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).
Why does agentic browsing matter?
So this is a really interesting new phase
in what I call agentic shopping, and it's
moved from the server to the browser,
and I'll explain to you what that means.
I'm gonna give you a little timeline
of where we are in the history of
this, a little, uh, overview of
what it's like to install Comet, and
then we're gonna go into the live
demo and show you some features.
And then we're gonna have a
special guest appearance later on.
Welcome to the Retail agentic Podcast,
where we explore the intersection of
retail, e-commerce, and AI agents,
which we call retail agentic.
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 Retail agentic, your
podcast partner in navigating
our AI shopping agent future.
And now here's your host, Scot Wingo.
Good afternoon, everybody, or good
morning, and welcome to our live stream.
Today, we are going to go over
Perplexity's Comet, and we're
gonna give you a live demo of
all the cool things Comet can do.
Uh, I am your host, Scot Wingo,
the author of the Retail Gentic
Podcast, as well as the Substack.
Excited to have you today.
Let's jump in
So today we're gonna talk about
what is, uh, Retailgentic, uh, who
am I, and why are we doing this?
And then why does agentic browsing matter?
So this is a really interesting new phase
in what I call agentic shopping, and it's
moved from the server to the browser,
and I'll explain to you what that means.
I'm gonna give you a little timeline
of where we are in the history of
this, a little, uh, overview of
what it's like to install Comet, and
then we're gonna go into the live
demo and show you some features.
And then we're gonna have a
special guest appearance later on.
Retailgentic is a podcast we started
to really explore the intersection
of retail and agentic shopping,
thus the name Retailgentic.
You can find the substack
at retailgentic.com,
and then on there, if you look up
in the top nav, which I'm showing
over here in the right-hand side,
you can go navigate to our podcast.
It's available both on YouTube
as well as all audio players.
If you want to skip right to that, you can
go to YouTube and search for Retailgentic,
or in your favorite podcast player, just
search for Retailgentic and we'll be there
So one of the context things, let me
get rid of this so you can see the
slides better, is that, uh, we've
been trained, uh, as an industry
in the retail world around Google
and SEO and SEM and these things.
And in the Google context, the user is
typically giving you three words, uh,
or less, and that's their search term.
So that's what we have to anchor on.
On the back end, Google uses
PageRank to measure the popularity
and the authority of a page, how
many inbound links does it have?
And then they're increasingly
focused on this EEAT framework, which
stands for experience, expertise,
authoritativeness, and trustworthiness.
LLMs are different.
It seems like they're a little similar,
but they're actually quite different.
They crave context.
So when most people prompt an LLM compared
to, to a Google search term, which is
at three words, you're getting way more
context, and the engines love that.
Um, they're increasingly having larger
context windows with, uh, Google Gemini
at the largest with a million token, which
is, you know, a very large context window.
And the average prompt that people
give the engines today is 24 words.
This is increasing over time.
As we learn as users of these GenAI
systems that they want more context, we
are giving them more and more context.
They also want content.
So, um, a lot of these things, and I'll
show you how this works, they, they have
a database, so they've been trained,
which is their initial content, uh, on,
you know, all available information.
At this point, it's pretty well known that
all the internet has been trained on, uh,
LLMs have trained on the whole internet.
So, so all the available knowledge in
the world they have largely trained on.
Um, but they all want more content,
and one of the ways they do that
is they store it in another place
that's aside from the training called
retrieval-augmented generation.
Think of that as a database that's
got more current information in it.
And then finally, they want authority.
And this is different than PageRank,
where it's a popularity contest.
What they're doing is they're looking
for-- Because they have so much, uh,
training data, they can-- and they
understand natural language so well.
In the retail context, if you
have a product and someone on
Reddit has mentioned your product,
they're able to tie those together.
Google's not able to because it's just
a, you know, a-- It's not doing any
natural language, traditional Google.
Um, now Gemini, uh, does all that.
So when I say Google, I'm
talking traditional Google.
They, um, they don't-- they can't
pick up that authority that way,
um, the, the same way the engines
can, the GenAI engines can.
So that's a new twist.
Um, we were working with one, uh,
retailer, and they weren't understanding
why their product was showing up.
And what we found is someone had
written a Reddit view, kind of a
scathing negative review, and that
had really diminished their authority.
Um, and they didn't even know
that existed until we kind of went
exploring what had happened there.
Um, so here's kind of a, a
example of what that looks like.
So here on the left, uh, you
know, we've given-- we've kind
of asked Google for some…
Oops, uh, from information about
some pickleball paddles, and
that's kind of the context it has.
Um, whereas you can tell when most
people are using the GenAI engines,
they tend to give very verbose prompts.
So, so that is, uh, super helpful to the
engines 'cause they love that context.
We're gonna be spending a lot of
time talking about Perplexity today.
Um, we're not affiliated with Perplexity,
just kind of a fan, and they, they're
a little bit out on the bleeding edge.
So, so that's why we've covered
them on two live streams.
They're coming out with some
interesting stuff, uh, and then
there's some quick followers out there.
Um, uh, Aravind Srinivas is the founder.
Uh, he was at OpenAI and
left to form Perplexity.
Uh, I recommend you follow
him on social media.
He's kind of a fun follow.
He's put very spicy comments and,
uh, always poking the Google bear.
They've raised over one point four
billion dollars at a fourteen billion
dollar valuation, so that's what
we'd call a decacorn in startup
lands, meaning a 10X over unicorn.
Um, they're growing very, very rapidly.
You can see their revenue, uh,
trends over here on the right that
show this kind of tremendous revenue
growth that they're anticipating.
And they were very early out of the
gate with this idea of how do you take
GenAI LMS and marry it with search?
Um, and then also this is one of the
first-- of the big guys, the big five that
we'll talk about, um, in the GenAI world.
They're the first to
come out with a browser.
Their investors include Jeff Bezos, uh,
who we know from the e-commerce world,
Tobi, uh, from Shopify, and then NVIDIA,
which is where all the GPUs come from.
So many of you that are readers of Retail
GenTech, you've probably seen this diagram
before, where we talk about the five
levels of agentic shopping autonomy.
And one of the reasons we were excited
about Comet is we believed that it would
move Perplexity, which is currently
at level three with Buy with Pro,
over to level four, Advanced Buy, and
we're gonna check in on that today.
Um, also, Google Gemini has announced
they're moving to level three based
on Google IO, Meta, et cetera.
Now, one of the surprising things
that did happen recently is ChatGPT,
uh, did come out and say, uh…
Ooh, that's interesting.
They did come out and say, um, that
they are coming out of the browser, so
maybe they will jump into level four.
We'll kind of have to wait
to see what that looks like.
So here's-- There, there's been
a number of agentic browsers.
The first is one called Brave,
and theirs is called Leo.
And, uh, that came out in late 2024,
so it was one of the first ones out.
And then Opera, the, the browser company,
came out with Neon in May this year.
Then in June, there's a company called The
Browser Company, and they came out with
a bro- uh, a product called Dia, D-I-A.
Um, and one of their taglines
was, uh, "Shop with your tabs,"
and, uh, that was interesting.
And then here we are, Comet, which was
announced last week, so we're less than a
week into this, uh, and then, uh, dropped.
It's been announced for a while.
And then ChatGPT on that same
day announced that they're coming
out with their own browser.
Uh, we don't know much about that.
They're-- It's gonna be derived off
the Chromium open source browser
package, which is what Chrome is
based off of, as well as Comet, and
now we're gonna have a third one.
So it's gonna be interesting to see
if they can match the capabilities
of Comet I'm gonna show you today,
or if they take a different route.
So it's important to note
what's different about this
The-- So over on the left,
we have a variety of…
And, and I'm just gonna give the context
of e-commerce because that's what
our audience is most interested in.
But this is generally true for any
information, um, that's in the engines.
So let's start at the left and
work our way over to the user.
So in the left, in that line,
the black line's important.
That's the dividing line
between server space.
So on the left of this slide, it--
we're in kind of server land, and
over on the right, we're in the
individual user's device land.
This will come up later as we
start to expand on this concept.
So there's all these e-commerce sites out
there, um, you know, certainly tens of
thousands, if not hundreds of thousands.
And what happens is the, uh, they are
trained on some data, but they don't
have-- The training runs take so long
that most of the engines are twelve months
out from the last time they trained.
So to give you, they couldn't really do
a good web search, um, experience for
you or tell you about products because
all their information's a year old.
So what they do is they create a
retrieval-augmented generation database,
which is a web index, just like Google
has, but theirs is specialized for AI.
So they send crawlers out, and those
crawlers go through e-commerce sites,
gather information, and put them into
a web index that then is searched
by, um, the, the sys- the engines.
And here's the big five.
So we've got ChatGPT at the top.
It's hosted on Azure for the most part.
Uh, Perplexity is mostly on AWS.
Anthropic is the third one.
It's on AWS, which is Amazon's, um…
We used to call these cloud, um,
servers, now they call them hyperscalers.
Uh, so o- I've kind of matched
them with the hyperscaler because
we're out in the server land.
And then Copilot is obviously owned
by Microsoft, and they own Azure.
Uh, and then the, the rogue one here
is Elon Musk xAI, uh, also known
as Grok, and they have their own.
They just went and built
their whole own everything.
So, so Elon believes in vertical
integration, and he's kinda like,
"I don't need a hyperscaler.
I'll just go build my own."
Um, and this Colossus thing is the
biggest cluster of, uh, next-generation
NVIDIA chips ever put together.
It's pretty, pretty crazy
if you wanna research that.
And then over here we have our user.
Um, in this example, he is on a
device that is a laptop, uh, or
a-- looks like a desktop, actually.
Could be a laptop, could be a phone.
Um, and the context that, that whatever
system is being used here is they're
gonna have the user prompt They-- Most
of these systems now have memory, so they
can kind of remember the things you've
asked, and if you specifically tell them,
they will remember things about you.
Like I, uh, once memory came out, I
made sure it understood things like my
shoe size, my preferences of, uh, shirt
size, pant size, all that kind of thing.
So as I shop, I don't have to enter all
that, my work address and my home address.
You can actually…
This is a fun experiment.
You can actually go to any-- your
favorite gen AI system that you use and
say, "Tell me what you know about me."
And, uh, it's always surprising
to see what it's gonna tell you.
Um, it also, because it has this
web index and retrieval augmented
generation, it basically has access
to that, so that's a lot of context.
It has the whole training
set of data as context.
And then some of these have agents that
go out independently of the crawlers.
Crawlers are kind of lazily
going through the web over time.
And then if you need immediate
new information, a lot of them…
Like let's say you want to double-check a
price, they'll send an agent in the form
of a different crawler to go specifically
look for a piece of information and pull
that off the website and pull it in.
Let's zoom in on Perplexity.
And, and this, this diagram that I'm gonna
show you would be the same for all these.
Um, so this, this is easily
exchangeable with OpenAI.
I just wanna make sure you understand,
um, this so that you understand
why the agentic is different.
So now same kind of setup,
but I've kind of simplified it
that we're on Perplexity only.
So y-you see their actual
bot is called Perplexity Bot.
Oops.
And, uh, it's here in the
middle, uh, up at the top.
It actually sends an agent
frequently to go look for
things, uh, as you're doing them.
As you ask Perplexity some questions,
it may wanna go get the most fresh
information, and it will actually
send a bot to do that, and that's
called Perplexity User Agent,
and that's its technical name.
Um, if you have a website, you can see
this coming in, and it identifies itself.
Perplexity Bot is its web-- uh,
it's kind of lazy-- more lazy web
crawler to build the web index.
Now, we've added a whole
thing at the bottom.
So as, um, Perplexity built Buy
with Pro, their agentic shopping
option, they needed another database.
They need a product catalog.
And this is-- I actually think
of it as an offer catalog.
It's a product catalog, um, that takes
all the offers it finds, canonicalizes
them, meaning if it finds 13 GoPros,
um, or 100 GoPro 13s on offer out
across the internet, it's gonna put
them all into one listing and then
tell you where all the offers are.
We call that canonicalization
in the industry.
So what it does is it now, um, as
it's crawling, it is, uh, using
the we-web index is where it's
getting a lot of its information.
It augments that with the agent, and then
also many of these, uh, companies are
offering up a merchant center, and this
is very similar to the old comparison
shopping engine days, or if you've
ever sold on a marketplace like eBay,
Amazon, Walmart, Target marketplaces,
um, they all have a s- a thing like
Seller Central where you interact with
the system to get your inventory in.
Um, now, those also have orders out.
Right now, you put inventory in,
and then the agent comes and does
the order, so it's a little bit
different, but, but very similar
front-end merchant center experience.
Um, now those merchant centers,
they can take product catalog,
uh, information or offers directly
from you through a variety of ways.
The most…
kinda like the Rosetta Stone of
e-commerce is what's called a data feed.
And, um, Google has the most
popular format in their Google
Shopping, so many people just
use the Google Shopping format.
Um, I had a company,
Channelyzer, that did this.
There's Feedonomics, there's,
there's now probably 30 or 40,
maybe even 50, 100 companies that
can help you manage your feeds.
That's one way to get the data in there.
That, that's pretty old school, and
it lacks some of the context that the
engines are gonna want, so many of them
would prefer to get the data through a
different mechanism that's more robust.
Um, it's like what we call a flat file,
file, and a lot of the context, like
variations and, and things of this
nature, are more like three-dimensional.
It's hard to represent
those in a flat file.
Um, there's a protocol called MCP
that some of them can use or an API.
So let's say you're sitting
on Shopify as your platform.
Uh, Perplexity's already integrated
with Shopify to some degree, both on
the checkout, I presume on a catalog
capability, so they can actually go
through a back-end server to get to
your catalog without you having to get
involved, and that's a feature that your
platform provider may or may not provide
Okay, so that's, that's kind of
where we were prior to July 9th.
Now take that, and I've shrunk
it down into a little box here.
So that's there.
And the blue box you see is all
the server stuff that's happening.
But now we've added
something really interesting.
So now our user's there
and they're using Comet.
So what happens is because you're,
you're, you're elevating the experience
between your local, your thin client
HTML and the internet on the server
side, now we actually have a bunch
more stuff going on on the user side,
so meaning actually on your device.
Um, so I'm gonna be running all this on
a, a pretty needy MacBook because I'm
gonna put it through its paces here.
But all this should work on a phone.
They haven't released the
browser for the phone.
They've said that one's
coming, an iOS and an Android.
Um, PC is already available, so
on Windows, Comet is available on
Windows and Mac as a, a native app.
Um, and then this will be coming
to your, your mobile devices.
So you're sitting there running Comet,
and now because it is your browser,
think of all that you do in your browser.
I'm normally a Chrome user.
Many people use Safari, Firefox, uh, or
Microsoft Edge, uh, Opera, Brave, whatever
you're doing, the browser knows a lot
about what you are doing and have done.
If you have a bunch of tabs
open, it knows all that.
And- Because of the popularity of
autofill, uh, it tends to know a lot of
your payment information already, your
login credentials, and that kind of thing.
So this is great for the LLM
because remember, the one thing
they want the most is context.
So we went from giving them the context
when we were kind of a thin client
working with lots of me-- lots of, um,
power on the cycles on the server side.
Now we've opened up the client side and
they can see, um, whatever you've talked
to the local agent about, your email, your
calendar, and all this is permissioned.
So, so I have made the personal
decision to give it all this
context because I get…
What I get back is time.
Um, and I think a lot of consumers
will make the same decision, and,
and you're about to see some of
the crazy stuff that this can do.
Um, my purchase history.
Now, now we're talking, um, there's a
whole part of this where it's made me
like, it's increased my productivity
probably 15% just on mundane things.
Um, we're not gonna talk-- I'm
gonna show you one example of that,
but the rest is all e-commerce.
So on the e-commerce side, it can
see my purchase history, my travel
history, all my open tabs, my browsing
history, my autofill, and much more.
That gives it the context to just
seamlessly do things without me
having to tell it what to do.
And it also has a very…
So, so it has yet another…
So there's still that Perplexity
user agent that Comet can call on for
server-side things, but it can-- now it
lives in your browser, it can open up
sites and do things right locally, um,
which is much, much faster than pinging
through all the servers back, you know.
So you going to their server, their
server goes to a website and back again.
There's less hops because we're just
going right to the, the information.
So the, the installation
process of Comet is, was A+.
So a lot of these things that
are kind of super early design
programs, I, I usually kind of have
a problem in the install process.
This is the, by far the best
install of a browser I've ever had.
So in step one, you can see it
had this very pleasant-- It has…
Turn your velt-- sound, sound
on when you download this.
Um, it has this very pleasant music,
and it's got like a c- a planet
theme because of the word comet.
Um, then what's nice is it actually looks,
I have four different, uh, I forget what
they call them in Chrome, uh, profiles.
So, so I have different identities.
I have my personal one.
I also work with this tweener
fund, uh, and then my refi
buy, um, and yet another one.
And, um, I'm constantly switching
between those, you know, those
different personas, if you will,
or those different, um, profiles.
So the first thing it did,
it said, "Hey, which of these
profiles do you wanna bring over?"
And then as I did that, it said
for each one, and y- this is
number three, it, it very, it gave
me the, the ability to opt out.
But it said, "Look, if you give me
your browsing history, your cookies,
your credit card information, all
the stuff that's in Chrome," it
pulled it out so that everything
was set up just like I have Chrome.
My…
And, and it was very seamless.
It was as if I was still in
Chrome, which was amazing.
Then you can do some customization.
Um, you get to…
Right now, you can't
update your own avatar.
You have to pick from this,
let's see, what is it?
The, the different planets here.
I was in kind of a Saturn mood that day
So that was the setup.
Now let's jump into the main
attraction of why you're here.
Okay, hopefully you can all see this.
So I've got four, uh, Comet browser
windows open, and I'm gonna do this
because some of the things I'm gonna ask
Comet to do are gonna take quite a while,
and I don't want you to have to sit there
twiddling your thumbs while I do it.
So we're gonna go into kind of a
hyper, a hyper, uh, multitasking mode.
So, um, hopefully that doesn't
make you, uh, kind of lose track
of where we are on the whole thing.
Okay, here we go.
All right, so over here I have Amazon.
I'm logged in, you can
see it recognizes me.
So I'm just gonna, um…
The way it works is you have
this new thing over here.
This is the Comet assistant, and
it's in the upper right-hand corner.
When I open it, it actually gives me
some prompts that says, "Hey, I see
you're looking at Amazon," context,
"Here's some things I can do on Amazon."
Well, for purposes of this, so I
don't have to slow down the demo by
typing, I have pre-put in a prompt.
Uh, "I previously ordered bully
sticks for my dog from Amazon.
Find my past order and find out if that
is still a good deal by comparing…
'cause I ordered these during Prime
Day, I don't wanna just reorder
'cause I know the price has gone up.
Um, uh, see if that's a good deal,
and I want you to compare the
current Amazon price, not what I
paid, but the current one, to what's
available from other retailers.
And I wanna make sure it understands
to ship to my work address."
So it's gonna get going on that.
Now, if you've ever used one of these
kind of more thinking models, um,
uh, they all have different names
like deep research and, and whatnot.
Um, so you can see here, it's
telling me what it's doing.
So it, it's actually op- and I can, I can
zoom in on this if I want, so I can see
what it's doing in this little window.
So it's going in there, and it's
gonna start cooking, as we say in AI.
So it's over here.
It's found the bully sticks.
While it's doing that, I'm gonna
go over to Target over here and I'm
gonna give it something to work on.
And, uh, I'm gonna actually type this one.
So keep an eye on the
Amazon while it's going.
And so, so the other thing is this is
hyper multitasking, uh, on steroids.
Um, so right now it's looking up the
current price of the bully stick.
You see it opening up a window doing that.
And then over here I'm gonna
say, "My daughter is going
to college in thirty days.
Uh, add to, to my cart the items
she will need as a freshman."
Uh, so this is a true story.
So now I've given it a-- So you've
seen me give it a task which is
like a reorder, but on steroids.
Now I'm also gonna give it another
task, which is going to be in this
model I like of research, find, buy.
I'm giving it a hyper research phase.
It's gonna have to go find all the
stuff I'm gonna-- my daughter's gonna
need for college, find it at Target,
and then add it to the cart, and
then we're gonna see how it does.
Back over on Amazon, it has
found this, uh, this item, and
we can see that it is $21.99
now.
I got it for $15 on Prime Day.
That was a good buy.
I should have ordered more because
obviously I need to reorder.
Uh, we have a, a dog, office dog, and she,
uh, Elsie, she loves the bully sticks.
Okay.
So now it's over checking Walmart.
Over here on Target, it has gone and
done a bunch of research to find out
what do people need for college, and
now it is going and looking for, um…
Looks like it's looking for bedding
While that's going on,
let's do a shoe exercise.
So here I just have it in my email,
and I know that I ordered these pair…
I would show you, but they're on my feet.
I have this pairs of shoes I really like.
They're the Cloud shoes.
Um, I can't remember when I ordered
them, and I don't wanna go look
through my email for the order conf.
So I'm telling it, "Look, I bought
some shoes, and I wanna reorder them.
Go figure it out."
And it's gonna go do that, and it's gonna
go-- It's gonna have to look in my email
'cause I actually remember, shh, don't
tell it, I ordered them from Backcountry,
but we're not gonna tell it that.
And, uh, I don't have the microphone.
It's not listening to the microphone.
You can actually do this all
verbally, which is pretty wild.
Uh, I didn't wanna blow
your minds too bad on that.
Uh, so we're gonna go do that.
So it's gonna go find it.
Okay, it is somewhere, and it's trying to
figure out shipping on the Amazon task.
Where is
it?
It's on Walmart.
Okay, over here--
So sometimes you'll see, you
see how it does this John Doe?
To, to do what it does to figure out
shop, um, shipping is frequently it'll
just put a fake address in there, but
get the zip code right, and then it…
that's how it calculates shipping.
So it's going through and
finding shipping cost on Target
for things in the upper right.
Over here, let's see how it's doing.
I bought some Cloudrock shoes.
Okay.
Based on your request, here's
everything you need to know about them.
Um, variants, things, thing,
who they're for, pricing,
durability, best places to buy.
Okay, it found them on Backcountry.
I want to reorder the same color
You can find it in my email.
I ordered eleven point five,
and I don't remember color name.
Look specifically for the
shoes I ordered previously.
So I didn't give it enough context
It's gonna go figure that.
Okay, over here on the Amazon side,
purchasing the Billy sticks from
Walmart at the best available price
for delivery to your work address.
Okay, so it has found that they
are available for $15 on Walmart.
So Walmart must have price matched
Amazon, who's currently at $21.99
that you can see in the upper left
screen, and it's found them on Walmart.
Target is still running, so you can
see it is adding things to cart.
Like s- looking-- looks like it's
putting a little storage crate in there.
So it's got eight things in
the cart if you look at Target.
Here on the right, it is now
going into the order tracker.
It found the order number
Uh-oh.
There's, uh, no 11 and a half
available, so it's gonna, it's,
it's gonna determine all that
Um, this is telling me to open Walmart
It's put it in the cart, and it's
telling me to finish the checkout.
I'm not gonna do that, um,
so we'll see how it goes.
Oh, it's actually below the $35
order, so I may actually buy two
or three of these to get the $15
price, 'cause that's pretty good.
I'm, um … On, on another idea,
I have Walmart Plus, so I, I did
not log in under that ID, but I
could have done that to save money.
So you, you see there it basically
found the $15 bully sticks for me,
and if I'd just gone and reordered
from Amazon, I would have lost $7.
So it saved me money in that example.
So that one's done.
Let's close that down.
Um, over here.
So I'm telling it, "Look,
I'm not gonna sacrifice size.
I really need that size."
I'll make this bigger for you guys.
Okay, so over here, it's added
eight things to my cart: a shipping
crate, a surge protector, a
task lamp … Oh, that's nice.
A laundry basket.
Gotta have one of those.
I doubt she's gonna be
doing a ton of laundry.
Flip-flops.
Nice.
And a shower caddy.
A … Ooh, a bath towel
set, antimicrobial.
And a dorm kit.
Now I'm gonna throw it a curveball.
I'm gonna give it more context.
She is going to the University of South
Carolina and staying in Williams Quad.
Will all this work there?
The best currently in stock
price, the on thing is, oh, good.
No, this is Zach Malone calling.
Let's see what we got here
I don't even know what end clothing
is, so it's found it for me.
So I'm saying take me over there.
Yes
Okay
All right, so it found it at this…
These are actually out of stock,
these old-- These are some o-o-ones
that are out of stock right now.
So it actually found them.
That's funny, it didn't do that last time.
All right, so now it's gone over
here in my target window, and it
has checked to see how we're doing,
and it's checking the bed size.
Your bedding kit is correct.
Desk drawer, plenty of room.
Your stacking bin and
items are gonna work.
Your lamp is permitted.
Additional tips.
Um, let's say, can I bring
a, uh, camp kerosene heater?
Well, I think we all know what
it's gonna say on that one,
but I'm doing it on purpose.
Okay, so we're done with my shoe exercise.
It helped me find shoes, pulled it
out of my email, and verified the
best price and place to get those
in my color and size preference.
Oh, I can't bring a kerosene heater
to University of South Carolina.
Okay.
So then I could go check out on this if I
wanted to, or I could say, let's do this.
I'm kind of going off the fly here.
Let's pick the most expensive item.
This bed set is pretty
expensive, but, but I like it.
Can you find it cheaper somewhere else?
Okay?
So it's gonna go think about that.
While it's doing it, I wanna show you,
this is the one example I wanted to
show you of where, um, you know, it,
it actually is super productivity.
So I get a lot of emails like this
that are saying, uh, I do a lot of
scheduling and calls with people.
And here's one internally where,
uh, Becca is saying, "Hey,
here's some dates you can meet."
So I'm just gonna pop this open.
Remember, it has the
context of my calendar.
I'm gonna say, "Check these dates."
Sometimes it-- I'm using Superhuman,
and it glitches out a little bit
on doing things, so we'll see
what it comes back with there.
I'm gonna hide that 'cause it's
got some calendar stuff on it.
You're just gonna have to…
We'll see how it comes out at the end.
All right.
Uh, over here, I searched for a
better price on this, and it's
still listed at Target, which
appears to be standard price.
Other retailers have other things, so
it, it, because I said I liked this
one, it went for a very specific thing
w- and it's only available at Target.
Okay, so let's shut that down
And then over here
And yeah.
Okay, so over here it has said it
has found that this is the best slot
for me, so I'm gonna send the email
And then I'm gonna
schedule the appointment.
Do you see that?
That just saved me probably 20 minutes,
and I do that a lot during the day.
Okay, so that gives you an idea of
all the cool stuff that Comet can do.
So our special guest is Alexa.
But this isn't just any Alexa.
Alexa, say hello to the live stream
audience and explain who you are.
Hey there, live stream folks.
I'm Alexa, your friendly
neighborhood AI assistant.
I've been given a makeover this year.
Think of me as Alexa 2.0
with more personality and
a better sense of humor
Alexa, tell us a little bit about
what you can do in Alexa Plus
I've got so much to show you.
Whenever you ask me, what can you do?
I'll help you discover and
learn about Alexa Plus features.
I can also share ideas to help you
get more done or stay connected.
Which one would you like to try?
Alexa, stop
Okay, so
Let's see.
Let me get this on here.
There we go.
So let me just show you this real quick.
So Alexa is our surprise guest, and
one of the things that's interesting
is it can connect to many of your
accounts that do various services.
Um, so you've got the…
All the shopping stuff is on the Amazon
side, and that's what I'm gonna show you.
But you can also do reservations,
order an Uber, get Uber Eats, um,
Thumbtack for tasks, ask it Wolfram
questions around math and that kind
of stuff, um, order event tickets.
It's really a good experience
with Ticketmaster, um, and, uh,
GrubHub and all your food order
and restaurant kind of stuff.
So, so that's pretty powerful.
I-I've, I, I'm gonna try to h-
keep this to 45 minutes, so I may
show you some of that at the end.
All right.
But let's go back to that.
All right, I'm gonna show
you a shopping experience.
Oh, look, she's already showing you
some of the things that she can do.
I know it's a little hard to read.
Here, let me do this.
I've gotta do it at an angle
or else it reflects too much.
That's probably about as
good as we're gonna get.
Alexa, I need to buy a toaster.
Can you show me some good options?
Working on it
Now what Alexa, what, what
the plus version brings
I found several toaster options for you.
The top result- Is you can interrupt
is the Amazon Basics two
slice
toaster.
Alexa,
I'm sorry.
I was actually looking for a toaster oven.
I, I need to do some pizza and
chicken in this, and I don't think
those are gonna work for that.
Got it.
Let's find a toaster oven
So as I was saying, you can interrupt it
now and it's much more of like a chatty
type interface I understand you're
looking for a toaster oven that
can handle pizza and chicken.
The top result is the Black + Decker
crisp and bake air fry countertop
convection toaster oven.
It has a medium
capacity A- Alexa, I already
tried that Black & Decker and
it, uh, it broke in a month.
I, I really want, uh, the
Breville or what's it called?
Braver?
The one that starts with a B
Breville got it let me check
I found several Breville
toaster ovens for you.
The top result is the Breville Smart
Oven Air Fryer Pro and Convection
Oven in brushed stainless steel.
What are the dimensions of that…
Alexa, what are the
dimensions of that mini one?
Let me check that for you
So as you can see, I've actually…
That is my toaster, so I know
a good bit about this one.
The exterior dimensions of the
Breville Mini Smart Oven are 15.5
inches in width, 14
inches in depth, and 8.75
inches in height
What are the reviews,
uh, for this toaster?
Checking the reviews now
According to customer reviews, the
Breville Mini Smart Countertop Oven
is highly praised for its compact
size and versatile functionality.
Users appreciate its ability to toast-
Alexa, what do the negative
reviewers have to say about it?
Let's check the negative reviews.
While customers generally appreciate
its compact size and versatile
functionality, some note that it has a
smaller interior capacity than expected.
A few reviewers also mention that the
crumb tray can be difficult to clean.
Alexa, w- could I cook a
whole chicken in this thing?
Let me check that for you
The answer's no
Based on the product information,
it does not seem suitable
for cooking a whole chicken.
The compact size and 0.45
cubic foot
cook capacity- Alexa, does Breville
have one that can hold a whole
chicken and, um, is a good price?
Let me check that for you
I found a Breville option
that might work for you.
The Breville Smart Oven Air Fryer
Pro and Convection Oven is large
enough to fit a whole chicken.
Okay.
It's a
versatile
option
with air-
Alexa, can it also do pizza?
Let me see
Yes, the Breville Smart Oven
Air Fryer Pro can cook pizza.
It comes with a 13-inch pizza pan and
has a dedicated pizza function that
allows you to bake delicious pizzas.
Alexa, what do the negative
reviewers say about this one?
Checking the negative reviews
While most customers are satisfied,
some report issues with uneven
cooking and a complex interface.
A few mention that the
oven is difficult to Alexa,
I see this has a four point two
A 4 point- I think that rating-
Alexa, I see this has a 4.5.
Are there any toaster
ovens that have a 4.7?
This is where I usually get off the
rails Let me check that for you
It has a rating of 4.3
stars from more than 11,000 customers.
To purchase this one, say, "Buy it now."
It, uh, it really likes
this Black + Decker.
It must be the Amazon Choice, um, one.
Um, Alexa, go back to that Breville
Are you referring to our chat
about toaster ovens for pizza and
chicken, or the one about finding
a highly rated model under $200?
The one about pizza and chicken.
So it's got memory of my pa- I,
ob- obviously this isn't the first
time I've had this conversation.
Oops.
Alexa, the one about pizza and chicken
We discussed toaster ovens
for pizza and chicken.
The Black+Decker Crisp'N Bake
can fit a nine inches pizza
and has air fry technology.
Some users find- Alexa, go
back to the Breville
We were discussing toaster
ovens for pizza and chicken.
The Black + Decker Crisp'N Bake
can fit a nine inches pizza and has
air fry technology, but some users
find it hard to clean frequently.
Alexa.
Are you-
Stop.
Alexa, are there any interesting
concerts in my area this weekend?
There are some exciting
concerts happening this weekend.
Big Time Rush is performing
at Coastal Credit Union Music
Park on Friday at seven PM.
Also on Friday, Casey Rocket has two
shows at Good Night's Comedy Club, one
at seven thirty PM and another at ten PM.
Alexa, are there any tickets
available for the Big Time Rush?
Sure.
Let me look that up
Great news.
There are tickets available for Big Time
Rush's In Real Life worldwide concert this
Friday, July 18th at 7:00 PM in Raleigh.
The event is on sale, but there
are only a few tickets left.
Would you like me to check
ticket prices for you?
Yes.
I'm particularly interested
in the front row.
I'm a huge fan
Certainly.
Let me check for front
row tickets for you.
As I understand, you're a
huge fan of Big Time Rush.
I found some great tickets, but
unfortunately, front row seats
aren't available right now.
The closest seats I can find are in
section two, row J, which are reserved
seats in the pavilion, right of center
facing the stage under the pavilion roof.
These tickets are one hundred and twenty-
Alexa, stop
All right, so hopefully that gives
you a little idea of, you know, we're,
we're in this really interesting next
gen world where we've now got a next
gen Alexa, and we've got browsers
that can go shopping for us and
add nine, ten things to our cart.
Uh, things are changing very fast in the
world of agentic shopping, and we are
your resource for staying on top of it.
Thanks for joining us today.
Uh, we will go through and, uh,
due to brevity, uh, we didn't get
a chance to go through all the Q&A.
I saw a lot of cool questions coming in,
and want to go ahead and wrap it up here
so we don't burn up too much of your day.
But we'll be answering those
questions, uh, in a post or over
on the platform where you were.
Thanks for joining us today.
Have a great rest of your day.
Follow, rate, and share to stay up
to date with Retailgenic, your guide
to the future of AI shopping agents.