Hello everyone. [music] Good morning,

Bill.

>> Good morning.

>> Great to have you.

You've had 10 straight quarters of

double-digit user growth. Most recently,

631

million global active monthly users.

It's pretty good.

How is the consumer seeming right now

between Iran, tariffs, gas prices? Like,

what are you seeing?

>> There's just a little bit going on in

the world, right?

>> You know, it it's uh you know,

interesting like obviously, you know, no

comments on current quarter, all those

sort of disclaimers as a public company.

>> Check.

>> Um

but, you know, we've talked about this

on our our last earnings call and pretty

consistently that, you know, there's

been real resilience amongst consumers.

Um

and I think not only has there been

resilience, I think it's also there's so

much changing

in terms of the landscape with AI and

all these kinds of things.

Consumers, I think,

interestingly, are craving authenticity

and meaning and purpose and all these

things that are not really answered by

AI. And and so, I think that's something

that we're really paying close attention

to is that, you know, I'm an

entrepreneur uh by background, an

engineer, a technologist. And there's so

many great uses of AI, but there's also

a bit of a sort of counter movement to

that with consumers of as so much is

changing,

what do they reach for to sort of have

ballast in the ship uh to to sort of

hold on to in terms of like where they

find trust or meaning or purpose in

these things.

Um

but, from a, you know, shopping

perspective, you know, of those 630

million users, more than half the

platform's Gen Z. So, Gen Z's our

largest, fastest-growing demographic. Um

and more than half of our users are

there to shop. And so, we've seen that

shopping activity uh particularly in a

visual-first, AI-driven world, uh

there's tremendous engagement.

>> So, let's come back to AI, of course. Uh

but you have this incredibly intentful,

purposeful, excited, passionate user

base, but you've said yourself, we have

more work to do to ensure monetization

more fully reflects the strength of that

user activity. So, what's holding the

company back on monetization?

>> Well, I wouldn't say there's anything

holding us back. I mean, we've made I

came in as CEO 4 years ago, and the

purpose was

twofold. One, we wanted to prove there

was a more positive business model for

social media, that you didn't have to

use engagement via enragement to sort of

keep people glued to a screen, we could

actually try to build a more positive

space, one that would leave people

feeling better, and one that would leave

people feeling better about their time

spent, that it wasn't about maximizing

time spent, but about maximizing time

well spent.

And then the other side of that was

building out as a shopping platform. And

over the last 4 years,

you know, we're now

80 billion plus searches per month.

Uh

you know, which is just a massive amount

of search volume. Those searches are

primarily visual, so proving out a new

way to search. More than half of those

are commercial. And as is true with any

platform, you want to get the users and

the engagement first, and then the

monetization follows behind that. But

we've also 5x the number of clicks to

advertisers over the last 3 years. So,

so the ad platform has been growing, you

know, very nicely. We quadrupled the

growth rate of the revenue. And so it is

following, but of course you want to get

the user engagement first, and the

revenue follows in behind that.

>> Is part of it just the way advertisers

measure success? I I recently read in

Media Stack one of your executives said,

"95% of Pinterest impact is invisible to

advertisers who look at last click

data." Is that a problem that

advertisers are looking at the wrong

thing?

>> Well, I think that's not just about us.

I think that's been a big breakthrough

in advertising generally.

Uh is

it's one of those things not all that

matters is measured and not all that

measure not all that is measured is is

what matters. And, you know, great CMOs

have always known this that even though

the last click was the thing that was

the most measurable, uh you know, any

great CMO would know that well, okay, do

you really believe the whole decision

was made in the last 30 seconds before

that last click? Or

were there things before that around

building your brand and your customer

connection and how customers should

think of you and how they were

researching your product? So much of

that happens before that. And just to be

very fact-based about it, for most

commercial journeys, they tend to be

multi-day, multi-week, multi-month. And

depending on how high consideration the

purchase is, the longer that time period

tends to be. But even something like

apparel tends to be a multi-day

consideration period. So, if you're only

measuring the last click, you've

probably just found the tollbooth, but

you didn't find sort of actually the

sort of map somebody used to navigate

the journey. And that's not just about

us. I think just generally that's been a

major issue. But the good news is with,

you know, things like clean rooms and

other privacy-preserving technology, the

ability to go understand the full

journey of the user is getting better

and better and better both with that and

with AI-driven systems that can help you

understand the multiple touchpoints that

were on that journey. And when viewed

through that lens, you know, yes, we 5x

the number of clicks to advertisers over

the last 3 years, so we're driving a

tremendous amount of last click. But of

those 80 billion searches a month that

are happening, more than half are

commercial. And you know, it's a great

moment to engage when you can meet the

user in a place where they have intent

but haven't yet decided what to buy. So,

95% plus of our queries are unbranded.

So, back to like are you a toll booth or

are you actually helping drive a

decision? That's a very fact-based way

to measure that of like how many of the

queries actually have the brand in it

already? If the brand's in it already,

then they sort of made up some part of

their mind. Well, 95% of our queries are

unbranded, which means the user has

clear commercial intent, but hasn't yet

decided what to buy. But, I think part

of this that you're touching on around

measurement, not just for us, but just

in the industry generally, I think it's

what the great CMOs have wanted for

years and years. They just couldn't

convince their CFO um if it wasn't

measurable. Well, now that more of that

journey is measurable, that's letting

great CMOs say, "Yes, brand building

matters. Customer connection matters.

How our product is understood by the

customer matters." And those things

matter way before those last 10 seconds

before that last click.

>> So, I know it we're getting in the

nitty-gritty here, but here we are at

Cannes. So, how are you helping

advertisers make that case? Tell us is

this Pinterest performance plus or like

what are the tools you're giving to help

make the case?

>> Yeah, so we've uh so we have our own AI

automation suite that makes it really

easy to create campaigns, connect to

that high commercial intent on our

platform.

Uh Pinterest performance plus, uh our AI

automation suite has cut campaign

creation time in half. We're doing

things like creative optimization where,

you know, um

you know, great marketers, great

creatives are always going to want to

know that they're showing up the way

they want in front of their customers.

But, the ability to personalize and say,

"Okay, well, you know, say you have like

a great, you know, kitchen appliance,

you know, that you want to show up."

Well, that person is if they're going to

go spend $5,000 on a coffee maker or

something,

they're probably thinking about how

that's going to look in their kitchen.

And so, when you can personalize down to

a level of like not just showing them a

great coffee maker, but showing them a

great coffee maker that is like sitting

in a kitchen

like their kitchen, you're going to have

better conversion rates on that and it's

going to be brand enhancing. These are

the kinds of things you can do with with

AI-driven creative optimization that

don't have to be about turning over

control to a black box, but can be about

how do you take, you know, your great

brand architecture and then marry that

with great personalization. So, there's

things like that that we're doing. And

then, we've integrated with 30 plus

uh measurement partners to sort of meet

advertisers where they are around these

issues of measuring the full funnel.

>> Is that the Is that the issue with the

resistance to some of these tools? It's

a black box and some

advertisers worry they're seating

control over their campaigns.

>> I I think um to some degree, yes. I I

think there's an interesting

um

uh tension to work through is that on

the one hand,

uh you know, these AI-driven systems can

do a really tremendous job of you know,

finding opportunities that you never

could have found on your own,

personalizing to to levels that you you

couldn't have done on your own. At the

same time, great marketers,

of course, want to know how their brand

shows up and are understandably going to

want to

be able to sort of track that through

and see that their brand shows up in a

high-quality way that they feel good

about. It shows up in a positive

environment. You know, we we see this

One of the reasons we've uh put so much

effort into building more positive

space. One, I just think it needs to

exist in the world um as a parent, um as

a citizen, I think we we need to demand

more from our tech platforms than what

we've been getting uh in terms of safety

and what it means for our our

our children. Um

but, it's also a good business decision

because study after study after study

shows that there's better engagement and

better conversion when users see

products in a positive context, right?

And somehow as an industry, we got

convinced over the last, you know, 20

years that you would measure the you'd

look at the thing that came right before

your ad and the thing that came right

after your ad, that the 10 seconds

before your ad, the 10 seconds after

your ad, and as long as those were safe,

we call it brand safe. If I had pitched

you on that in sort of the

TV world to say, "Oh, I've got this show

that's like really terrible, it shows

all kinds of horrible stuff,

but for 10 seconds before your ad and 10

seconds after your ad, that's going to

be safe stuff, so go ahead and run your

ad on my TV show."

You'd have laughed me out of the room,

but we've started to accept that, not

started to, we you know, the industry

accepted that for the last 20-plus

years.

We want to prove and have been proving

that actually the broader environment

around that matters a lot more.

And in the studies are very consistent

around that.

And and I I think again, I think great

marketers have always known that, so

we're trying to give them tools to to

engage in a more positive space.

>> Positive the whole way, AI for

positivity. Let's let's get into it on

AI. So,

how much content on Pinterest is

AI-generated and do people care?

>> So, it's interesting, you know, we've

seen this multiple times, you know, the

ability to create content has been

democratized

tremendously, and it's not just with AI,

you know, when we all got a computer in

our pocket and then there was a phone,

you know, that on that phone you had a a

camera,

anybody can go create content. And

consistently you've seen that as content

creation has become easier, the volume

of content goes up tremendously. The

average quality goes down because

anybody can go shoot a photo. Um you

don't have to be a professional

photographer to take a photo on your

phone, uh including me, you know,

that'll like get half of, you know,

someone's head in the shot or whatever.

Anybody can take a photo. You get way

more content, but while the average

quality goes down, the total volume goes

up, and in the fact that

everybody can create and the distance

between someone seeing something in

their mind's eye and being able to go

create that piece of content. The fact

that that is collapsing, I think is

actually bringing a lot more human

creativity in the world. So, our job as

a platform is to not judge whether it's,

you know, digitally created, you know,

or AI, you know,

modified, but to say, "Is this a

high-quality piece of content that's

really relevant to this user?" And so,

as a platform, we're doing a lot to make

sure the quality of the content is is is

high, that the relevancy for the user is

good, and then also we're giving users

agency, which I think we've led the way

on this relative to other platforms in

we're labeling AI-generated content and

giving users tools to say, "Show me less

AI." If they don't want to see AI cuz

there is personal taste in this.

Um in not only is there personal taste,

it varies from moment to moment. There

are some moments where the AI-generated

content, you know,

uh

may appear, you know, as something that

you couldn't really do in real life.

Well, sometimes that's okay, right? Um

you know, how many, you know, if you my

my daughter's 12. Anybody that has young

kids, like this past Halloween, you

know, how many kids did you see with

like, you know, purple hair because of

like K-pop demon hunters? Well, like

that's digitally created content that

inspired something in the real world.

So, there's certain certain times when

that AI content is helping to inspire.

There's other times when somebody say

like, "No, I'm like actually trying to

buy I'm, you know, trying to redecorate

a room, and I want to see a sofa I can

actually buy in that room." This is one

of the things that we do is is help

translate between here's a piece of

content that may have inspired, that

maybe was created with AI, but then with

our visual search technology we make it

so that you can see the the buyable

thing in the real world. And we're

seeing that really resonating with

users, but we're in this transition

period where users are still trying to

figure out their own tastes and

preferences and our response to this is

do as much as we can to to filter and

sort high-quality content and give the

user agency so that they know when it's

AI generated and they have their own

controls to say show me less if they

don't want to see it.

>> Okay, working your way through. And then

just on the back end

uh you've been so all in on AI. How

about those bills? You got some token

maxing going on or how are you

controlling your costs there?

>> Well, you know, it's um

fortunately because we're a very high

commercial intent platform, we've

actually expanded profitability over the

last 3 years even as we have

dramatically increased the use of AI on

our platform because we're aligning use

of AI with high commercial intent

audience and high commercial intent

experiences that monetize really well.

Um

but I do think this is something that

people need to pay very close attention

to because I think we've been in this

mode as an industry of go go go, don't

be left out and then now, you know,

especially you start to move to

uses-based pricing,

the onus is much higher on platforms to

prove that there was a good ROI on that.

Um

and we've been very uh as a smaller

platform,

uh

you know, relative to some of the

giants,

we've had to take that approach from the

start. Uh you know, I talked about how

we are able to use open-source models

and post-train them on our high

commercial intent data and see, you

know, not only 90% lower cost than the

closed proprietary models, but also

better performance because we can train

it on our own data. And so I think some

of these issues of how do you get the

best performance,

open source can do a lot there because

you can post-train on your own data in

ways that you can't with proprietary

models, and you can get much much lower

costs on these things. But, I do think

we're just at the start of sort of a

reckoning of what are the real costs of

of these AI platforms where in some use

cases it totally makes sense, and other

use cases, once you start to pay the

real price of those tokens, it it may

not.

>> So, that's a conversation internally and

what keeping an eye on this kind of

thing?

>> Yeah, absolutely. And and one that we're

trying to help our our clients and

partners with as well.

>> You've talked in management about the

value of fast feedback loops, products,

and management. Could you just say what

what you mean by that and give an

example?

>> So, it's interesting like this um

I gave a talk on this in Davos in 2018

um

that was turned into a bit of a white

paper that uh back then I was saying

like, "Okay, AI is coming.

Like cloud compute, AI will be the

domain in terms like the sort of core

foundational models that AI will be

created by a small number of players

because it's so

uh you know, so CAPEX intensive. But,

in the same way that cloud compute, you

know, created all kinds of new

businesses that couldn't have existed

otherwise, and all kinds of

opportunities that wouldn't have existed

otherwise, that AI would do the same

thing. That AI would be put in the hands

of the many. And the thing that people

should think about is what are the

unique feedback loops in your business.

AI thrives on feedback loops. And so,

while most companies are not going to go

create their own foundational models,

what every company can do is think about

how do I serve my customer? What do I

what problem am I solving for them? What

makes me uniquely equipped to solve it?

And then, what are the feedback loops in

my business

that are unique to me that I can train

my that that that I can train AI on to

make my business better. I think that's

something that every single business can

think about. You don't need computer

science PhDs to do that. Um but

everybody should be thinking about what

are the feedback loops in their

business.

How do you tap into those and then how

do you how do you find you systems and

platforms are going to help you do that?

And again, part of what we're you know

what we're trying to do with our

platform is not only are we making that

commercial intent available, we're

trying to make it available in ways that

help our advertisers gain a customer,

not just a transaction. Part of why I

was so interested in doing visual search

and discovery is that you know more than

half the human brain is wired for visual

processing.

You know, you still have a lot of these

things that are like you know very text

based, but most commercial journeys are

tremendously visual. And how do people

tell their brand story? How do they

differentiate? How do they create

connection with customers?

Visuals are a huge part of that and we

want to make that accessible to all the

advertisers on our platform and do it in

a way where not only do they gain a

transaction, but they actually we take

the customers to our advertisers so that

our advertisers actually gain a deeper

relationship so they can start to have

their own feedback loops with those

customers to deepen the relationship.

>> Thank you, Bill. Great to have you.

Thank you, everyone.