Retailgentic, presented by the ReFiBuy Media Network and hosted by Scot Wingo, is where Agentic Commerce meets retail innovation. We help retailers and brands prepare for the future of agent-driven shopping with actionable news, expert insights, dxis, forward-looking predictions, and practical guidance on Agentic Commerce Optimization (ACO).
All of them, whenever they introduce new
features, always talk about, you know,
co- whether it's Copilot or ChatGPT or
Perplexity, the, or even Meta AI, right?
Talk about the personalization aspect,
because they now all have a memory,
so they remember your conversations.
And for ChatGPT, you can type, "Remember
that my grocer is Whole Foods," or,
"Remember that I am gluten-free."
Um, and then it just remembers it.
So, uh, again, and Mark Zuckerberg has
said that he thinks that Meta AI will
sort of dominate the space because not
only does it remember your interactions,
but it pulls in all of your social media
interactions, so it's an even richer
profile, and that people will keep using
the agent that knows the most about you.
Welcome to the Retailgentic Podcast,
where we explore the intersection
of retail, e-commerce, and AI
agents, which we call retailgentic.
Our mission is simple.
We are shining a bright light
on the new agentic AI shopping
megatrend, which will disrupt the
trillion-dollar digital retail market.
The stakes are high.
Google, OpenAI, Microsoft,
Perplexity, Amazon, Apple, Meta, and
Anthropic are all fighting to be the
consumer AI shopping agent winner.
There are gonna be some tough decisions
for online brands and retailers to make.
What happens to checkout?
How should payments work?
Should you continue to invest
in retail media networks?
What does the world look like when
over half of your traffic is not
from humans, but from AI agents?
These existential questions
are the tip of the iceberg.
On this podcast, we're talking to industry
experts, analysts, and influencers
to get answers to these questions,
identify trends and best practices.
This is Retailgentic, your
podcast partner in navigating
our AI shopping agent future.
And now here's your host, Scot Wingo.
Welcome to the Retailgentic Podcast.
On today's episode, I'm excited
to host Michelle Grant, Director
of Strategy and Insights, Retail
and Consumer Goods at Salesforce.
I first met Michelle about 10 years
ago on the e-commerce trade show
circuit, and she was with Euromonitor.
She's always been a huge wealth of
information, and we always have had her
on the Jason and Scot Show to help us
dig into lots of interesting trends.
At Euromonitor, a lot of it was around
cross-border trade and what was going
on in Europe, as you would imagine.
In this podcast, I catch up with
Michelle and, uh, understand more
about what her role is and then at
Salesforce, which is kind of newer to me.
Uh, and then we jump into what
they're seeing in their data.
Salesforce has two sources of data.
They have survey data, which
is what we would call stated
preference or qualitative data.
But then they also have, thanks
to the Salesforce cloud platform,
previously known as Demandware,
a lot of quantitative or observed
data from consumer behavior across
their, their wide range of retailers.
Thanks to this data, Michelle reveals,
uh, based on some Q1 insights that they
have, some details around e-commerce
traffic trends, and then picks out
what's going on with the ChatGPT and
other LLM component of that, how it's
converting compared to other types
of traffic, which I think is a really
interesting harbinger of, of what that's
gonna mean as that traffic increases.
Uh, and then, you know, if you followed
Salesforce at all, Marc Benioff, their
CEO, is really leaning in on agentic.
So they're-- they have Agentforce.
Uh, they may even rebrand as
Agentforce at some time, I don't know.
Uh, and that's a huge area for them.
So she has a lot of really…
They're having a lot of conversations
with retailers about agents, and
she shares where they're getting
traction, where, where retailers are
most interested in how they're gonna
use agents in their own operations.
Then we jump into some generational,
back into the consumer behavior.
We jump into what's going on
with some generational data
she has that's interesting.
And also she surfaces some data that
indicates consumers are going to be really
a bit cautious with agents at first, so
there's a trust element that we talk about
that I think you'll find interesting.
At the end, I always like to get
people, uh, out of kind of the
current quarter and thinking forward,
which is always hard with public
company folks, but Michelle was game.
And we talked a little bit
about where she sees this going.
And one question I like to ask is, will we
need websites if, if 100% of people shop
with agents in this thought experiment?
Are websites this vestigial thing
we just don't really even need
anymore or, or what happens to those?
And Michelle has opinions on
that that I, I really enjoy.
I hope you enjoy the
conversation as much as I did.
And now here is my conversation
with Michelle Grant from Salesforce.
Hey, Michelle.
Welcome to Retail Gen Tech Podcast.
Thank you for having me.
I'm delighted to be here.
Cool.
So we were just, uh, in
the green room, as it were.
We were chatting, and we've known
each other for over 10 years.
So, uh, you've had…
Uh, I've been out of it for a while
doing some other things and back in
e-commerce, and you have Salesforce
there on your title, so I'm excited
to hear, uh, um, you know, how, uh…
And you've, you shortened your
title, your official title.
Uh, I love it, 'cause I
always give Jason a hard time.
You may have more words than he does.
I'll have to fact check this, but
Director of Strategy and Insights,
Retail and Consumer Goods at Salesforce.
That seems like a pretty big umbrella.
Um, tell us what you do at
Salesforce and how you got there.
Yeah.
So if we take a step back, when we
first met each other, I was, uh,
the head of research for the retail
industry at a market research firm
called Euromonitor, and my role there
was to oversee the program that we
run, uh, or they run still, on market
sizes, market shares, forecasts, uh,
for the retail industry, 30 different
channels in almost 100 countries.
Um, but I was looking for a new challenge,
and came across a great job description
and posting from Salesforce where I could
use those market research skills and that
industry expertise, but apply it and look
at it through the lens of technology.
So I joined, uh, Salesforce in March
of 2020, which is an interesting
time to be starting a new job, and
that's a whole separate discussion.
Um, but here I use those same skill sets
to really track the trends impacting
the retail and consumer goods industries
using a lot of different methodologies.
So I get to keep being a student of
the industry and sharing my insights
with our customers and with the world.
So really, really excited to, to be in
the tech industry, uh, but also still
do market research and then study the
industries I'm really passionate about,
which are retail and consumer goods.
So yeah, my title is
probably longer than Jason's.
Awesome.
We'll have to add a l-
another six words to this.
Um, have you got to meet Marc Benioff yet?
I haven't.
Um, he did stop by our booth, uh, at NRF,
um, last year, but I was at a customer
event, so I missed that, unfortunately.
Y- yeah.
At least you weren't, like,
uh, just chilling out or, like,
late to the booth or something.
Exactly.
Uh, yeah.
And then do you get to go to
the, their big show every year?
I do get to go to
Dreamforce in San Francisco.
Yeah.
Yeah, I've, I have not ma- I went to
a very early one, but I've heard it's
turned into quite a spectacle, and there's
usually, like, a really big, like, you
know, band and a whole, a whole thing.
Yeah.
Multiple bands.
He always has, like, really
cool shoes and stuff.
A big concert.
Yeah, it's…
Yeah, and the weather in San Francisco
is wonderful, so there's so many
outdoor activities and activations.
It's, it's very impressive.
Now, are you attached to the Commerce
Cloud group, or you kind of can, you
kind of support these categories across?
So maybe you would talk to a
customer that's just using CRM or
one of the other Salesforce clouds-
Exactly
… that I'm trying to think about.
Yeah.
You're in a, you're in an industry
vertical, not necess- Now, obviously,
you probably do talk to the commerce
folks the most because that's your
people, but, uh, I'm sure it comes up
even in a broader sense given the wide
scale of Salesforce's customer base.
Definitely.
So, yeah, so the, the clients, you
know, have to be in the vertical
of retail and consumer goods.
That's where my expertise is.
Um, but they could have any
different product like Service Cloud,
Marketing Cloud, Commerce Cloud,
a combination of all the clouds.
Um, so I have to know all, about
all of our products as well to
see how, you know, the trends
impact that, uh, uh, capability.
So yeah, it's really fascinating.
My remit is very, very wide.
Yeah.
One of the f- reasons you were one
of the first people I reached out
to is both you and I are data nerds,
so we, we share a love of data
and stats and, and these things.
Um, and then being embedded in Salesforce,
you're kind of at, uh, y- you have access
to a lot of data, which we'll talk about.
But also, uh, you know, agentic, uh,
is, is a big focus, uh, of not only
Benioff, but the whole company now, so I'm
excited to kind of dig into that and…
Is that, are customers asking you
about that kind of thing at this point?
Has it rippled into your,
into the retail and CPG?
Okay.
Oh, yeah.
Yeah.
Definitely.
That's top of
mind.
Yeah.
Good.
And again, in a lot of different
areas that we'll talk about.
Yeah.
Let's talk about data, because
you guys have Commerce Cloud
and then all the other clouds.
Um, so, so you have kind
of aggregate data there.
Um, and then you guys also do…
So that, that gives you kind of the
transactional observed behavior.
Uh, you know, I'm not a statistician, but
I guess they would call that quantitative.
And then you guys also get the kind
of opinion type of data as well
because you run a lot of surveys,
which is more the qualitative.
And, and that's pretty interesting
'cause it allows you guys to kind of
see, you know, what are people saying,
which is a little bit of a glimpse
in the future, and then what are they
doing, which is kind of like the, you
know, what are they doing right now?
Um, any-- tell us a little bit
about where that data comes from
and, and like the last time you
ran the samples and, or whatnot.
Right.
So yeah, our, um, sort of, like you
said, the first thing, uh, kind of our
proprietary data set is being able to look
at what's happening on Commerce Cloud.
Um, and we publish, uh, a lot of metrics
quarterly on our shopping index, and
that data is coming from Commerce Cloud.
Obviously, we clean it, we anonymize it,
we privatize it, um, you know, to be all
sorts of compliant before we publish it.
But that does give us a look
into what's happening in, in
real time on the platform.
Um, you know, it's not indicative
of the whole industry, right?
Um, because it is a subset of e-commerce
websites, so it wouldn't necessarily…
You know, it's a good look, a good
barometer, but it's not necessarily
reflective of the entire industry.
So, you know, every data source
has sort of their asterisks.
Uh, and I want to give
that kind of caveat.
But I did dive into Q1 2025 data.
Um, you know, we've been keeping an eye
on the generative AI or agentic trend for,
you know, basically since the beginning.
Um, and we're able to see the
sources of traffic to our website.
So took Q1 2025, you know, again, that
quarter kind of has caveats, like Easter
was in March, it's in April this year.
Uh, it was a leap year,
it's not this year.
We didn't adjust for any of that.
And so it was interesting when we
look at traffic overall, it was down.
Um, and I don't know if it's because
of those events or the weath- uh,
the fires in LA, unfortunately.
But it was a, it was a
tough quarter overall.
Um, but when we look at, uh, traffic
that's coming from agents like ChatGPT,
Google Gemini, uh, Perplexity, we see the
traffic is, uh, like from those sources
is less than 1% of all traffic sources.
So it's still really small.
Um, we did see-- I did look at the
verticals, and it's more impactful
for, uh, skincare and health and
beauty, interestingly enough.
And the biggest source of
that traffic is ChatGPT.
So people are using that,
uh, to find, uh, websites.
Um, and it was the fastest
growing channel for traffic.
Uh, it was up 80%.
Again, that's from a really small
base, less than one percentage of
the, of overall traffic, um, in--
but overall double-digit growth
when overall traffic declined.
So that's, that's pretty interesting.
You know, last week, uh, Eddy Cue
was on the witness stand as part of
this lawsuit that, that Apple's--
I guess it's the Google antitrust
lawsuit, but Apple got drug into it.
And, um, you know, they said for
the first time ever, traffic from
the Safari integration they have
with Google was, was down month
o- over month, so, um, and then he
specifically cited people using ChatGPT.
So it's gonna be interesting to
see when that starts to spill
over more into the retail world.
We, we've heard, uh, and I shared this
with you earlier, but we, we've heard
from some folks we're talking to that have
reached out to us, um, and, you know, I
don't know if it's just bot traffic, but
there's, there's a lot of noise out there
of, you know, these unknown things hitting
people's website trying to track data.
Um, and then, um, you know, they
can't tell if it's, if it's people
using the bots and kind of coming
from citations or, or what's going on.
So it's a little bit chaotic
right now in retail land from what
we've, we've been able to tell,
uh, from talking to customers.
It really is.
And so I did look at, like, traffic
coming from search, and it was flat,
um, from Q1 2025 versus Q1 2024.
But what was up was really s-
what, what was up was social,
uh, traffic from social media.
So I think the Gen AI and the
agentic, um, storyline has kind of
overshadowed the storyline that, you
know, research from eMarketer, TikTok,
and Google itself showed that younger
generations are actually starting
their product searches on social media
first instead of the social engine,
or social-- or the search engines.
So, kind of Google is maybe facing a lot
of different, uh, areas where its search
dominance is getting nibbled away at.
Uh, and I know it's a small
sliver, but was there anything else
interesting about that traffic?
Does it convert as good or better?
And, um, I know you mentioned
ChatGPT is the dominant one.
Is Perplexity showing up at
all, or is it, like, it, it's
just kind of noise in the data?
It's kind of noise.
It really is, uh, dominated by ChatGPT.
And that- Yeah … when you look at
the data as far as, like, downloads
and monthly active users, like,
ChatGPT is so much bigger than
everyone else really at this stage.
Yeah.
How does it, um, convert
compared to other traffic?
Yeah, it's on par with direct traffic and,
um, and search traffic, so that's good.
Um, it's pretty healthy,
uh, conversion rate.
Um, and I mentioned social media
that there's a whole separate podcast
on like, why is social media con-
traffic conversion rates so low?
Um, especially since
the traffic is growing.
Um, but yeah, so even though
social media traffic grew very
strongly, that type of traffic has
never really converted very well.
So is it a consumer thing or is
it a technical integration thing?
I don't know.
But yeah, as far as conversion rates
go, when you're, when someone is looking
at a ChatGPT, it's, it's up there
with, with, um, direct traffic and
search, so they're very highly linked.
Yeah.
If I were to hazard a guess, it's kind of
where they are in their consideration set.
So if they're kind of researching, which
is social media, like, "Oh, I was just
doom scrolling and I saw this interesting
beauty product," you know, they're,
they're not really in the purchase funnel.
Like, you may serendipitously
grab them there, but it's
gonna be pretty low converting.
Whereas if they're researching, um, if
they're going direct to your site or
researching, they've done a bunch of
beauty care product research on ChatGPT,
and they're getting pretty close, then,
then they're further down the funnel.
It's probably a funnel, some kind
of a funnel thing would be my guess.
Yeah.
I agree.
How about, uh, okay, so that's
our, our, uh, quantitative data.
How about qualitative?
What are shoppers saying, uh,
when you guys give them surveys?
So we just released at the end
of March our, uh, Connected
Shoppers Report, the sixth edition.
Um, we update this about every two years.
And this year we surveyed 8,350 shoppers
and 1,700 retailers, uh, in 21 countries.
So it's a really big,
uh, research undertaking.
And we did ask about, um, about using
agents for product research and discovery.
39% of, uh, survey respondents said
they'd used it for, um, product research
and discovery, and of that 39%, uh, 12%
said that they were using it frequently.
But when we look at generation splits,
54% of Gen Z said they've used, uh, like
a ChatGPT to research a product, um, and
then 20% of them are using it frequently.
So again, there's kind of that what
you would normally see in generations
when it comes to adoption, Gen Z
is really leading the pack here.
Interesting.
So, so it's again tying that…
So is, is the counterpoint that the
other generations are still kind of using
social more for these types of things or?
Yeah.
Social or search, direct,
uh, could be- Yeah … email.
Um, but yeah.
Yeah.
Interesting.
There's a whole bunch of TikToks and
Instagram about, uh, for Gen Z showing
people how they're using ChatGPT
for s- outfit styling or beauty.
So it's really interesting that- The
generation is using social media to teach
other people how to get the most out of
these agents when it comes to shopping.
Did you, um, uh, so you
looked at generations, did
you also slice it by category?
Any interesting category data in there?
We did.
We asked, you know, kind of what is
your interest level when it comes
to doing different activities?
And the number one result was,
uh, getting gift ideas, right?
So plugging in the demographics of, you
know, I don't know, I never know what
to get my dad for Father's Day, right?
Uh, and so you put in, you
know, he's, you know, St.
Louis Cardinals fan, uh, he lives in
Chicago though, um, you know, and you
put in all these, uh, demographics and
characteristics, and then you can get any
kind of gift idea, whether it's a book or
a shoe or, you know, a Cardinals apparel.
Um, so yeah, I think
that's really fascinating.
And the second one that was a
really close tie was researching
electronics and appliances.
And I think that's a little bit of like
kind of an early adopter of technology.
You probably purchase a lot of technology.
Um, men tend to use Gen AI a
little bit more than women, so it
could be a gender thing as well.
Um, so that was sort of the, the
second most popular use case.
And then, um, the trend we're following
most closely is-- So, so a lot of this is
just kind of maybe comparison, compares
to comparison shopping engines, right?
So, or Google.
So, so sending traffic downstream.
We have a theory that, uh, and it's right
in our name with the, the agentic part
of it, um, that eventually these agents
are going to transact, uh, at a higher
level and go visit the, the website for
the human and the human, the, the human
operator will never get to the website.
Um, is that something you guys think
about and, and have a point of view on?
Oh, yeah, definitely.
Um-
Being so into agentic, I'm, I'm
sure you guys think a lot about it.
Exactly.
Uh, pretty much everything we do
has to have an agentic angle to it.
Uh, I wouldn't be surprised
if Benioff does change the
company's name to Agentforce.
Um, but yeah, so we, we, I mean,
to be honest, I thought That this
technology was kind of like years away.
We've already seen some examples
come to life, like with, uh, OpenAI's
Operator and Alexa+ being beta tested.
So I feel very, uh, I feel very
smart that we had the foresight to
ask consumers about what they would
like an AI agent to do on their
behalf without any input from them.
And what we found was that
shoppers want to take the
friction out of the experience.
So they were very interested
in having the AI agent, uh, do
customer service for them, right?
"I have this problem, go
talk to the retailer and get
it resolved on my behalf."
Um, same with returns.
"If I have to fill out forms or
schedule a delivery or pickup for
my return, have the agent do it."
The fun one that I snuck in was
optimize loyalty points, right?
We have credit cards, we have loyalty
points in, uh, maybe with a brand or
a grocer, and having an agent sort
of analyze where you should redeem
your loyalty points to get the best
value out, um, was up there too.
So always thinking about kind
of how loyalty programs play
into the retail experience.
But sort of when we asked about
actual transactions, um, purchases
made on behalf of someone, um, the
interest level there was a bit lower.
It was kind of at the bottom of the range.
Um, so we asked, "Would you be
comfortable having an AI agent
buy complimentary products based
on what you've already bought?"
Right?
So this, these AI agents are
gonna remember everything about
you, including past purchases.
So shouldn't they just-- if you
bought shoes, shouldn't they
just like buy you the socks?
Um, that ranked low as well as
just making purchases that it
automatically recommends for you.
So people aren't really…
And then when we asked like,
"Okay, what would change your
mind about using, uh, AI agents?"
Uh, like what sort of factors
would boost your trust in them?
And the, the top one was data privacy
and security protections, right?
Everyone does understand how
much data these, uh, ass- these
agents, um, gather on you.
So really making sure that that's
all secure and private is important.
And the next one is the ability
to turn it off and on easily, and
then approval for any purchase.
So like It's cool, but I don't think
people are really ready to, like,
completely outsource their shopping
without any sort of input into that.
Yeah.
They wanna kind of stay human
in the middle and have the agent
go and kind of com- ping them
back and that kind of thing.
Interesting.
A lot of the new wallets that
have been announced have that.
They have a consumer-facing piece to
them where the agent has, uh, you know,
kind of like a budget, if you will.
And then you can set a flag, you know,
before you spend your budget, I, I wanna
check off on everything kind of a deal.
It'll be interesting to see how long we go
before people get comfortable enough that
they start to take those rails off and-
Exactly.
And maybe, you know, maybe it's
certain categories, you know,
that you can automate, right?
Like I, I don't need…
You know, here's my weekly grocery.
I don't need to be approving that
all the time because I, it's a list.
I already know what I'm going to buy.
Um, I'm just doing a replenishment run.
So that'll be, it'll be interesting to
see how it, if there are differences
in product categories as well.
So that's the consumer side of things.
On the retailer side, um You know,
our, uh, my understanding is within
Agentforce, one of the first ones
you guys had was kind of in the
customer service kind of arena.
Um, has that worked its way
into your, your vertical at all?
Or is that, you know…
And, and are people actively kind
of researching what it would look
like to have, um, you know…
And, and I think, I think your model
is to just kind of have them on the
team, not have them take over everything
because there's been, um, there's this
company, Klarna, that went all in on it,
and then they found the agents weren't
handling a lot of the, the, the human
edge cases that kind of happen very well.
So, so, um, w- what have you seen
on, uh, as far as retailers adopting
agentic for their operations?
Definitely.
It's more for those low-level
repetitive tasks, um, that don't
really need a human to answer it.
So, you know, for service for
retailers, the biggest inquiry
you get is, "Where is my order?"
You know, you can program an agent in Sil-
in Service Cloud to answer that, and you
don't need a person to do that at all.
Um, you know, uh, we've seen people,
you know, build in sort of FAQ,
like their FAQs into service agents.
So any sort of frequently asked
questions, maybe it's about warranty or
your subscription policy or even your
return policy, the agent, the non-human
agent just answers that right away.
So we're seeing a lot of the kind of,
um, l- really straightforward, low-level,
repetitive questions that you get
being automated, uh, by, in that agent.
And then anything that's
more complex, then the agent
hands it over to the person.
And I think what is missing in that
conversation is that when that a- the
actual person gets that inquiry, the agent
summarizes the case, has already pulled up
all of the information on the person, you
know, you know, what orders are, are open,
what's their loyalty status, have they
complained before about this issue, right?
So when the person gets to talk to a
service agent, that service agent has
been briefed and knows what's going
on, and doesn't have to toggle through
a million screens or like search
for, you know, what the history was.
So the tran- the conversation
is much faster, much more, um,
you know, uh, personalized, uh,
much more engaging because that
agent, uh, the, the AI agent…
I always like-
Yeah.
Human and AI.
Yeah.
Has prepped the person, uh, really well.
And so we've seen great, um,
success that this increases
productivity for the service agent.
It makes their job easier because a
lot of that stuff has just already
been brought to their attention before
they engage with the actual shopper.
So the-- it really is kind of working
in tandem, um, when it comes to, to
customer service and in other, in other
skills that we've seen retailers adopt.
It really is about taking out like
kind of these manual, repetitive,
low-value tasks to free up humans to
do the more complex or creative work.
Yeah.
Interesting.
Any other, um, areas that, you know…
I've heard some retailers talk about,
um, leveraging some agents around
maybe like supply chain, and there's
some mundane things that kind of
happen in that, you know, kind of
like keeping an eye on levels and
tweaking some forecasts and things.
Uh, I know Amazon's got a whole,
you know, automated system that
runs most of their supply chain.
Um, and then also I've heard some
of the marketing functions where,
um, you know, there's, there's the
analytics part of marketing can get
to be lots of moving data between CSVs
and things to get to the, the data
you want and freeing up the marketers.
Are you guys seeing anyone use
agents there, uh, on the platform?
Definitely, um, in marketing is
a big area, um, for, for agents.
I mean, you know, you can basically
just have it write a campaign brief for
you, you review it, and then it, uh,
the AI agent will create that customer
segmentation, will decide what the
promotion that would resinate, resonate
the most with that segment, with the cop-
you know, write the copy for the email
and then send it out and then automate--
and then automatically adjust if it's
not hitting the goals that you set it to.
So yeah, marketing is a huge function
where there's going to be a lot of
productivity gains because a lot of the
stuff can be done, uh, by an AI agent now.
And we do see it, we do see
interest in inventory management.
Um- As well as like where,
where do you put your inventory?
At what time?
You know, what, what area of the country?
Is it in store?
Is it in your fulfillment center?
Um, the thing that, uh, we have Sales
Cloud, so we have agents, um, that
are-- one of our other first agents
to market were all about sales, so
having an agent conduct outbound,
um, uh, outreach, uh, for prospects,
but also, uh, training agents.
So we asked retailers if they were
interested in, you know, a training
agent for their own employees,
especially those in the store, and
that rates really highly, right?
There's a lot of turnover in the store.
Um, there's a lot of things that
store associates now have to do.
They're responsible for more
than just checking people out.
So having an agent, you know, kind
of monitor their performance, provide
feedback, or give them instruction very
clearly and easily on how to open or close
a store instead of them like, you know,
flipping through a manual in the back
office, like that's also, um, resonated
with, uh, the retailers in our survey.
Interesting.
Yeah.
I wonder, like even the customer service
example you gave, you could see that
being in a retailing context, you
know, like, so here comes, uh, Mrs.
Smith.
She's, uh, you know, a gold tier loyalty
person and, you know, she previously
purchased some shoes and some X, Y, and Z.
Um, you know, being able to have an agent
present that to someone in the store
right at, you know, very rapidly will…
Is much better than, you know…
I've been in the stores that have
tried this with the tablets and stuff,
and it's, like, super awkward 'cause
they're, like, catching up on, you know.
It doesn't add a lot of value if you're
gonna have to wait there five minutes for
them to poke around on the iPad to figure
out who you are and what you're doing.
They can never find me either.
Yeah.
It's my curse.
It's because you only have one T.
That's why.
That, that, that is…
But Wingo's so unique, you would think
that would help, but it, it, it doesn't.
Yeah.
How about, uh, I don't know if you're
allowed to pontificate, so maybe this
is, like, not a Salesforce point of
view, but, but have you spent much time
thinking about where does this all go?
Like, um, uh, you know, I've heard
some people say we may have a day where
websites don't get traffic, that these
agents will just be out there, you know,
maybe, maybe the website goes from less of
a human readable, maybe it has a double,
maybe it's a machine readable version and
a human readable, because there, there's
gonna be a lot of these agents out there
doing things, researching and finding
products and ultimately buying them.
Um, do you think that's where we go,
or do you think that's overblown?
I think that is where we
go for some people, right?
Um, I think it's a
little bit more nuanced.
Uh, I liken it, and maybe
it's not the best analogy, but
right, um, we still have stores.
Not everything is bought online.
Um, and online, you know,
desktop is still around.
Uh, it didn't all go to mobile.
So just kind of looking at those past
behaviors, I'm hesitant to say that
everything is gonna be done through
one of these agents, that we're
all going to just shop on ChatGPT.
Um, certainly not in the near term.
I don't think, you know, what I've
seen from their latest release
of shopping features, it's not as
thorough of a, of a research maybe
that, uh, some consumers would like.
Uh, I also like to, to remind
people that shopping for many
is like a leisure activity.
It's a fun activity.
You know, I, I go back to those
recommendations and how people
were hesitant to outsource that.
You know, when it comes to beauty or
fashion, it's personal style, right?
I don't think you, you know,
you scrolling multiple websites
is not a pain for some people.
It's fun.
So I'm really hesitant that, um,
that maybe the catalog version
of the website goes away entirely
in the near term given, like, the
user interfaces that I've seen now.
Um, but I do see sort of a future
where you, um- You-- I mean, we kind
of already see it now, I guess, with
Amazon Rufus, is what I'm saying, is
like search gets a lot better, like
the search bar gets a lot better.
You can have a conversation
with the website.
You can have a conversation on the
PDP page too, not just summaries.
So the website gets a lot more
interactive as far as like
how you find product on site.
But then I also think both to help the
shopper enjoy the experience and, um,
sort of learn more about your products,
and then probably also to s- to find,
to be found in these agents, is you're
going to get a more context-rich
website, more video, more blogs, more
information on the product detail page.
And then the third component is just like
a hyper-personalized experience, right?
So when I go to, um, Sephora, uh, when
I go to that website and I'm logged in
with my loyalty program, they're able
in real time to really personalize it.
Like these are the categories that I
shop, these are the brands that I shop.
Um, and so, you know, when I do see
something, it is hyper-personalized to me,
and then I'm able to like talk to it, and
there's gonna be all this amazing content
for me to really sit back and, and enjoy
the online shopping experience instead
of just kinda like, "Oh, I need this.
I'm gonna search for X thing," and
then whatever results pop up, pop up,
whether they're relevant to me or not.
So that's kind of my like
vision of the, of the website.
It stays, there are product catalogs,
heavy on content, really interactive,
hyper-personalized, um Uh, experiences
Interesting.
Maybe the human experience gets
so much better from AI than
people don't wanna use agents.
It, it's probably gonna be like,
you know, if I just need to reorder
some replenishables, an agent will,
will be probably the, you know,
it's not gonna be the fun part
of, of ecommerce to your point.
But if you're, if you're kind
of like looking for, you know,
apparel or something you care a
lot more than, than, you know…
I, I agree, it's not
gonna be all or nothing.
Yeah.
And it, it could be like
a, a companion too, right?
Like you hel- it helps you narrow down
some things if you're really overwhelmed
by all of the choices and places to go.
Um, you know, I, I'm, you know, I listened
to your last Jason and Scot podcast,
and you were using it in the store.
I've used it in the store.
I've asked ChatGPT to give me
feedback on whether or not these
jeans match this top, right?
Like it's a personal shopper, um,
in my, i- in, in the store as well.
So I, I don't think it replaces anything.
It can enhance things, and yeah,
some transactions will probably,
depending on the moment, be
outsourced, uh, to those agents.
But you'll still-- those agents will
need to pull data from somewhere.
Now, some people I've talked to, and
I'd be interested in your take on this,
is like, at the end of the day, you
do risk having your website just being
a fancy product data catalog, right?
So is it just a feed and, and not…
Uh, I'm a little bit more optimistic
that shoppers do wanna have that kind of
engaging experience on your website where
you can control how you're presented and,
and what storytelling you want to do.
Um, but yes, I-- some people
think that maybe you just need
a, a product feed in the future.
I, I kind of, you know, I've been at this
long enough, like you, that, you know, we
all thought certainly ecommerce is such
an awesome experience, it'll go to 90%,
like, like a lot of the Asian countries
have, you know, 90, 98% marketplace.
Even there, they've even skipped
websites and gone to marketplaces.
Um, every region's different and, you
know, we're, we're kind of at the 16,
17%, and I don't really see us-- we're
not like w- the growth of ecommerce
is, is really kind of relatively slow
compare, or in line with, with offline.
You know, it used to be like
3X, and now it's like, you
know, maybe 10% or 20% more.
So we're gonna get into the
20s, but you know, it's gonna
take a long time at this pace.
So yeah, I, I think, I think it's
gonna be a mix and, you know, I
think what retailers should do,
everyone always has this knee-jerk
reaction to block stuff and be like,
"I don't understand it, block it."
But you gotta go where the customer
is, and if the customer wants to use an
agent You know, and they're buying their
sixth pair of athleisure from you and,
and they just know, like I'm a-- I buy
multiples of everything because I don't
like to make decisions about that stuff.
So, so, you know, for me, a
lot of that decision is just
like, "Oh, make my life easier.
I want to buy yet another one of these,
you know, pairs of pants or something."
Um, so you just got to meet the customer
where they are, and if they want to be on
an agent, then yeah, you're just a product
detail page and give the agent what
it needs to make an informed decision.
I think it'll be inter- I wonder if
we're going to have some y- you know, we
always have direct to consumer and all.
I wonder if, I wonder if you'll be able
to say to the agent, "Hey," you know,
the agent-- I think there's going to
be an interesting friction between, it
feels like the agent companies, let's
just use ChatGPT, but I think there's
going to be like eight to 10 of them.
They're going to want the, the
personalization and kind of
loyalty to live up at that level.
But you're going to want, as a retailer,
you're going to want it at your level.
So at some point you may say to the
agent, "If your user would tell me who
they are and if they're in my loyalty
program, I could offer X, Y, and Z."
And I think there's going to
be some really interesting
friction and opportunities in
there that'll be, you know.
And then like the pricing engine,
maybe, maybe there's, maybe you offer
a different price to an agent than
a human at some point or something
because, you know, a, a, a human
Humans don't love dynamic pricing.
It's like a, you know, they-- we all
kind of get it, but like Uber had to
turn off surge pricing because it was
like, had such a negative connotation.
Uh, so, so I think there's gonna
be some really interesting edge
stuff there that's gonna happen in
this world that we're gonna have
to navigate, and it'll be fun to…
Uh, and as technology providers, uh,
folks like, like Salesforce Commerce
Cloud is gonna have to kind of figure
out, you know, how do we help our
customers navigate this ourselves?
So it's gonna be a-- that's, that's
where I think the really interesting
intersections are gonna be.
Yeah.
To that personalization point, that's
one of the things I really study when,
um, all of these, uh, you know, news
releases come out because it always,
all of them, whenever they introduce new
features, always talk about, you know,
co- whether it's Copilot or ChatGPT or
Perplexity, the-- or even Meta AI, right?
Talk about the personalization aspect
because they now all have memory,
so they remember your conversations.
And for ChatGPT, you can type, "Remember
that my grocer is Whole Foods," or,
"Remember that I am gluten-free."
Um, and then it just remembers it.
So, uh, again, and Mark Zuckerberg has
said that he thinks that Meta AI will
sort of dominate the space because not
only does it remember your interactions,
but it pulls in all of your social media
interactions, so it's an even richer
profile, and that people will keep using
the agent that knows the most about you.
Um, and so what I've seen
kind of with our, um…
We just released our personal shopping
agent, and in one of the, the, the
demos I've seen is, um, you can use
it to actually replace a quiz, right?
Why, you know, you don't need a,
a quiz anymore to figure out what,
uh, foundation type you need.
You just need to have a conversation with
the website again about like, okay, these
are my-- this is, you know, the skin.
And it can prompt you, right?
Like, do you have dry skin, you
know, or, or, um, oily skin.
Are you a summer or a fall?
What's your color
palette?
Gen Z women are into using ChatGPT for
that, uploading a photo and figuring out
what their, their color, uh, analysis is.
Um, so I was thinking like, you know,
we see a lot of, um, mobile apps in
particular and retailers, you know,
have like a fill-- tell us about us,
you know, you, because we're multi-
like Sephora, I'll stick with that.
You have a big profile of like skin
tone, hair type, all of these things
that I have to fill out, and so fill
out rates are generally very low.
Um, it can be very varied, but again,
like I was saying earlier, like, you
know, if you log in and the technology
now seems to be good enough where like
it just remembers that I asked the chat,
um, that I'm looking for s- uh, like
sunscreen for, um, with a SPF of 50.
And so it'll always know that that's
what I want when I look at sunscreens.
So I think, yeah, there's a
hyper-personalization component
that becomes like embedded in
the experience, both in those
agents and then in that website.
And you can then get re- do really
cool things and make the shopping
experience more interesting.
I always am like, always find it
funny when I have retailers sending me
like men's clothing things because I
don't shop the men's category at all.
My husband does his own shopping.
Um, you know, so there's a lot still to
be done in basic personalization, but
this technology now can really unlock
just kind of this automatic memory of you.
And then instead of, you know, filling
out a quiz or filling out a profile,
you can just chat with the s- uh, the
shopping assistant, and it'll remember
everything that you already told it, and
you can then personalize it both, you
know, on, on the website, but then link
that data to the marketing profile, right?
And so now you can send me emails that
are only women's items and not send
me, um, men's items because I don't,
I've never purchased that from you.
Yeah.
And I think they're way more likely
to have a chat about that stuff,
um, than fill out a quiz that's kind
of like a never-ending, you know.
You get pulled in these quizzes,
you're like, "Oh man, this took
like, you know, a whole hour."
Um, so the, the, the…
Just for clarity, the, the
shopping assistant you're talking
about is part of Commerce Club.
Retailers can put that on their site
and then it, it-- I'm sure the retailer
can program the types of things that, in
conversations that it has and whatnot.
Yep.
It is generally
available.
Totally customizable.
Yeah.
Generally available and totally
customizable across product
recommendations, service inquiries.
Yeah.
Neat.
I'll have to, uh, play with that.
Um, well, uh, we're getting close on
time and you've been very generous
with your time and I appreciate it.
Uh, I did see you guys are having a, um,
a show coming up, so tell us about that.
Yes.
Every year in Chicago where I'm based,
uh, we have Salesforce Connections.
This year it's June 11th and 12th,
the ideal time to be in Chicago.
Um, thankfully it's not in February.
Um- That's, so the, the week
you guys have summer, right?
Yeah.
Yeah.
Yeah.
Just those two days.
Um, and it's at McCormick Place, and
that's where we talk about all of
our innovation, all of our products,
but, uh, particularly for, um,
marketers and commerce professionals.
So we have a lot of content
geared to how to leverage
Marketing Cloud, Commerce Cloud.
Um, there will be, every session
will touch upon agents, so you'll be
able to get a really good informed
perspective on what is, uh, an
agent and what it can do for you,
uh, both in marketing and commerce.
So if we wanna see the agenda, we
just, uh, Google Salesforce Connections
2025, and we can find that out?
Yeah.
And I'll be there, so you can-
Yeah … reach out to me and, uh-
Awesome … and come meet me in person.
Yeah.
And, um, are you speaking on
anything, or are you just generally-
I am … part of the crowd?
Um, I will be on stage to talk
about service, um, and what
agents can do in, in service.
And then I have a few round tables
around, um, personalization and loyalty.
I've never been to that.
Is it, is it a Dreamforce giant thing?
Like, how many people…
Or is it more of an intimate kind
of a hanging out with the, the, a
small subset of the Salesforce crowd?
Um, it's a smaller subset than Dreamforce.
So, um, if I had to guess around
attendance, yeah, 5,000 people maybe.
Maybe up to 10,000.
Yeah.
It's at McCormick, um, and so
it's, yeah, it's a lot of fun.
Um, and there's lots of, you know, dinners
and happy hours that go along with it.
Uh, but yeah, it's a great chance
to reconnect with people in the
ecosystem, both at the retailers and
consumer goods companies, but all
of our wonderful, um, partners, our
solutions implementers, our ISVs.
So it is like a commerce marketing
reunion every year, whereas like
Dreamforce is for every vertical.
A lot of developers come and admins come,
and it's, it's, it's very overwhelming.
Awesome.
Uh, is it one where you guys
drop news, or does that get
kind of reserved for Dreamforce?
Or do you guys like announce typically
the next year's platform or anything
like that, that people can…
Yeah.
Yeah.
Interesting.
So- We'll make announcements
around marketing
and e-commerce … could be
some news coming out of this.
Yeah.
Could be.
Yeah.
Very cool.
Yeah.
Hopefully more agents.
Oh, yes.
Cool.
Thanks, Michelle.
You're welcome.
You know, next time you get any breaking
cool news, let us know and we'll have
you back on the pod, 'cause we love
to hear data about what's going on
out there as we try to all read these
tea leaves of our agentic future and
the intersection of agents and retail.
Definitely.
Thank you so much for having me.
It was a pleasure, and I look forward to
seeing- Absolutely.
Yeah.
It's good to be back in the game and, uh,
I'll see you at all the usual trade shows.
And, um, we'll see.
I think we're gonna do a little
summer trip during Salesforce
connections, but maybe next year.
Okay, great.
Would love to see you there.
Follow, rate, and share to stay up
to date with RetailGenic, your guide
to the future of AI shopping agents.