Activate Trusted Data Everywhere | Data Cloud Keynote, Dreamforce 2025

We had a lot of fragmentation.

Puddles of customer information everywhere.

16 different systems. None of it

connected. My problem is only getting worse.

What's at risk is that we focus on the

wrong things. Customer focus comes in

jeopardy. We have to unlock that somehow.

Using data cloud.

The data is available. The data

is clean. The data can be

trusted. Bring it into one place.

Delivering much richer personalized

content to delight you in the moment. What

used to be a series of spaghetti

integrations transformed into a single source

of truth for customers. And that in

turn is the data that's activated into

marketing cloud. All together unified.

That's where we see the biggest

value. That's exactly what we need

to make generative AI successful.

With Salesforce, we can move at a speed

we've never been able to move before.

You simply activate your data cloud to

give that agent access to real-time data.

And serving it up to Salesforce so

that we can take action on it. data cloud

delivers then that's critical to give

them that personalized experience can't

think of a technology before that that

would really accomplish that to that level

that for me is really

really powerful please

welcome executive vice president and general

manager salesforce rahul aradkar okay thank

you thank you what an inspiring video that

was Every time I see it, it inspires me.

I was here last year, about a year back,

delivering the first data cloud keynote.

Thanks to our customers, we

have had an incredible year.

And thanks to our partners, who have been

helping our customers deliver the outcomes.

Now, how about the data blazers? We

started that community a year back. Now

we have 70,000 of them. We have a few

of them right here sitting amongst us.

Yes, thank you.

And the employees who are watching

this, who are sitting here,

your creativity, your dedication,

and your perseverance is the

reason why we are here today.

So we are seeing success across

every geo, every industry,

and every segment.

And frankly, it's blown us away. it's very

humbling, and your outcomes have been

phenomenal. There are a lot of numbers here. We

share a lot of numbers in terms of usage,

numbers in trillions, numbers in

quadrillions, but those really, what's behind

those numbers are real customer stories.

I was at the customer advisory

board yesterday, and a small business

customer who is a physician,

Romi Chopra, he's here with us, or I

think he's here with us here today.

Romy Chopra is a physician, small business

customer, using data cloud and

delivering value within weeks for his patients

and saving over one and a half million

with a small footprint of data cloud.

Or take the other end of the spectrum,

somebody like Indeed,

completely transforming their business

to being an agentic enterprise.

So it's the entire range that we have seen

from a success standpoint. point. Now, the

creativity and the experimentation and the

outcomes that our customers have provided has

been really behind what we do with innovation.

It's been our fuel that drives innovation.

Let me take an example.

SiriusXM helped us learn about streaming

data and what we need to do with

streaming data. The customers in

Japan told us about IBM mainframe data

being zero copied into data cloud.

Now, IBM mainframe data with iceberg

standards is zero copy in data cloud.

That's correct. You heard it right.

Even their DB2 data.

Then there are data blazers telling us about

unstructured data. You're going to hear a

lot of that in the next few minutes about

what what we're doing with unstructured data.

Now, for those of you

who are using data cloud

and who already are driving outcomes,

there's a lot of innovation coming.

And for those of you who are just getting

started or want to get started and

want to learn, what's this all about?

How does it work?

What difference does it make?

Let's unpack that together.

You see, data has always been the foundation

for everything customer, knowing your

customers, engaging with your customers.

We have built incredible systems over

the years to store data, to ingest

data, to analyze it, to drive

insights, to know more about your customers,

to engage with your customers.

But here's the problem. All that knowledge

is trapped, is siloed, is disconnected.

And the paradox is that this very

disconnected data is actually interconnected.

Let me take an example.

real story my daughter and i

were buying a car recently and

we were researching the vehicle

and we bought the car and after that she

gets incessant reach outs for the same car

what's the what's the what's the point

in reaching out to her with the same car

it's unlikely she's going to buy the car

again within a one within a month and i

get very expensive ads about the same car

I'm not buying another car with her within

a week. So that's wasted media spend.

That's wasted data that is sitting around.

For that brand, it's wasted

data and wasted media spend.

So between the time that you engage with

customers and what data you need, you've

built a lot of pipelines. Those pipelines are

brittle, and they're not available in the

moment that is needed for your customers.

So, what if the gap between knowing your

customers and actioning on it disappeared?

As my friend Rafa from Brodesco said

yesterday at the cab, it is effectively

connecting data to the place where it's needed

at the time at which it's needed to deliver

value across all the fragmented data.

That is the vision that we have been

living at Salesforce. That's the vision

we have been partnering with you.

That's the future we believe in. and

that's the future we are going towards.

And given that, we are introducing Data360.

Data360 is the new data cloud. It's the

same product you have come to love and use,

has a new name, with a ton

more innovation that is coming.

So data cloud, simply put, is an evolution.

Data360 is an evolution of data cloud.

It is activating all your trusted data

everywhere. We have talked about activation

repeatedly. Pretty much everybody in the cab

yesterday told us it's all about activation.

Now, what does that mean?

It's simply two things.

It's personalized experiences

across all touch points, all

modalities, and all channels.

And perhaps the most interesting thing

you're going to hear for the rest of the

conference is this word known as context.

providing context across all your fragmented

data in a meaningful manner in the moment

that AI agents need. That agent that was

trying to reach out to my daughter and me

had no context.

Now, going forward with the innovations

that we're going to talk to you about

today, that interaction will have context.

You came to this conference, you're hearing

a lot about what could be, what should

be with the data and AI transformation.

What we have focused on for the

next 40 minutes is what already is.

There are three broadly

used use cases here.

Number one is we've seen significant focus

on and our customers delivering value

with digital engagement and delivering

personalized real-time experiences with it.

The second one is what I refer to using

context to deliver outcomes that AI agents

can deliver on your behalf with humans and

AI working together. And the third one is

binding the customer 360 so you have

integrated experiences across the entire 360.

So for the next 40 minutes, we're

going to unpack all of these three.

And to do that, I want to get started

with the first one, delivering

personalized real -time experiences.

And I want to invite my colleague,

a leader in data cloud,

Chandrika, to the stage.

Thank you, Rahul.

I'm super excited to be here to talk about

how Data360 is enabling our customers to

deliver personalized, real-time experiences.

Now, across the industry, marketers

are using Data360 to connect

to data across their ecosystem.

to unify and harmonize all that data,

know their customers,

and activate that data to reach

their customers at the right

moment with the right message.

Who better to showcase

this use case than Meta?

Meta is actually using Data360 as

their data platform. They're using

that to power their B2B scenarios.

They're connecting to all of their

fragmented data across their ecosystem,

unifying all of that to deliver personalized

experiences to their customers' advertisers

in a consistent and connected fashion.

Now, let's hear how they're doing it and

share more. In order to do that, I'd like to

invite Nick Harris and Corbin Tho to stage.

Welcome Nick.

Welcome Corbin.

Thanks for being here. Now, before

we begin, could you please

introduce yourselves to the audience?

Thanks Shandrika.

Nick Harris, glad to be here, lead

a data analytics team at Meta.

Hey, everyone. Excited to be here. Corbin

Tho. I am the head of MarTech Platforms

for our Scaled Business Marketing Group.

Now, before we begin, maybe you can

start by talking a little bit

about Meta's vision in the B2B space

and how does that relate to data?

Sure. So one of the main things that we're

trying to do in the B2B marketing space

is ensure that advertisers or clients

like you all are able to reach your customers

on our platforms at the right time.

That starts with educating you about how to

use our products, giving you the best

best practices on that, as well as ensuring

that you know what to do so that you're

reaching your customers to get the return

that you want from advertising with us.

You know, a company your size has a vast

swath of data, planet -scale data, right?

And a lot of complexity that comes with

handling that kind of data. Can you share

a little bit more about how you were

dealing with that complexity before you

took on the journey to go on with Data360?

Yeah, certainly. So we were effectively

operating on two different Martech stacks. Our

small business team was operating on one set

of tools. Our global business team was

operating on a completely separate set of tools.

And the result of that was fragmentation,

not just within our marketing org, but also

with product and sales as well. So building

those integrations was very challenging. Passing

that data back and forth was very messy.

And this just resulted in a very

inconsistent experience for our advertisers.

And that was the key right there. As Rahul

mentioned, as customers are migrating from

one part of our marketing funnel to another or

one part of our segment to another, we were

losing the context because that lived in a

different data ecosystem, and we couldn't

serve that customer with intelligence right

when they needed it the most. You know, this

is so similar to everything that we heard at

the CAB yesterday, Customer Advisory Board,

from many other enterprise customers as

well. Tell us how Data360 helped solve this

problem for you. Yeah, certainly. So I alluded

to already, you know, two different MarTech

stacks, but now we have a single platform

where we can build a consistent connected

experience for our advertisers. Not just that,

we can actually see our advertisers as they

start in the acquisition tier and as they

graduate all the way up into our growth tier.

And as we are bringing in signals and

context into the platform, were able to act

on that data in a much more rapid fashion.

And as Corbyn mentioned, we're trying

to make the experience between marketing

and sales a lot closer. And being

able to provide those sales agents

with qualified leads has been key to

making this product successful for us.

That's impressive. You know, I'm

sure folks in the audience are wondering

about the same question I am.

What were the outcomes from using Data360?

What were some business metrics

you were able to achieve?

Yeah, the results, absolutely.

So, you know, we've moved off six

MarTech platforms down into three.

And so we were able to save on

operating and licensing expenses.

And with that added efficiency, you know,

we were able to deliver, you know, 20 to

30 plus campaigns quarter over quarter.

And then on the growth side, we're able to

target more specific audiences to drive

some of that revenue. And then for the first

time, we were actually able to attribute

revenue to our mid-market audience, which was

a major accomplishment for us this year.

And it's not just about the efficiencies

or opportunities that we had as a company.

It's also about our advertisers ensuring

that they're able to reach their customers

better as well. And so we're seeing

that as a result of these actions as well.

Those are impressive numbers. It's been an

amazing journey partnering with you. You

know, we've learned a lot along the way.

I know you guys are thinking about the next

set of use cases and how you're going to

take B2B forward with AI. And we're looking

forward to continuing that journey with

you guys. And thank you so much for being on

the stage here and sharing your insights.

Thank you, Nick. Thank you, Corbin.

Give it up for Nick and Corbin.

Now, folks, what you just heard is a powerful

story of how Data360 can be used to

activate your data. And that's what Meta did.

You know, that's not unlike how many of our

customers are doing using Data360. FedEx,

Air India, Heathrow, Elevance. Wyndham, all

of them. You know, that's the reason why

Data360 is the number one CDP for marketers.

It's also recognized by our

industry analyst as the leading

vendor in the B2B and B2C space.

At the heart of it all is connectivity.

You can connect to any data anywhere,

unlock that data, whether you're ingesting

that data or whether you're federating with

that data through zero copy and make that

available for your marketing scenarios. You

can unify and harmonize all that data in

Data360 and personalize and create real-time

experiences for your customers anywhere,

whether you're using Customer360, sales,

service, marketing, commerce, if you're

driving AI and agentic scenarios, or if

you're doing BI and analytics as well.

And at the heart of it all

is governance and security.

You can tag your metadata, you can classify

your metadata, and you can build policies

using our unified policy framework, whether

it's access policies, security policies,

masking policies, or you want to just do

retention policies across the board,

consistently across your metadata. So you can

have confidence in targeting customers with

precision and comprehension with your

unified data. And of course, with our

activation stack, you can reach your customers'

touch points anywhere they are with

confidence across both paid and owned media.

Now, folks, that's not all.

Now, as we talk to Meta and some of our

other big customers, we realize that

your customers don't just live just within

the walls of your business. In fact,

they live across your entire partner

network, whether it's retailers, healthcare

services financial institutions

across all sorts of data systems, right?

Now, that's a pain point.

Wouldn't it be nice if you could collaborate

across your entire partner network, across

the vast swath of data that lives in

their ecosystems, such that you can safeguard

the sensitive data in a privacy-safe way

and discover new hidden audiences and target

those audiences for additional campaigns,

grow your business, have happy customers.

We think that's the next innovation.

So today, we're announcing

Data360 Clean Rooms.

Data360 Clean Rooms is designed to give you

a way to collaborate safely with your

partners using that shared data. Now, it's

all founded on the zero copy foundation, which

means you can untrap, unlock all that

trap data that's located everywhere in your

network and your partner network and make

that available for collaboration and ensure

that sensitive data is not shared. You

can discover hidden audiences. This is all

interoperable across many clouds, so you can

access data, discover hidden audiences

everywhere and anywhere. And of course, the

insight that you gain from unlocking all of

that, the aggregated anonymized insights

can be turned into instant action across

your entire ecosystem, across the Salesforce's

entire ecosystem, across customer 360,

across AI and agent force,

as well as BI and analytics.

Now, let's actually take an instance,

a use case, to talk through this.

For all of you to be in this room here,

you all traveled via different airlines.

You're staying all at different hotels,

right? And if you're like me, you have

preferred airlines, preferred hotels,

and you're booking them separately. And

of course, you're also tracking loyalty

points across all of these separately.

That's a big pain point, certainly for

me. I'm sure it is all for you as well.

Now, wouldn't it be nice if airlines and

hotels could collaborate across all of

the vast swath of data they have spread

across their entire ecosystem and be able

to find these hidden audiences across

both of their systems, across both of

their organizations, in a secure fashion.

Let's actually get into a demo. Now I'm

going to invite my demo partners here,

April and David, to help me with the demo.

April is a marketer at a fictitious hotel,

Polonia Hotels, and she wants to collaborate

with me, a media representative at Astro

Airlines. Now she wants to be able to

leverage my data such that she can pinpoint

audiences, an overlap segment, if you will,

in a clean room environment so she can

target that newfound segment, if you will,

using my media network,

you know, feedback screens,

mobile phones, emails, potentially, and

even signage at the gates as well. Now let's

see how we do this. We're going to go into

the clean room environment, April and I,

and she's going to bring her high value

customer. She cares about this segment

deeply, and she wants to map that with all

the segments I, as Astro Airlines, offers

in this secure, clean room environment. It's

a neutral environment, right? And so

she queries high-value customers against all

of my segments to create this overlap

segment. Let's take a look at the results.

April,

can you share with us and describe to us

what this looks like, what this means? Yeah,

so we actually collaborate with a lot of

different partners, and this is our collaboration

with you. I have all of my customer

segments over here on the left, and I have

all of your customer segments that match on

the right. And I can just query those high

-value guests and see that international

travelers and business travelers, that is where

we're going to get the key perspective

reach and get all of that information to you

guys. Not just that, though, is we are all

integrated here, and I can share this to

Slack so that we can continue to collaborate

to make the best campaign moving forward.

That's amazing. So to bring it all together

to the pain point we just talked about

earlier, first and foremost, this is a

secure, neutral, clean room environment,

right? I, as Astro Airlines, is a provider

of data. And I can securely, safely bring the

data that I choose to and pick the partner

that I want to share this with. So these are

anonymized, aggregated segments I'm bringing

to the table, and April is interested in

specifically finding pinpointed audiences,

prospects, and combine that with my shared

audiences to create this overlapping segment.

Now, it's a win-win, right? April, as a

marketer, can target the specific newly found

audiences. I, as Astro Airlines, get ad spend

from Polonia, and you all, or we all, as

customers, are able to leverage the two businesses

collaborating together. Now, we don't

have to separately track them and map our

identities to them, right? So that, folks, is

the power of Data360 Clean Rooms. Now, you

might all be wondering, how is our Data360 Clean

Room any different than any of the solutions

that are available in the market today?

Database vendors,

DataLakehouse vendors offer clean room

solutions, right? Let me tell you, first

and foremost, with data 360 clean rooms,

you are not moving raw tables. You're

not having to deal with raw tables and

have to then figure out what that means.

Raw tables are not here. We're actually dealing

with harmonized, unified context, rich

context, aggregated, anonymized, chosen by the

provider to share in the secure clean room

environment. Once you find that aggregated

insight, you can have actionability right

here in the flow of work within Salesforce. You

can use your customer 360, sales, service,

commerce, marketing. You can use AI and agent

force, or you can use BI and analytics to

immediately action on that newfound insight

in the flow of your work as a business user.

And of course, governance and security is

fully baked in. You can have policy controls

over that data. You can choose who you want

to share your data with and what query that

organization, that partner can run, and on

what columns. You can set security policies

over all of that and compliance is fully met

in the secure environment and all of this

is built on the zero copy foundation which

means you can get at that trap data and unlock

that trap data and make that available

in the secure clean room environment now we're

going to move on to the next innovation

to talk about how data 360 is powering smarter

more trusted ai agents with context and

to do that I'd like to invite my partner and

friend and an amazing leader in data 360

marketing, Erika Early. Erika, take it away.

Thank you, Chandrika.

Hello, Data Blazers. I'm so happy to

be here. I'm very grateful to learn about

all the phenomenal innovation from

many of our customers, just like Meta.

And I've also been learning a lot

from many of our customers who

are doing agent -forced solutions.

And as someone that deeply loves

and understands data and AI,

lot of the innovation that you're going

to be seeing here today and throughout

this event is going to completely

transform the way that you drive trust

and accuracy with your AI and agents.

Now, we are in a great time,

and this is a time of AI.

This is the era of innovation, and

this is the era where data is definitely

a place where you can drive a lot.

And one of the things that I wanted to

share today is no matter what kind of

data you have, whether it's structured

or unstructured data, you can process

all that. And what really matters is a

semantic understanding of all your data.

But something that is interesting

that's happening in the industry is a

lot of our customers are telling us

that it's hard to process data in AI.

And MIT is saying that about 95

percent of Gen AI pilots are failing.

Why do you think it's that?

This is happening because agents need

context to take action. And we've done a

lot of advancements with unstructured

data. You can process files like PDFs,

images, and videos. But there is a very

interesting type of data, of unstructured

complex data, that it's hard to process.

And the reason why this is interesting is

because there's a lot of deep domain

knowledge. And to truly drive understanding

and be able to unify all your data, you

need to understand domain knowledge so that

you can drive trusted AI and insights.

I'll give you a few examples. Imagine

you have documents that have flowcharts or

road tables or graphs. It's really hard

to parse. For AI, it's a hard problem.

But there's a lot of knowledge in here,

and you need it to drive insights.

And common AI models, if you use a common

retrieval augmented generation model or

a large language model, it's going to be

hard for them to process that. In fact,

you might get about 30% to 40% accuracy.

And the reason why is they cannot really

parse and understand this information.

It's really hard for agents to see that.

And this is why, today, Today, we're very

excited to announce that we're advancing our

unstructured data capabilities in Data360.

With intelligent context, you have

now the easiest and fastest way to drive

AI insights and trust and accuracy.

You can process complex documents.

And the best part, you can do that

with governance and with security.

Think about this. You would not use

your important information in a public

AI tool. And this is why it matters.

But what better way to talk

about this than with a live demo?

So let's talk about Cumulus Health.

They are a leader in medicine and

healthcare and technology. And

they have many, many patients. And

one of those patients is Harper.

And Harper has diabetes, and

she's been wondering, is she's

eligible for a clinical trial.

So for that, let's go see how an employee

of this clinic can work with agents and try

to help this patient. So I'm going to bring

back April and David to do an amazing demo.

Awesome. So I am going to play the part of

the employee that's getting Harper's question.

And I'm right here in my home health, and

I can just use AgentForce to query and see

if Harper-Clark is eligible for the trial.

Um, Erika, something seems

to be a little off here.

Yeah, I mean, this information that

you're seeing on the screen is not

really going to help this patient.

It's not understanding the context and the

information from her or from the decision

-making process to tell her she's eligible.

How do you think we can fix this?

Well, we could spend all day parsing

through this lovely document that's

up here and Harper's medical history,

but I'm not sure the whole audience

here wants to watch us do that.

Yes, but I want to pause here to say,

look, in this document, there's a lot of

deep domain knowledge, and it's very

important so that we can drive accuracy.

You need to know the age of the patient,

you need to know their lab results, all the

medical notes, all that information is

stored somewhere. Now all the logic is here.

What happens if we can use this

document and use intelligent context to

understand if we can solve this

problem and help the patient, and then

the employee can use AI and agents.

One step ahead of you. I already have

my agent over here. Let me just go

ahead and quickly kick that off. To remind

everyone, this over here, that's the

statement that we don't want, right?

And while the agent is over here thinking

really hard and compiling the

information using our structured data, our

unstructured data built in with governance.

Let's take a peek under the covers,

okay? All I did was upload a PDF and then

let the agent take all the default

configuration. And what it has done is taken

this entire document, it has generated

the chunks for me, and I can even ask

AgentForce, how do you qualify for tier one?

And it's going to parse all that information

for me so that I can check for accuracy

and even reference back each of those

individual chunks to make sure that the

agent is set up the way that I want to.

Let's go back over here, and we can see

the agent has already answered, and Harper

is conditionally eligible for Tier 2,

so my job is done in seconds, and I can

just get that information to her so that

she can get started with the paperwork.

Okay, that's awesome.

So what we just saw is, look, we have all

this customer information. We have the patient

information stored. We also have lots of

unstructured notes from the doctor. we also

have now this document that we're able to

process. And it's working. Intelligent context

is amazing. And it's going to be a lot of

difference in how we drive trust and accuracy.

And this is why Data360 is really

what's driving everything that

you need for trust and accuracy.

You can unify and harmonize and govern

your data. Something important on the demo

is this data is very secure. It's secure

for our patients. So we can do this

on a trusted foundation of governance and

trust that is built into our platform.

Now, I have another amazing

customer Salesforce.

Salesforce runs on Data360.

And we have many different use

cases, from marketing to agent force

and even more customer 360 apps.

And to talk about all the amazing

innovation and how we build this, it's

a true honor to welcome on stage

Michael Andrew, our chief data officer.

Take it away, Michael. All right.

Thanks so much.

All right.

Hi everybody.

So I'm going to talk to you a little

bit about Salesforce and how we use

Data360 to solve our own problems

across the whole experience, from

marketing to agents to the customer 360.

And I'm our chief data officer. And what

does that mean? Well, first of all,

I love data. I'm obsessed with data. If

you know me, you'll know this is true.

And I wake up every day saying,

how can I help Salesforce get

a little bit better at data?

And I think so many of

you have that same dream.

You're trying to help your companies get

better with data. You're trying to help

your companies get smarter and take

better action, whether it's turning it into

insights and analytics, forecasting

to make better decisions, or ultimately

into action, where you can turn that data

into agents, turn that data into better

experiences across your Customer 360,

activate that data for audiences to grow

your business, or automate your business

to be more efficient. And you want

all of that to be trusted and governed.

Now, at Salesforce, we haven't always

had our act all the way together.

You know, our own data was fragmented. We

had hundreds of millions of profiles and

systems. We have hundreds of petabytes of

data. We have a massive data lake on

Amazon. We have a massive, amazing enterprise

data warehouse on Snowflake, hundreds of

thousands of applications. And that meant

we couldn't always simple route leads to

the right sales team. When you filed a

case, we didn't always have all the information

we needed to solve your problem. And

as we started to think about agents, we

realized that our context and knowledge

was all over the place. That was not going

to help us build the agentic enterprise.

What I'm so excited to share with you

is we believe we're making significant

strides solving this with data 360.

The Salesforce now runs on data 360.

We've taken those fragmented profiles.

We now have 153 million unified profiles

with 68 billion signals attached to them.

So every time you're opening an email,

visiting the website,

talking to our sales team, filing a

ticket, coming to an event like this,

or acting in Trailhead, we

can now unify that to you.

And then our knowledge, already more than

700,000 documents are in vector indexes that

power the agents. And what does this mean?

Now when you submit a lead, we can see

who you are, score it, and route it

to our 25,000 sellers around the world.

Now when you file a ticket, we might

know more things like, what's your

trailhead badge, what certifications you've

done, have greater context on you.

And this means we can now

power agent experiences.

On our website, we've now served

more than 1.8 million of you,

helping to solve your problems. We now

have almost 23,000 salespeople every day in

Slack talking to agents to help them prepare

for deals. And now our SCR agent is

helping to respond to leads that we can never

get to. And this is running 24-7, 365.

When this sneezes, I'm getting alert. I'm

getting pager duty. We're trying to solve

it. If it doesn't solve it, I'm calling my

friend Rahul and saying, hey, we've got to

work on this. And we do. We figure it out.

And this is what customer zero means to us.

Our pain and where we solve our own pain

turns into better software for all of you and

solutions to help you and your business.

But it's not just about

running and scaling. By the

build this, I want to have a shout out to

Janani and Abhishek. I don't know if

they're in the room here today, but they

were with us when we started with one Scrum

team, so let's give a big hand to them.

We've taken it from one Scrum team

to now this, and of course, so many

of the teams made this possible.

But it's not just about scale, it's about

impact. And we're now seeing incredible

results. For our marketing teams,

we've been able to take data that was trapped

and automate revenue in e-commerce.

As we start to take smarter data and

target our ads, target our marketing,

we're seeing huge lifts with more

personate, relevant advertising paying off.

For our sales team now, year over year,

every lead we send to them, we're

way more likely to convert into a deal.

Every time we have an alert, it helps

to prioritize high -interest prospects.

And for our service with the help agent,

we have now saved almost $100 million

in costs. That's right, $100 million.

How?

Because if the agent can solve your problem

right away, that means you never have to

file a case. It never has to go to a

support engineer. And that means our support

engineers can spend more time helping all

of you create success with your customers.

But we would not have been able

to do this without the power

of the platform and zero copy.

And we have a lot of data. We have more than

90, I think, 95 trillion records on Snowflake,

hundreds of petabytes on Amazon. I'm not

going to get rid of that. We spent a lot of

money and time building that infrastructure,

but we need to tap into it. By our estimates,

we've saved more than 10 million on

engineering costs that we would have had to pay

if we had to shuffle that data around. And

that means that we could take our data and

turn it into action to drive our own results.

But, you know, I want to actually

show you this in action. So

we're going to welcome back April.

Here we go. Let's do a demo.

Let's actually show it.

So this is our real production instance.

And by the way, we're in production

freeze during Dreamforce. We don't want

to mess up anything for you. So here's

our little friendly reminder, everybody.

You can see we have about 1,000 data

streams coming into our instance.

Those data streams then get

harmonized into the unified profile.

So if you look kind of on the right,

you can see all those profiles, now 46

% consolidation, they were fragmented,

now they're unified down

into single profiles.

This means we can then take

data. So let's take an example.

Here's a table we have on Snowflake that

our marketing data science team produced.

It gives an estimate of the

value of the touch points,

almost 23 million records.

Now we want to use this, so look at

this. We can now build a segment. taking

that same data, zero copy, no movement,

and we're able to build it,

we can, you know,

go through this, and, wait, wait,

wait, it's production, please don't do

that, don't do that, it's production.

All right,

so, this is a real five

million segment population,

it's real, it's real, we gave her

access, we'll take it back later, anyway.

So,

let's now show you this, every day

this is running, so you can see

this segment publishes to our marketing

cloud for European marketers,

and I'm glad it actually

was successful yesterday.

It's a good history, everyday publishing

out. And this is showing you kind of what

Meta showed you, how we activate data for

better marketing. But it's not just

activation, it's also about agents. So let's

look at that help agent. So when you go to

ask for help, and you start to say, hey,

tell me more about data cloud connectors.

The answer is, of course, we're going to

have it, right? We have all the answers.

But how does the agent know that? Agent

Force, think. You got it? Yes. Here we go.

It's telling you all about it. Well, behind

the scene, we have built those search

indexes. So all those documents now come into

knowledge indexes that every agent can take

advantage of. And now every single agent you

work with is coming from this data. The

structured data about you as a customer and

the indexes of knowledge that allow AgentForce

to give the right answers and serve you.

But I want to talk a little

more about zero copy.

And who better to talk about zero

copy than Christian Kleinerman,

the EVP of product at Snowflake.

Come on down, Christian. Let's

talk about some data stuff.

All right.

Hello, Dreamforce. How's it going?

Amazing.

Well, Christian, we've done a lot of

work together over the years. I'd

love, you know, maybe you can talk a

little bit about the history of our

partnership and what we built together.

Yeah, so at the heart of what Snowflake

does, it's all about helping organizations

get more value out of their data. And of

course, that includes Salesforce data. And

I think we saw many customers trying to

integrate our technologies and leverage data.

We built a connector, then we looked at

where there was additional friction, and

we've systematically been making it better

until we arrive at zero copy sharing, which

is the marquee integration between us.

Amazing.

Now, what are some of the common pain

points that our partnership is

helping our customers solve together?

I think at the very top of

the list is data copies.

You said you wake up and you think

about data. I don't think you think about,

I'm going to make a new copy today,

because copies are a pain.

A, because maintaining a new pipeline and

how do you keep it running, all of that is

painful. But the other one is security and

governance. Now you have an extra copy to

govern. And that is why we're so excited about

the ability to integrate and collaborate

without having to make an extra copy.

Yeah, you know, we showed the demo course,

taking our Snowflake data, activating it up

to data cloud. Our customers can go the

other way, right? Can take from data cloud into

Snowflake. I'm curious, what are you seeing

as use cases, is that bidirectional

fluidity of data. Yeah, we see a lot of value on

going, having data going both directions.

Can I take an example, Salesforce

has great customer data.

Someone may have product data in Snowflake.

So you can say zero copy from Salesforce to

Snowflake. I combine it with that product

data. I can use old school machine learning

that still exists and is still a thing and

be able to do, for example, prediction

creation or leads generation, and once I

have those insights, I can use zero copy

sharing back onto Salesforce and put it

in front of the users. That's one of many

use cases that we see on a regular basis.

Our two technologies working great together.

Yeah, we're doing that all the time.

Seems really great to share that. Now, I

think zero copy is really part of a bigger

story about openness. What does an open

ecosystem mean to you and to Snowflake?

We're very committed to interoperability

of technology and data in particular. And

it goes at the bottom of the stack, which

is Apache Iceberg-based open data. We also

do it for catalogs. And it was Salesforce

and Snowflake a few weeks ago. We were

making an announcement sharing how we want

semantic data, semantic models to be able

to interoperate. We announced the formation

of an open semantic interchange. And

it's, again, us collaborating towards data

and metadata not being locked in.

Amazing. And we've done so many great things

together, but there's a lot coming up. What

excites you about the future? What's next?

At the heart of it, we're all

living in amazing times. It's the

era of AI. It's the era of agents.

But in the enterprise, the

power of AI comes from data.

And I think that's where I see lots

of additional opportunities for us

to continue to collaborate and give

more power and more juice to those

agents and those amazing solutions.

Amazing. Well, thank you so much,

Christian, for coming today. Everybody,

please give a big round of applause

for Christian. Michael, thank you.

We really appreciate the partnership.

So we just showed you some success we're

having with zero copy in the ecosystem,

but we're not alone. We're seeing

so many companies really get success

with zero copy. And in fact, nearly half

of all the data and data cloud in H1

this year was accessed from zero copy.

And with the success we're seeing,

we're really proud to announce

a bigger zero copy network.

So now we have IBM, Zero Copy File Federation,

and more than 108 new sources now

available in beta for all of you to try. So no

matter what kind of data system you have,

you can seamlessly connect it to Data Cloud

and activate your data to create success.

Now with this, I'm going to hand back over

to my partner, Rahul, and take us on home.

Thank you, Michael. Thank you, thank you.

So Michael and Salesforce are our first

and best customer. Mark tells us that if

we don't make Michael and Customer Zero

happy, we have not earned the right to ship

software to you. So I'm delighted to see

Michael being successful. I want to move

on to another topic here. We're running

behind on time here. I want to move fast.

I want to acknowledge

Informatica. So Informatica was an

acquisition that we agreed to acquire

that we announced earlier this year.

We're waiting for regulatory

approval. We do have the leaders

from Informatica here with us.

Krish, who is the CPO, and Bala, who is the

CTO. They're here today. Acknowledge them.

So we are delighted about the fact that

we're going to provide end-to-end

data management experiences and data

processing experiences, data fluidity that

I've talked about across catalog or

integration governance

MDM, et cetera. They're going to be

phenomenal synergies for you, the

customers when informatica really completes

post-close. So I want to switch to

a topic that is near and dear to me,

which is the data blazer community.

There's 70,000 strong now. You can join the

community. There's a QR code out there. I

do want to switch and talk about a leader in

the community who is really led in terms

of what you can do and what the rest of you

really have leveraged his services as well.

Mehmet Oran, who's here today.

Mehmet, there you are. I didn't

see you walk up. There you go.

So, Mehmet, so you've been a data cloud

customer and a data blazer for a few,

now it's been a while. So, tell us more

about your experience with data cloud.

Prior to using data cloud, I owned a large

stack. We had MDM, data lakes, integration,

a large team, and a large maintenance budget

that went with it. What I like about the

fundamental philosophy behind Data Cloud is

every key feature you need to bring data

together, to improve it, to act on it, are all

combined together under a common feature

set. And you can still have the choice to

leverage Snowflake or leverage Informatica MDM.

That never goes away. That means a total

lower cost of ownership and faster velocity.

Our first implementation was more than a

year ago, it was April of 2024, where we

migrated from HubSpot to Salesforce CRM

in three weeks, part -time team of four.

We started presenting about it. We

brought the historical data into Data

Cloud. That gave us the data foundation.

In the meantime, we identified overlapping

customers to ensure no unmanaged

duplicates for marketing engagement.

Fast forward to spring of this year, we

went live with our sales agent that used the

historical data all those activities no

one looked at plus our transactional data

in sales cloud and service cloud four

weeks with a part -time dream of three is

angelico in the room she might be i just and

i want to recognize her because last year

we didn't have a lot of these features

yet so we figured out how to bring it

together it has been great our sales team

is being more efficient and more

effective, powered by better data every day.

That's awesome. So agility. So you have

really talked about going live in three

weeks, having migrated with HubSpot to

Salesforce with Data Cloud as being the foundation.

So what's been the magic? What's been

the success there? What's the story

behind the success? Honestly, it's not the

magic. You have a clear use case. You

understand the capability of the platform. You

find the right partner. You find the right

people. and then it's a matter of

executing with focus and keeping on learning.

So tell us more about your role at the

DataBlazer community. So I think the

birth of this is a good story to tell

because when people were skeptical that

you could do something in three weeks,

people wanted to know what we did. But

then we quickly realized people wanted

to know what we did but how we did it.

So the one-hour presentations

turned into day -long workshops.

We show hands-on practice how to deal

with messy data, messy schemas. And it is

now a global group of everybody's helping

each other because none of us have all

the answers. But together, we grow better

together, and our impact grows as well.

So I want to switch over to roadmap.

So you've been privy to our roadmap. We've

talked to you lately about a few capabilities

are coming up. There are upcoming capabilities.

What excites you about them? Look, not

only we talked about it, I really appreciated

the opportunity to try them out pre-release,

and I'm excited to show what I have been

spending the part of my time last month just

experimenting. So let's spend a minute if

you want to show it to us. So my day job is

I'm the GM and data strategist at Peernova.

We do software focused on data

reliability. So we want reliable data in

return. But what I want to deliver to

every one of my teams is better data,

faster to make business decisions.

So the first feature that we were looking

at is Simple Start. And what it does is

instead of coming to a homepage, as pretty as

it may be, When I come to data streams,

every single one of my home CRM objects are

automatically ingested, and not only they are

automatically ingested, they are mapped.

So instead of spending the time finding

a data model, looking at help and

training, figure out what I need to do,

everything is ready within minutes of my

having turned on data cloud in my org. This

means we can focus on acting on the data,

figuring out what the business use cases are.

Simple Start gives us the ability to

go from turning on the feature to

start working in almost no time at all.

The second feature I want to focus on,

which is even more meaningful, is how we use

agents to bring additional data and improve

the setup and maintenance experience.

In this case, I have account registration.

This is getting data from Snowflake

via zero copy. And what I want to do

is ask my agent to map this object.

Now, ordinarily, as a data architect,

you would look at help and

training, you would look at the schema,

confess if you will create a spreadsheet

for this, and you're going to try to find

all of your target objects and hopefully

drag and drag sufficiently and effectively

to figure out what the right target needs to

be. While that is thinking, the previous

topic that he referred to, Simple Start,

where CRM data just showed up automatically

inside Data Cloud, that also got mapped.

That is free right now. Yeah, that's

correct. It's free. All CRM data is coming

into Data Cloud now is free. You

don't get charged at all. Yeah, and no

reason not to bring it in to act on it.

Now, in the meantime, thanks

for helping me fill the time.

The mapping recommendations are already

available on the screen. And what you see here

is because these agents know the target

schema possibilities, it knows what is in the

object I just pointed to, it gives me

recommendations on what are all of the things

I can map to, and it doesn't take action on

my behalf. I'm still accountable for the

correct design, but I am now saving weeks'

worth of work to focus on the hard problems.

One thing that I would say is that it's not

just the mapping agents. We also have

segmentation agents. We also have the agents

that are for activation. They're all coming.

They're all getting GA'd in January.

So, Mehmet, I think we want to

wrap up with the segmentation

agent. We can just talk about it.

We're coming very close on action here.

So what I want to do, and I'll let

this run in the background while it's

happening, is the best data is not going

to be powerful unless you can act on

it in order to drive business results.

One of the things that the philosophies

I'm embracing is, if I can think something

I want to do, I can type it, which

brings me the results, so I can test it.

As a result, I tweak it. I'm the professional

that's responsible, and weeks are

becoming days, days are becoming hours.

Rahul, someone is going to break

our record, which we were so proud

of. When does this become available?

It's right on the screen. It's all available

in January. Thanks for your pilot. Thanks

for running pilots with us. And thanks for

all the data blazers who are running pilots

with us. All of this is available in January.

So you're excited about going zero to

activation on your first day or in a single day?

I think this is going to be real.

Okay.

So I do want to switch gears

and talk a little bit about...

Mehmet Oran is the winner of our golden

hoodie, the first golden hoodie for data blazers.

So can we get him the golden hoodie here?

Thank you.

So all that grief space, all the

feedback, all the hours, it's been

great to work with your team. Yeah,

so this is amazing. Thank you.

so at this stage i'm gonna i'm gonna give

you an um i would call people and tell them

about what they should be doing the cta so to

speak call to actions right would you do

the honor of actually telling the folks what

they should do at dreamforce yeah uh thank

you thank you everyone um look i'm going to

say that the way to be here to share success

stories is the power of all of us. So go find

another DataBlazer. We are hosting and

networking our downstairs, Moscone West, second

floor, immediately following the keynote.

Come find other customers that share their

success stories. I'm going to do a session

Wednesday with real production screenshots

before with better quality I get

permission to do this and after so you can

learn from story and implement it yourself.

Get your hands dirty with the labs. Experiment

it in a safe setting. When you're skeptical

go to a round table talk to customers talk

to consultants find out how they succeed

so we can all be better together and then break

our record be on the stage instead of me

tell your story let us learn from you rahul

thank you awesome thank you everybody have a

great dream force, Mehmet Orhan. Thank you.