The Deep View: Conversations

Three years ago, we announced Firefly,

or we're coming out with more Firefly integrations.

And at that point, it was just like,

that felt like impossible.

It was like, it went from a platform

to all these different AI tools that teams were employing.

And the conversations we're having now

are all about the agentic application,

which is like the jobs to be done using those tools,

which is this really interesting evolution

where you're really starting to see,

how do I think about the work being done in my organization?

What is a job to be done that could be turned into an agent?

And how does that free up folks to do something different?

And how do we think about that in like a really structured way

to like make sure that we're exploring,

but then finding the value out of these tools that are being created?

Hi, Pat.

Thank you so much for joining us today

on the Deep View Conversations.

That's a great idea.

Would love to start with you telling me a bit about you,

how you got to this role,

and how you've been involved in this whole world of AI

that we're living in?

Sure.

Yeah, so Pat Brown, so my role at Adobe

is I'm responsible for global marketing execution

across media and email.

So think of like any campaign that gets out into the world.

The analytics associated with that.

So like, who are the customers,

who are the accounts if it's a B2B program,

and then all the kind of channel performance?

So is it working, essentially?

And then the bottom is all the kind of attack.

And my background is on the software side.

So at one point I was a software engineer,

so I like the technology.

But like kind of the platform and like the AI's applications

and some of that.

And that's a really fast evolving space, as you might imagine.

Yeah, this is my, I think, third or fourth Adobe Summit.

And there's been a really big shift, I think,

in how we're talking about just AI and marketing as a whole,

but also how marketers have to think about AI,

not just using it, but also how AI's

going to pick up the materials that you put out into the world.

I'm assuming you're seeing a similar shift too.

Can you talk a bit about that?

How marketing is changing because of AI search results

and all of that?

Totally.

I mean, it's funny.

Even so like three years ago, it would have been three years ago.

We announced Firefly.

Or we're coming out with more Firefly integrations.

And at that point, it was just like,

that felt like impossibly cool.

It was like commercially safe.

This is something that brands can use to do the work

that they need to do.

But it was relatively defined space of how it would potentially

drive acceleration and the campaigns that you're running

or emails that you're putting in the world or whatever

it is, websites, that sort of thing.

I think what's evolved is that was a very,

and then in the last, after that, the next year

was again, kind of about the tools and the things

that were actually building, where there'd be measurement tools.

So it went from a platform to all these different AI tools

that teams were employing.

Now the conversations, and as you heard the keynotes

and the conversations we're having now

are all about the agentic application, which

is like the jobs to be done using those tools, which

is this really interesting evolution,

where you're really starting to see,

how do I think about the work being done in my organization?

What is a job to be done that could be turned into an agent?

And how does that free up folks to do something different?

And that's the most conversations that we're having

is how do we think about that in a really structured way

to make sure that we're exploring,

but then finding the value out of these tools that

are being created?

They've kind of included the evolution,

because I think actually, when it might have been my first summit

that I came to, was the whole Firefly moment.

And that was what that was the heart of the event.

And then we kind of segue it into agents a bit.

But this year, it feels a bit different.

I think, again, my perspective, is that agents last year

was a bit of like, there's this really big promise here.

There's things that they could do, but it was like an isolated

little workflows, right?

Where now it's like, you could automate an entire customer

experience or an entire client portfolio using these agents.

Have you seen that evolution too?

I think that's well said.

And even last year, I don't think that there

was a clear definition of what agent

that even really met last year.

It was being thrown around by, I think, a lot of folks,

even in the market.

And I don't know that everybody really

was on the same page for what it really was.

I think by now, we've more defined it.

And it's something where like, OK, so now I understand

this is the interface by which I'm going to get a bunch

of those tools or that work done.

It's going to be, it's going to go, and it's

going to use all these things that I want to do, and come back

and show me something.

And that's a really interesting evolution.

And you're right.

It's no longer bounded to just like AI tools,

but it's like agentec AI, non-AI tools.

There's a bunch of different ways.

And it becomes more the way that you're

going to engage with them.

And where are you?

Are you a cloud user?

Like, what is the UI?

I want to have access to those skills of those agents,

but they've got to be where I am.

And that's driving a really different conversation

for like how we work.

What is a creative professional sign

into every day?

What is a media person sign into every day?

And that's a big shift.

But they are really like about being here at Adobe Summit.

Specifically is that the entire audience is really marketing

professionals or creatives in some way.

So when we're talking about agentec solutions that

are like, again, now taking up an entire workflow

and doing it autonomously ideally,

how do you kind of marry the idea of these are creatives

who've gone into the field because they like using their creativity

and they like expressing their thoughts

and their ideas and visions in that way?

Well, also kind of automating it and giving it to a bot.

Yeah, it's interesting.

And this is a, we spend a lot of time

thinking about this.

And it's the way that I think about it

is the agent is just a function of like the context

that you've provided it.

And it might be brand-based context.

It could be performance.

It could be a customer-based knowledge set.

But somebody is working to make sure that that agent

or the information that is being created for that agent

is up to date and has passion around it.

So like if you take someone that feels like they're responsible

for like the brand articulation at a company,

there's going to be some agent within the ecosystem

that is interpreting that brand articulation.

It could be existing assets that have been run.

It could be a literal documentation of a brand

standard or a design system.

But that's an ever-changing thing.

And you can continue to think about how you want

to feed context into that agent.

But like that's a full-time job.

Think about the number of times, even from a Adobe perspective,

we develop an amazing campaign that we all love.

But it goes through these permutations of edits

and people's perspective to like make tweaks

because the world has changed.

How are we flowing those back into the context

of like the actual agent itself?

That's a full-time job for someone who thinks about that.

It's just, I think the concept is similar,

but the way that you do it is just really different

using these tools.

The agents are just a function of what we tell them

in the context we give them.

I couldn't not touch on the agents for creatives.

But I think that's everybody's first question

when you hear about like agent

and any capacity of image generation, video generation,

asset generation at all.

But I would love to talk a bit more about something

that's really called my attention here at this conference.

It's using these agents in marketing,

but maybe less so for the creative work,

but more so for understanding those analytics.

And I think that's your expertise too.

So I would love to hear about the value that there is there.

So that's where I grew up was in stats and computer science

and all the analytics and I love it.

And marketing science, I did that before that was a thing

before it was called that.

But I think that in some ways the creative side

is much further ahead grappling with some of these changes.

And I would say the analytics infrastructure

because it's so varied by every single company

is starting to work through that.

Same for us at Adobe.

And I'll show you even going all the way working back

from a dashboard or a BI sort of infrastructure.

We have lots of conversations because like any company,

we have a gazillion of these dashboards.

Everybody's got different ways that they want to consume

performance and how are things working.

And the question you have to ask yourself is

how do you, from a good analyst perspective,

reduce the time it takes for people to understand

what's happening?

Is that an agent that then is helping you understand

the best way to get to a visual construct?

Do you need a dashboard or do you need a design system that

articulates the way you want to look at your analytics?

And it's really evolving even how you think about and define

like the visuals that you're creating.

And then think the way that you're getting the analytics done.

And so same like we're having on the creative side,

we're thinking about like the equivalent of like a brand

system for like a visual system for your analytics.

Like, do you want to see trend lines?

Do you want to see decomposition?

And like modularizing kind of the jobs to be done.

And then like, where do we want the human time to be spent

on the synthesis across all these different pieces?

So I think it's, I saw a great, actually a great piece

of marketing.

I think it was HSBC.

And it was this AI-focused imagery.

And it basically said, what will you do when you don't have to do?

And I love that.

I thought the tagline was so cool.

And I give it to my analytics teams.

And I'm like, when you're not, you know,

you're nobody's writing SQL anymore.

When you're not doing this really grindy work anymore,

how are you driving contacts and like driving decisions?

And I think that has opened the teams up

to think a little bit differently.

Now it's time for a word from this week's sponsor area.

Bringing AI into your company is exciting,

but can also feel risky.

Between data privacy concerns, security threats,

and constantly evolving compliance requirements,

it's easy to hesitate.

That's where area comes in.

Area provides a unified security layer

with advanced threat detection and strong governance tools,

designed to protect your AI ecosystem from day one.

You can safeguard sensitive data, monitor activity,

and stay aligned with your compliance standards,

all while continuing to innovate at speed.

So instead of slowing down because of uncertainty,

you can move forward with clear guard rails and real confidence.

You don't have to choose between innovation and security.

Area helps you do both at the same time.

If risk has been the thing holding you back

from going bigger with AI, this is your green light.

Visit area.com to learn more or start a free trial.

Go to AIRA.com.

That's AIRA.com.

We thank area for their support of the deep view.

Now back to the show.

Right, I think there's something to be said there too about,

you don't have to necessarily spend maybe the time

in a dashboard digging out the analytics.

And maybe you could use a conversational prompt and be like,

hey, why was that?

I don't know, the click-through rate as low.

So was there this or that.

But it still takes the knowledge of understanding the landscape,

understanding the company's vision, brand voice, tone,

all that to make sense of that data.

Totally.

You still need the concept of an applied analyst

that is an analyst that understands the business inside and out.

We'll understand how to ask and articulate

the right questions even into any sort of infrastructure.

And then you still have the age old analytics problems

of, am I producing things that people can understand and self-serve?

That's something that you really try to write,

because you want to push the insights as close

as you can to people making decisions.

And then you still have the problem of, I need truth also.

So the agents are going to get me so far.

But I still need to have good analytics practices

to make sure I'm doing testing, I'm doing structured experiments,

I'm doing all my other things that feed into these capabilities.

This conference was also the first time I've heard

on a main stage, a term, geo, and AO.

And I just recently read an article,

I think it was two or three months ago,

and that was my first introduction.

And it's basically, well, you could probably explain it to me a bit better,

but how AI is affecting search engine results, right?

And that's impacting how brands reach their audiences.

Yeah, it is.

And this is a really fast and moving space, too.

And we see this, and we've seen this in even some of our websites,

and our website traffic, and early on,

one of the things that I think a lot of brands saw was,

as people were searching, and they were being surfaced either in,

they went natively to chat GPT, or they saw it in an AI result in Google,

or pick your engine, they were being provided informational results back.

And it was more like, how do I, for us, it might be,

how do I do an extension of a frame in Premiere?

Or like, how do I crop this image in Photoshop?

Now we have millions of pages on our adobe.com,

there's some that explain how to do this, and try to make this in all different ways.

But if that now is being served up by some of these tools,

you don't see that traffic coming to you.

So you're like, oh, interesting.

How do I think about that?

That informational search is now being, you know,

zero click is the term, because it shows up,

and like, oh, I got my answer, and I can move on.

That change is a just sort of thing,

think about higher intent things,

where people are trying to get work done.

If they're like, I want to crop this image,

and then there's an opportunity for them to do that live,

how do we think about the role of our surfaces,

as we start to work through, like our infrastructure?

And so this is the kind of way that we're starting to think about,

how are those engines actually collecting the information that they're collecting?

I think increasingly the chat GPT and even Google,

they've gotten better about citing the sources by which they're pulling their information,

because they're just repeating back in a parsed way what they see on the web.

And so what we've done is,

build into some of the tools,

better way for brands to see, like, which of those citations

reddit YouTube that are like, explaining the information that is being rendered,

because as a brand, you want to make sure that's accurate first off,

and increasingly like, it's updated,

and if you have a way to like, go out and like,

provide best information to like, give a better result than great.

But that is a little bit of like this.

It's like, early days of SEO is what it feels like.

Right, and like, to bring this all so full circle,

when I did hear about it on stage, it was during demos,

which I thought was pretty neat.

And it'd be again, someone like looking up a question like,

what can you help me understand,

and it would be a specific, and I like about why this page is maybe underperforming,

this landing page, or whatever, maybe,

and then they would use the term,

A-O or D-E-O, and then Adobe's tools were able to kind of give some context

and explain that a bit.

I'm curious as to how you've, I guess, trained the tools to be able to help with that,

even though it's such a nice and space, and I think we're all whether it's like media,

myself, you know, publications trying to figure out how do we still rank,

how do we still show up, or whether it is a brand trying to get their product,

or their small business trying to rank on Google,

we're all asking ourselves these questions.

So I was really curious as to how this product kind of has the answer that they were all looking for.

Honestly, it's a little bit of like using the data that you have,

and looking for absence of things that are showing up, like very specifically.

If there's one thing, we have a, you know, the bedrock of a lot of the data infrastructures,

all the web analytics data, it's all the targeting,

it's all the data that we work through.

If you start to see changes in this, for example,

we saw informational searches falling, like if you have your pages categorized,

we literally looked for the absence of a signal,

and then it was enough to throw a flag up,

which is now in the early day, the tool that we just announced.

It was able to say, like, looks like these actually have declined.

This category of pages are seeing less traffic.

In parallel, then we released the capability working with the JetGPT in Google

to like start to understand and ask questions of like,

if I were to understand this category is missing,

people looking for this information,

what are the results that are showing up in these platforms?

And you started to see a correlation between the way that it was answering,

and the fact that it was giving a result and the decline in pages.

And then from there, we started to like look across all kinds of other data,

to start to then connect these different journeys,

to be able to provide marketers with a perspective on these.

The SEMRUSH acquisition that we're hoping closes soon,

that we're really excited about, they're another example of this,

where they've been approaching it in a slightly different way,

where they're using some of the SEO tools,

where they're looking at like, again,

how are these searches showing up, where are their citations?

What is the ranking of those citations?

I feel like it's going to change every week,

just like SEO did in the early days until forever,

but like, we're able to at least help marketers figure out like,

where the patterns are starting to show up,

so that marketers can go and start to dig and figure out what's happening.

So this kind of reminds me of like,

what's going on right now in cyber security,

and honestly, what's always gone on in cyber security,

of like this cat and mouse chase,

and it's like, you know, the more advanced that technology becomes,

the more advanced that that tax become,

and then defenders become more advanced,

and then just keep all evolving,

becoming more advanced altogether.

Would you say it's kind of similar, I guess,

right now for marketers, for organizations,

whether it's a media publication or a small business,

whoever it may be, who was confronted with this,

signals are changing, search engine optimization,

all of it is not what we know anymore.

Would you say using these AI tools is kind of a way to almost,

again, combat the AI, maybe traffic crash that is happening?

I think it's a good way to think about it.

I think that what it's actually also reasserting is the importance of like,

having a brand, and then also like,

having an experience that customers feel value in,

because increasingly as these things are being surfaced for free,

effectively, you're not for free,

but like without having to do any additional work,

I do think customers still have a connection to a brand,

or like an experience,

or like a, and they want to still come in and experience that thing.

One of the things that we saw early on was,

while informational clicks in this category was declining,

some of the brand components of like the traffic

and people looking for, actually stayed up

and were relatively durable during this time,

and we didn't see any issues.

So I do think that there will continue to be this cat and mouse game

of like brands in the zero click of like how we think about that.

I think customers ultimately are going to adjudicate this.

I'm like, am I really just looking for like,

how to crop this thing?

Or like, what's my relationship with Adobe, or pick your brand?

Or like, and that's the thing that we have to rethink

even the value exchange with customers.

I'm like, how we're making sure that we're giving,

building a connection really closely.

It's like, this is like marketing one on one 50 years ago.

It's like bringing back some of those same concepts.

So from your perspective, again,

having seen these different shifts,

haven't studied it, again, before,

being a professional, as mainstream as it is,

are you worried about, again, SEO being dead?

Or even marketing or analytics jobs that like SEO professionals

are constantly like, what is going on with AI results?

Does that mean for their job stability?

What does that mean for brands?

Like, is it worth investing in websites and all these things?

People are just going to click on, you know, a chat,

to be tea result.

And from your perspective, is there still value?

I guess, investing all those things, still future there?

Or is AI really just decimating it all up?

I'm more bullish on all of this.

I think it's just a different paradigm

on how people are getting information

and how they're serving information.

I think that there will continue to be people

that are trying to navigate to get somewhere.

Like, there is still going to be a way that they want to go

somewhere, and they're going to be looking for information,

and they're going to be trying to complete a task.

And like, there's going to be some like bedrock components

of what people are trying to do on the internet,

and like on the web, and like in digital properties,

that will continue.

How they do it is going to change.

No question.

But I think the core components of like the practices

that we built around, like building some of these capabilities

up and like making sure that we're findable,

and people can come in, like, we're giving value back,

that will continue.

It'll just evolve a little bit.

Interesting.

And then if your early career, as you, again, we're starting off,

what tool that exists now would you wish you had then?

What tool would I wish now?

Yeah.

It could be a tool.

It could even be just a little workflow to like an agent.

Like something I'm just curious that you think

would have unlocked a lot of value then.

I think we spent so much time just consolidating and synthesizing.

If you think about these massive problems of like,

I've got all these different components,

and I just need to like summarize all the different things

that are happening.

Like consolidation is such a hard problem of like,

whether it be data sets, it could be text documents,

it could be a bunch of things.

I think that is just such an unbelievably powerful capability,

just to understand like what is happening

across all of these different systems,

so that then at least know like what the common themes are

that I can start to react to.

And I think that is at least one of my favorite use cases

for some of the actuals.

Yeah, so in with that, what would you advise

to someone who might be either a marketer working on

this new project or this new brand,

who they want to obviously perform and show up and do well,

or whether it's just a, again,

a business professional targeting it solo

and currently trying to get started,

what would you advise them to be first steps

into navigating this new space of, you know, AEO

or AEO and also using AI tools that are at hand now?

Yeah, I think so it's kind of two different things.

It's like the exploration required to like,

just be interested in the space,

which I think we're fortunate enough,

I've hired some really, really smart folks

that just really like to poke around in the tools

and understand what the journey is actually look like.

And how are people leveraging them?

But then I think it's also being grounded in

the core elements of even marketing and like customer journeys,

because at the end of the day, like independent

of all the AI stuff, we're still trying to like,

help customers do things.

And we, the superpower of marketing is that we are also

consumers and being marketed to.

And all of us have a perspective on like,

actually that was a really great experience

and actually got to what I wanted.

And so we can actually adjudicate that ourselves

on like, this is actually what good looks like.

So even as you're like exploring,

how do I employ those things to like make sure

there's a really, really great customer experience

at the end of the day.

And like customers are finding value

in the thing that I'm marketing.

And honestly, the biggest thing in Adobe is this has

really collapsed the world between marketing and product

and like website and all the different surfaces

into like one journey.

And I think that's a really exciting superpower

to like think holistically about like,

how customers are moving across all the different surfaces

regardless of where they sit organizationally.

Thank you so much, Pat.

It's been so fun to take your brain

of all of your years of expertise.

Thank you for being so generous of your time.

Likewise, super fun.