadly, as I
mentioned, the focus on onsite objectives and AI. Within onsite objectives,
the click-to-message ads, lead ads, shop ads, those are all things that we’re
investing a lot in. We’re seeing good traction. And that’s sort of the path
forward for us as opposed to focusing on what is the pre-ATT world look like.
Operator: Your next question comes from the line Ron Josey with Citi. Your line is
open.

-- 8 of 15 --

9
Ron Josey: Awesome. Thanks for taking the call, guys. Susan, I wanted to follow up on
just the messaging component of today’s call and just the strength we’re
seeing in click-to-message and on WhatsApp paid messaging. Can you just
talk more about the opportunities just to tap into not only advertising
budgets, but I think you mentioned customer service and new budgets as it
relates to just click-to-message?
And then, payments went live in Brazil this past quarter, I think. So, any
insights on just how that might add to closed loop attribution, transactions,
do you see that coming through more and more, is question one.
And if you could, just a really quick one. I think you said there’s a supply
constraint in Reels ads formats. And so, I’m just wondering like what you’ve
seen on ad load, are we sort of at ad load and now it's pricing on Reels, or do
you think there’s more room to go on ad load too? Thank you, Susan.
Susan Li: Oh, yes, no problem. OK, so I’m actually going to answer your Reels question
first then we’ll come back to the messaging questions. So on Reels, the ad
load is still lower on Reels than it is on other surfaces, and there is room for
the ad load and Reels to grow, so we’re not at max ad load.
The issue that I was flagging around sort of the structural issue as it pertains
to ads supply on Reels over time is that because of the difference in scroll
speed, it takes much longer to watch a Reel than it does to read a Feed,
Story, a piece of content in Feed or to look at a sort of static Story image.
That means that our current ad load mechanics, which primarily rely on
inserting ads in between pieces of content, would mean that if you held the
mechanic constant, you would end up with a much lower per-time ad load on
Reels than you would in Feed or Stories because of the scroll speed issue.
So that’s what I was getting at. Now actually even on a sort of -- ignoring
that for a moment just on a ads-per-number-of-organic-stories, Reels is at a
lower place, so that’s what I meant when I was saying that there’s room for
Reels ad load to grow. But when Reels ad load is the -- reaches whatever a
mature state is in terms of being apples-to-apples with Feed or Stories, the

-- 9 of 15 --

10
scroll speed issue would still mean that it would be lower monetization per
time.
So that’s a place where we are going to need to think about how to be more
creative about thinking through Reels ad load in the future as we get closer
to something that is a more mature or more steady state ad load, and that’s
something that will, I think, ultimately be an important factor in terms of the
long-run economics of how valuable Reels can be for us. So that is on the
Reels question.
OK, coming back to the -- oh, right -- the click-to-messaging opportunity. So
yes, we are also really excited about this. We’re seeing strong momentum.
We shared last quarter that click-to-messaging had reached a $10 billion
annual run rate. That is continuing to grow even though we haven’t updated
that number, and then even more, I think, exciting for us is that advertiser
adoption continues to broaden, and we estimate that over half of click-to-
message advertisers are exclusively using click-to-messaging ads on our
platform, which means they're really driving incremental demand, which is
so important for us.
So the places that we are investing and scaling -- I think I alluded to these a
little bit earlier -- ads creation, how do we make it easier to set this up in Ads
Manager, investing in native ads creation in WhatsApp, partnering with our
ecosystem of developers who can leverage the hosted API solutions that
we’re offering so they can also build experiences with click-to-messaging ads.
And then we’re really doing a lot of work to drive action further down
funnels, so we’re updating our ranking models across all of our ads, including
click-to-messaging ads, to try to make sure that they sort of are optimally
targeted and as performant as possible.
We are testing the ability for advertisers to optimize now for purchases
rather than just for message replies, so kind of changing the optimization
function to be closer to the sort of ultimate business outcome or objective
that an advertiser is looking for.

-- 10 of 15 --

11
And we’re investing in sort of new ways for businesses to drive the outcomes
that they're looking for. We talked about in-thread payments and other
commerce tools. And then I kind of talked about how this is the sort of laying
the foundation for our much, much more nascent paid messaging business in
terms of via by cultivating a ecosystem in which there is a lot of business-to-
consumer activity and trying to foster that behavior on our platform, we’re
laying the groundwork for our paid messaging ecosystem.
This is very early, but we’re seeing good growth. And we believe there’s a
big paid messaging opportunity here, and WhatsApp has such global reach.
It has over 2 billion daily active users.
So we’re focused on growing this area. We’re sort of updating the pricing
model to offer differentiated pricing for different use cases. We’re trying to
make it easier to buy. I think you alluded to payments in Brazil and people in
Brazil can now find and message and buy from small businesses on
WhatsApp directly. Anyone on WhatsApp in Brazil can try it now, and we’re
making it available to more merchants soon.
And we’re continuing to invest in other in-thread experiences so that sort of
the entire purchase funnel can take place within an app. So I think that’s --
it’s a big opportunity there. Is there anything you would add, Chad?
Chad Heaton: No. I think you covered it.
Susan Li: OK.
Chad Heaton: I guess I would just add, you mentioned P2M in Brazil. Well just note that in
India we have similar end-to-end shopping experiences with JioMart, transit
solutions with Bangalore Metro and more, so similarly powered by the
WhatsApp Business API. So hopefully this is something that we can leverage
beyond this specific Brazil case over time.
Operator: Your next question comes from the line of Brad Erickson with RBC Capital
Markets. Your line is open.

-- 11 of 15 --

12
Brad Erickson: Yes, thanks. I had a couple. So first, just geographically, when you think
about the results and the guidance, can you kind of walk through the
geographies individually in terms of being better in-line or worse than your
expectations? I know you mentioned China, but maybe if you could unpack
some of the others there?
And then second, just on Advantage+, what sort of timeframe and
expectations do you have there for “full or fuller adoption?” Is it something
where within maybe 12 or 18 months most advertisers are going to get a
look at it and adopt it or not, or is this something that’s like a multi year thing
where you're going to see more and more advertisers move over? Thanks.
Susan Li: Chad, do you want to take the regional ad revenue question and then I’ll talk
about Advantage+?
Chad Heaton: Yes. So I’ll kick off with that. And I would just note this is based on user
location. So ads revenue growth accelerated across all regions as pricing
trends improved and the currency headwinds eased from Q4.
I’ll just tick in through the regions quickly. In North America, the acceleration
in year-over-year growth was driven in part by improved demand, especially
from online commerce advertisers and lapping a weaker demand period than
in Q4 as well as the accelerated impression growth that we disclosed.
In EU, the big factors contributing to the acceleration were lapping a
challenging year-ago period and the start of the Ukraine War as well as the
lessened currency headwinds. In APAC, the eight-point acceleration was
primarily driven by lessened currency headwinds. And then finally in Rest of
World, the acceleration of growth was driven by pricing, which is partly
explained by broad based strength from the Middle East and Mexico.
Susan Li: And then on the Advantage+ question, so, I mean, I would say that we are
still rolling out a variety of features. I think I talked about some of them
earlier, and we’re continuing to iterate on these and make sure that basically
where we find the opportunity to deploy more ways to automate parts of the
ad buying, ad targeting, et cetera experience for consumers that we basically

-- 12 of 15 --

13
-- sorry, for advertisers, that we find ways to basically make that as
automated and as easy to use as possible.
There is a suite of different tools. And so, I don't expect that all advertisers
will use all of them. I think some will be better suited to certain subsets of
advertisers than others. There are a number, for example, that are really
focused on Shopping Campaigns that are very much optimized for
ecommerce advertisers whereas there are others that are focused on
creative optimization or audience optimization. So I think it’ll depend a little
bit on an advertiser’s given use case.
We are seeing good adoption with the ecommerce and retail advertisers that
we originally designed the solution for and I expect we will continue to try to
design features and offerings that continue to expand it’s usability for
broader segments of advertisers going forward.
Deborah Crawford: Operator, we’re going to take one last question.
Operator: Certainly. Thank you. That will come from the line of Deepak Mathivanan
with Wolfe Research. Your line is open.
Deepak Mathivanan: Great. Thanks for taking the question. Susan, on Reels still under-
monetizing and your expectations that it likely stays that way until this year,
how much of that is due to the growth and engagement on Reels over the
past like say 12 to 18 months.
Again, what we’re trying to figure out is it taking time to close the
monetization gap because engagement is growing faster, kind of widening
the gap or is it because the demand from advertisers is yet to ramp there,
which one is the bigger driver?
Susan Li: Thanks for the question. So I think -- I just want to make sure I’m being clear,
you’re referring to sort of the overall Reels headwind as opposed to the gap
in monetization per time of Reels relative to more mature Stories.

-- 13 of 15 --

14
So Reels engagement -- you know Reels is actually becoming more
incremental to engagement over time. So it’s been incremental for a long
time, which we’ve been talking about.
But nonetheless, some of the engagement is cannibalizing higher
monetization services but it’s becoming more incremental to engagement
over time and that is really the piece that is enabling us to close the gap,
even though its monetization per time continues to be below that of Feed
and Stories kind of for the structural reasons that we described.
And then in terms of demand for Reels ads themselves, we feel like there’s a
lot of good progress here on -- we’ve talked in the past about the advertiser
demand, making sure that we are unlocking advertiser objectives on Reels.
We’re continuing that work in Q1. We’ve brought Click to WhatsApp ads to
Facebook Reels. And we’re now testing them on Instagram. We’re enabling
advertisers to boost organic images from Facebook on to Instagram Reels ads
with photo albums. So we’re continuing to sort of unlock more advertiser
demand.
As I mentioned last quarter, the majority of our ads objectives are already
eligible for Reels but we’re kind of working on the incremental remainder.
And then in terms of making Reels ads more performant, we’re very focused
on testing improvements to creative optimizations, improving templates that
automatically convert creative that is not yet fully immersive to Reels ads,
improving the quality of low resolutions Reels ads, testing ways to make
Reels more interactive to help more people purchase, including call to action
optimizations that we hope will drive conversion.
So I would say that there’s a lot of good work being done on the sort of
monetization efficiency side, which we hope over time in combination with
the incremental engagement that comes from Reels will help take Reels from
being not only a revenue headwind to neutral by end of this year, early next
to a revenue tailwind going forward.

-- 14 of 15 --

15
Deborah Crawford: Great. Thank you for joining us today. We appreciate your time and we
look forward to speaking with you again.
Operator: And all this concludes today’s conference call. Thank you for joining us. You
may now disconnect your lines.

-- 15 of 15 --