Facebook : Third Quarter 2021 Follow Up Call

Facebook, Inc. (FB)

Third Quarter 2021 Follow Up Call

October 25th, 2021

Operator:	Good afternoon. My name is France, and I will be your conference operator today.
	At this time, I would like to welcome everyone to Facebook's Third Quarter Results
	Follow up Q&A.
	All lines have been placed on mute to prevent any background noise. To ask a
	question, please press one and then the number four on your telephone keypad.
	This call will be recorded. Thank you very much. Ms. Deborah Crawford, Facebook
	Vice President of Investor Relations, you may now begin.

Deborah Crawford: Thank you. Good afternoon, and welcome to the follow up Q&A call. With me on today's call is Dave Wehner, CFO, and Susan Li, V.P. of Finance. Before we get started I'd 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 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 may present both GAAP and certain non-GAAP financial
	measures. The 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 one then the number four on your touchtone phone. Please pick up
	your handset before asking your question to ensure clarity.
	If you are streaming today's call, please mute your computer speakers. Your first
	question will be from the line of Michael Nathanson with MoffettNathanson. Please
	go ahead.

Michael Nathanson:Thanks. Hey, Dave. How are you?

David Wehner:

Hey, Michael.

Michael Nathanson: So firstly, thanks for giving us a lot of time to ponder on the IDFA changes. I remember talking to you last summer. Now that it's upon us what have you learned that may have surprised you?

1

	Clearly things are coming a bit more disruptive given the guide, but where have the
	biggest surprises been? What sectors have maybe been hit the hardest? And any
	update on the rollout of the Aggregated Event Measurement adoption, client
	feedback, anything to help us there? I really appreciate it. Thanks.
David Wehner:	Sure. Yes. I mean, I think we have, I think, been out in front of saying that we
	expected this to be disruptive for the industry and it clearly has been. It's been
	difficult to forecast exactly how advertisers would react to the rollout.
	I would say we've seen the impact probably be a little bit more disruptive than we
	anticipated. I think just the retooling of all of the - all of the targeting and
	measurement to basically work for aggregated events is just it's difficult especially
	for smaller advertisers with smaller budgets.
	So I think it's certainly been - it's certainly been challenging. Obviously the areas
	that are hardest impacted are those with offline objectives. So the two I'd probably
	call out there, online commerce and gaming.
	Online commerce as a vertical continues to grow but it contributed less to our
	growth than in prior quarters, and it has been - I think the ATT headwinds is a big
	part of it, but there's also a broader macroeconomic overlay where we had a big
	acceleration in ecommerce growth last year due to COVID.
	So it's a little bit hard to unpack where we are, kind of what's - how to kind of pull
	out what exactly the different drivers of it are, but we do think the ATT headwinds
	are a big part of it.
	Gaming also has been challenges - particularly challenged this quarter, and app
	objectives took a hit from Apple's changes and from what we've seen from third-
	party reports, Apple's changes seem to have helped shift share to Apple's own
	captive app-install advertising business, so it's clearly been beneficial for Apple, so
	that's -- I think that's something that we've seen.
	I don't know if there's anything you wanted to add on some of the Aggregated Event
	Measurement Adoption or anything on that front, Susan?
Susan Li:	Yes, I think echoing what Dave mentioned earlier, we had a range of expected
	impact from the ITT changes and ultimately what we've seen is in that range but it's
	really on the higher end of what we had expected, and I think the underreporting of
	web conversions has really been a bigger issue than we expected, but it's something
	that we're very focused on helping to mitigate through better modeling techniques.

Michael Nathanson:All right. Can I just ask -- thanks, Susan, on the idea that you guys improved underreporting into the end of the year is -- does the report maybe -- has Apple changed any of their own tools or privacy limits that may be helping on closing the reporting gap?

2

Susan Li:	Has Apple made any changes to their own tools? Sorry, was that the question?
David Wehner:	To help with the situation?

Michael Nathanson: Yes, have you seen any..

David Wehner:	I think it's more work we're doing on the modeling front to try and improve the
	attribution of conversion events.
Susan Li:	Yes.
David Wehner:	So I think it's more work that we're doing to try and bridge that gap rather than
	helpful work that Apple's doing. So...
Susan Li:	Yes, their tools are in the ecosystem, like SKAN, and we're seeing that there are
	measurement issues and so we're trying to mitigate that.
Michael Nathanson:Thank you, guys.
David Wehner:	Thanks, Michael.
Operator:	Our next question is from Alan Gould with Loop Capital. Please go ahead.
Alan Gould:	Yes, thank you. With respect to Facebook Reality Labs, can you give us some sense
	of how much that cost you last year? And are there any milestones that we should
	be looking for as you roll out your investment into this area?
David Wehner:	Yes, thanks, Alan. We'll be providing historic detail when we do the segment
	breakdown for Q4. Obviously that continues to be a growth area, so we've
	increased the investment over time and that will be reflected in the breakout when
	we provide it, but I'd say stay tuned for a more granular historical breakdown when
	we report in Q4. We're just giving a snapshot of the impact in 2021.
	In terms of milestones, there's obviously a wide range of products that we've been
	launching in this space. The Quest products, we launched the Ray-Ban Stories
	product, there'll be additional product rollouts, both in the sort of hardware arena
	and in the software and experience arena, so I think that's something to look at.
	I would also encourage you to join Facebook Connect on Thursday, Mark and the
	FRL team are going to be talking about a lot of the different initiatives we have in
	this area and be sharing some of the experiences that we expect to roll out, so that's
	probably going to be the best opportunity to see where that vision is heading.
Alan Gould:	OK, thank you, Dave.
David Wehner:	Yes, thanks, Alan.
Operator:	Our next question is from Brian Fitzgerald with Wells Fargo. Please go ahead.
	3

Brian Fitzgerald:	Thanks, guys. I wanted to follow up on the -- on the measurement undercounting
	conversions by 50 percent, I think the blog --- 15, 15 -- I think was the blog post.
David Wehner:	Yes.
Brian Fitzgerald:	So when you guys talk about, by year end we should be able to cut that in half. So
	just for clarification that means hey, we're going to be undercounting by 7 percent
	as we roll into '22. So just clarification on that point.
	And then when you talk about applying A.I. and machine learning to trust and
	safety, as you lean into Reels and video, is it -- is it harder to police that then syntax
	or images, is there more -- are you more prone to A.I. biases in video -- cleaning up
	video versus cleaning up some of the traditional media?
David Wehner:	Thanks, Brian. On the underreporting, and Susan correct me, but one of the things I
	would say is it tends to be really client sized does effect this. So when we have a
	larger advertiser it's easier to model the conversion. So what we'll do is we'll
	probably be more effective at correcting for the underreporting for larger
	advertisers and it will take -- it'll take more time to correct it.
	So I think it's the underreporting that we get to. It just sort of gets - maybe we get
	the 7 percent but that's not -- that's not consistent across client. So that's where
	we're saying this tends to be something that's more disruptive for smaller
	advertisers because it's harder to model when you've got smaller budgets and
	smaller cohorts because of the aggregated event approach as opposed to individual
	attribution.
Brian Fitzgerald:	Got it.
Susan Li:	Yes, that's right. The 15 percent's an aggregate number, we do expect that number
	gets reduced by half by end of year and we hope to make more progress in 2022
	plugging the other half.
	But the distribution is very variable by advertiser size. I think we'll get very close to
	correct for large advertisers but it will just take much longer for smaller advertisers.

David Wehner: Yes. And on number two, I mean video's already a big part of the experience so we -

• we obviously use A.I. and machine learning to look at content today but it's -- it is something that is a ongoing effort area of investment to be able to understand what content is out there and how it -- how it applies to our community standards. But we apply our community standards to video content as well as text based content.

Brian Fitzgerald:	Awesome. Thank you.
David Wehner:	And by the way, for stuff like -- certainly for like -- it gets less play but certainly for
	things like kind of more content that violates things like nudity and that front, we're
	able to use A.I. and machine learning to catch that whether it's video or not.
	4

	So that's a pretty clear example of where we're using A.I. to make sure that the
	content doesn't violate even if it's video.
Brian Fitzgerald:	Got it. Thanks.
David Wehner:	Yes, thanks, Brian.
Operator:	Our next -- our next question is from Stephen Ju with Credit Suisse. Please go ahead.
Stephen Ju:	OK. Thank you. So hi, Dave. So …
David Wehner:	Hey there.
Stephen Ju:	Presumably you're undercounting the events so the advertiser is still getting the
	benefits. So kind of similar to what SNAP said the other day. I think the value that
	you're providing to the advertisers is probably pretty much the same.
	So I was wondering if you can characterize -- I guess for -- I guess the very long tail
	advertisers are -- is there a certain percentage in the overall revenue in which
	Facebook is acting as principal as well as the agent in terms of spending on behalf of
	your clients because they have a certain fixed budget and there's a certain ROI goal
	that they're trying to achieve. And in your view you are still delivering on that.
Susan Li:	So …
Stephen Ju:	I guess in a follow -- yes, go ahead. Yes.
Susan Li:	Sorry. Go ahead.
Stephen Ju:	And yes, I was just wondering if you can talk about how many business profiles and
	how many advertisers you might have ended the third quarter with since it's been a
	little while since you guys have updated us with those metrics? Thanks.
Susan Li:	So on the first question we're seeing - we're definitely seeing that the data
	limitations from ATT are resulting in performance drops in addition to measurement
	drops, so I think those are two sort of separate things.
	Now the measurement drops are something that we can mitigate, and that's
	something that we're actively working on in the short-term, and then the
	performance drops are something that we are working on sort of longer-term
	mitigations either for in terms of figuring out how to do better modeling with more
	limited data or in the longer-term things like our own sort of first-party Shops and
	things like that.
	The other thing related to the measurement piece is because you're unable to
	measure as effectively you're also unable to target as effectively. The explicit
	targeting parameters are still available, but a lot of the targeting now is done
	5