Travel Buddy with Switchfly

In this episode of Travel Buddy, Rachel Satow and Ian Andersen unpack the rise of conversational loyalty and its impact on customer experience. The conversation explores how AI-powered chatbots are reshaping loyalty programs, making them more intuitive and engaging, while raising important questions about data, privacy, and the need for seamless integration across business channels. With real-world examples, strong perspectives, and practical guidance on technology and user experience, this episode offers business leaders, marketers, and technologists timely insights on where travel loyalty is headed next. Tune in to discover how genuine conversation and smart technology can transform the future of customer relationships.

Chapters
(00:00) Introduction to Conversational Loyalty
(01:00) Transforming Traditional Loyalty Programs
(06:00) AI Implementation and Practical Considerations
(10:00) Unified Customer Experience Across Channels
(13:00) Data Integration Challenges and Personalization
(19:00) Privacy, Convenience, and Regulatory Debates
(28:00) Multichannel Communication Preferences
(31:00) Human Touch and Natural Conversation Flow

Connect with Switchfly
Website: https://www.switchfly.com/
LinkedIn: https://www.linkedin.com/company/switchfly/
X: https://twitter.com/switchfly
YouTube: https://www.youtube.com/@SwitchflyOfficial

Creators and Guests

Host
Ian Andersen
Demand Gen & Marketing Operations Manager
Host
Rachel Satow
Senior Marketing Strategist

What is Travel Buddy with Switchfly?

See more at Switchfly.com

Welcome to Travel Buddy,
presented by Switchfly.

In this podcast, we talk about all
things travel, rewards, and loyalty.

Let's get to it.

Brandon Giella: today we're talking
about conversational loyalty.

Yes, it's about AI, but there's a lot
of practical use cases here to think

about the customer journey from, say,
a chatbot into a customer loyalty bot

and the conversions between all of these
different departments and how they use AI.

I think what would be really helpful,
Rachel, I'll start with you, is talk

to me, what is conversational loyalty?

I know this is a growing trend.

A lot of people are talking about this.

There's a lot of use
cases out there, I think.

But define this for us, and
then we'll go from there

Rachel Satow: Yeah.

So conversational loyalty is essentially
the next phase of what we see today

in terms of someone coming to your
site, your program, your app, etc.,

having a conversation with a chatbot.

When you think about it from a
loyalty perspective, traditional

programs often feel like a catalog.

Someone is needing to, browse
through, really dig into some of the

redemption options that they have.

The idea of conversational loyalty
essentially turns that into a

conversation, one, but a two-way street.

So someone who is looking to redeem, say,
60,000 points can come to that chatbot,

come to that tool within your program
and say, "I would like to redeem 60,000

points for a beach vacation in March."

And the conversation agent will
allow them to surface ideas and

options in a more streamlined manner.

So … And then ask more about it.

So for example, if … Let's use
the example I just, just used.

A beach vacation in March.

I want to redeem 60,000 points.

What can I get for it?

It may be able to surface that I could fly
on specific, airlines, go through specific

hotels, have this room versus that room,
and help me make decisions faster while

also inquiring more details, all without
having to necessarily leave and go to a

different browser to do that research.

So the idea of conversational loyalty
is that you, you essentially have

a companion to help you in the
decision-making process so that you

can move through the funnel faster.

Brandon Giella: I love that.

I love the aspiration.

I'm feeling a headache growing in
legal teams and engineering teams

everywhere, which is totally fine.

This is obviously the direction
in which we are headed with AI.

We can clearly see that case happening.

so it's kind of like a get ready for this.

This is for sure happening.

And Claude, everybody's familiar
with using AI tools today.

We're chatting with these
things all the time.

a lot of hallucinations in there.

You gotta be really careful, especially
if you're in certain industries.

Totally get that.

And yet, you can see this
could be super helpful.

We've talked about in the past using AI
to do like trip planning and itineraries,

potentially booking tickets for you
and car rentals and so on, of course.

you guys have a lot of great
technology at Switchfly.

but there's a lot of technical
things to, uh, get in place.

So before we get there and talk about
more, more of the practical side of

things, talk to us a little bit about how
you see the traditional loyalty landscape

and how people are using booking redeeming
points now, and then kinda shifting on

where we're going and how to get there.

Rachel Satow: Yeah, I mean, even if you
have the strongest rewards catalog in

the world, they can feel less valuable
if there's friction points in that

process, in the redemption experience.

So when your program becomes difficult to
use from a, in today's, you know, quote

unquote traditional program structure,
if those reward, rewards and redemptions

are difficult to use, that's when you
start to feel some loyalty fatigue.

Like, people are getting
inundated with offers from a

bajillion different channels.

You're essentially pushing all of
these different things to someone,

and it is that one-way street.

So this is the natural next step
in being able to reduce some of

that friction within your program
and help move past, you know, like

confusing redemption charts or hidden
restrictions that people don't know.

It's allowing you, it will be eventually
a two-way street to a- allow further

conversation, deeper conversation.

But as it exists and can exist today, it
helps to surface answers a lot faster.

So it can talk through some of
those program rules that might

not be readily available unless
you're digging through it.

It may help with comparing options.

and all of that is something that is
totally available today and to tap into,

to your point, with, LLMs as they exist.

but yeah, I think from a member
experience, they will be able to

find more value in your program if
your ex- your overall process is just

easier to navigate because you're
implementing tools that may be trained

on your own documentation, etc.,

but you're implementing tools
that help expedite things for them

Brandon Giella: That's right.

That's right.

I'll, I'll turn this next question to you.

what's the worst chatbot
experience you've ever had?

No, I'm just kidding.

no,

so no, I'm kinda joking there,
because obviously like some of these

tools, it's like it sounds great in
theory, and then you start working

with it, and you're like, "This is--
Just let me talk to a human," you

Ian Andersen: Yeah

Brandon Giella: and you see a lot
of like AI assistants in like call

centers, for example, and being able to
surface, you know, it's like reading the

transcript and surface relevant articles.

I actually had a wonderful chatbot
experience a couple of weeks ago.

It was very, very well-trained on a
knowledge base, and it actually did

answer my question really well, and
I had kind of a particular question.

but I wanna talk to you, Ian, a little
bit about like, tell us like some of

the, you know, guardrails or risks or
ways to like, yes, we can understand

how this could be more personalized
and help the decision-making, but

there's definitely some things to
think about, of course, from an

engineering perspective, but also from
like a policy or a program perspective

that might help build some of these.

Ian Andersen: when we're thinking
about AI models in general, to

include chatbots, clearly there are
still mistakes being made, right?

And still, gaps to where it really does--
There is a point in which a human on

at least one end of that interaction
needs to be, involved to, to make some

sort of determination, or at least, you
know, spot if something's off or not.

I think we're seeing it get
better and better all the time.

I know in training AI agents,
you can be, pretty draconian in

instructions given about, you know,
not leaping to conclusions, not,

filling in gaps with extraneous info.

I think as these companies are building,
building their, their AI models and

training them, especially when it
comes to things like handling money,

there's going to be some very severe,
restrictions and limitations put on it.

That said, the, the industry
is advancing rapidly.

there's, there's-- In virtually every
situation you can, you can foresee

a

way in which, AI will at least, if
not replace, at least expediate a

lot of the process, and this is,
is clearly one of those areas.

I think when we're talking loyalty,
it's, it's understandable we

get bogged down a little in like
the, redemption, conversation.

But I think even before that,
what we're seeing now is companies

building, buying and purchasing
sort of assistants if you will.

you know, you can go on the Amazon
app and say-- and, and talk to Alexa

on there and say, "Hey, you know, I
kinda want this thing and, but I'm not

really sure," whatever, and, and the AI
will help lead you down the direction

of to looking what you're buying.

I think the loyalty part of that's
just a natural extension, right?

You're, helping a user, whether they're a
member or not, come to a buying, decision,

helping them make the purchase, right?

And then, on top of that, if you can
add in that loyalty aspect of like,

"Hey, if you become a member and you
wanna buy this again, you get you

get discounts, you get whatever."

Or if they are a member, you know, during
that buying process, if I'm looking for

shoes and, you know, they can-- the, the
AI can kinda say, "Hey- You can pay cash

for this, or there's these other shoes
or this brand you're not even aware of

that we have a agreement with, you know,
that you get double the points for or

what, you know, whatever the case may be.

just, there's, there's infinite room
for expansion in that sort of area.

So it becomes not-- you're not having
your, your customer support AI chatbot,

and your buyer support AI chatbot,
and your loyalty support AI chatbot.

It really is a whole organization that,
that really takes the, the prosp- you

know, the person from prospect through
customer and then repeat buyer, right?

Through, engagement and loyalty.

And it, it, obviously the word
conversational, we tend to

think, know, immediate direct
conversation when you're interacting.

But I mean, on top of that, there's so
many, additional marketing and growth

opportunities that, that tie in, right?

If, if the AI's looking at your,
your, chat history and your purchasing

history, they're gonna be able to
recommend certain things for you.

They're gonna be able to email you or send
you notifications or, you know, whatever,

even after the conversation that, that
brings you back into, into it, you know?

as Rachel mentioned, you know,
if you're-- if, if I say I want a

beach vacation, you know, in this
timeframe, what can you do for me?

maybe it can't do anything good
right now and, but a month later

it says, "Hey, because of X, Y, and
Z, these hotels are now on sale.

You wanna come back and revisit it?"

You know?

So I think it's better to look
at conversation not as just that

immediate chatbot interaction, but
that ongoing total conversation

you're having with a buyer, right?

Brandon Giella: The, yeah, gosh, I, I
have, I have, I'm gonna go on a diatribe.

Rachel, you go ahead.

Yeah, yeah, you go

Rachel Satow: Yeah.

I mean, as I have two,
I have two things there.

I think the first that I want to address
is that there is such an opportunity

to streamline channels, to Ian's point.

Like, yes, all of the information that
those conversations will be giving

to you and that the reference points
of your data that is training the AI

can pr- surface to the user, that is

That's amazing.

The biggest thing that I see in the
future is that all of those different

aspects, the, you know, the buyer's
agent, the customer service agent,

the, the, loyalty agent, they'll all
eventually converge into one just simple,

straightforward agent that becomes your
conversation companion, someone that you

can ask all of those different things to.

And while Ian's absolutely right
in that all that information you're

getting from that can be dispersed
across your marketing channels, can

go into your platform to be able
to serve better options if they're,

you know, looking manually, etc.,

the other aspect of it is you'll
be able to have a conversation

more proactive conversations.

So if you think about it from somebody
booking a trip, yes, eventually I think

that conversation can be, "I want to
redeem for a beach vacation in March,"

surface later like, "Hey, if you wait a
month, things may be a little bit cheaper.

Let's revisit back then."

And then they can come in and say, "Hey,
like remember we talked about this?"

So you're able to, to kind of circle back.

But then it … once they actually
book, you're able to have an ongoing

conversation leading up to that trip,
throughout that trip, after that trip,

all in one place, rather than it coming
from, like if somebody's booking with

one agent now, then your emails are
eventually going to drip to them come

time for them to actually travel.

You'll have a, a net new, a, a net
new channel to be able to have those

conversations to prepare your travelers.

The caveat to all of that, going back to
your question, Brandon, about risk I think

that programs today will find a challenge
with their data being complete and

connected enough to be able to get there.

You know, conversational loyalty and a-
an AI bot, the way we're discussing it

aspirationally, needs the right context
and the right training, and it needs to

be holistic in order to get to that point.

So many lo- mo- loyalty programs are
going to run into the challenge the,

of fragmented data across a CRM and
their booking platform, and the loyalty

platform that they're using, the customer
service tools that they're using.

They're all…

They all carry their own specific
data, and all of that data is in those

platforms, and it's hard to, for loyalty
programs to find a way to get all

of that data in one source of truth.

and if the conversation agent only sees
one aspect of the member relationship

because that's the only part of
the data that it has access to, the

recommendations are going to feel
incomplete or not personalized enough.

And when you think about creating loyalty,
accuracy is going to matter more than

a casual AI engagement or use the way
we're doing it to Acclaim Trips now.

Brandon Giella: Mm-hmm.

Rachel Satow: Them being able to serve,
the most accurate piece because they

have complete profiles on individual
members is going to be more important

in cultivating loyalty than just having
the ability to have a conversation

Ian Andersen: Yeah, I don't, I don't know.

I, I don't know if I
wholeheartedly agree with that.

The-- I think directionally you're
right, for, th-there's-- every

part of the organization has their
own data silo and hoard, right?

And, and yes, the, the out-of-the-box,
like, solution to turn on the

chatbot isn't necessarily gonna work
across the board right out the gate.

Which why-- is why I think it sort of
has to grow internally, from a single

starting point to start incorporating
the, the other areas, right?

As, as it gets better and better
and more, more and more trained,

more and more knowledgeable.

I don't think we're gonna come
out tomorrow and say, like,

"Here's your loyalty chatbot.

it in, you know, integrate with your
CRM and off to the races," right?

It has to be that of
growing holistic thing.

And I think…

Brandon, to your point earlier, you ha-
you, you mentioned having a really great

conversation with AI chatbot, earlier.

I've had, I think, some relatively
decent ones recently, at

least like ultimately helpful.

I've had a few, you know, recently
that probably weren't as helpful, but

a few that you could almost not know
you're talking to an AI chatbot, right?

If you t- sort of take it out of it.

and I think people are willing to forgive
a little bit of and, and headache if

they're seeing the ultimate reward, right?

If, if you're showing we're trying
to develop this thing, it may not be

perfect, it may not get you to the
one hundred percent end state, but

it'll get you eighty-five, ninety
percent of the way there, and then

we'll give you a link or whatever to,
to get the extra five, ten percent.

I think people might be willing
to give it a little grace.

I mean, we, we all know l- this
conversation two years ago, we'd be

some theoretical day in the future
AI might be able to do a couple

of these things, and now we're
talking about it being here, right?

It's just it being hard to, to implement.

But it is here, and, I think people
understand with how rapid the pace

of technology is changing in this,
in this, that yes, we're gonna have

some growing pains, but I know a
little bit of investment now is gonna

really pay off for me, you know, six
months, eight months down the road.

Rachel Satow: I would, I, I would argue
back, that I think the three of us and

many of our listen- listeners probably
have a slight bias because we are on

the more tech savvy, cutting edge…

Like, we're, we're working with
this, this type of tool all the

time, and the typical user might not.

The, the end user might not.

So I think we would be more forgiving in
those instances than, say, m- my parents

or grandparents or anything like that.

Ian Andersen: Yeah.

Brandon Giella: I, I actually did just
read a report, that, that if the, if the

chatbot apologized for getting something
wrong, that it was actually a better…

Like, folks rated it a better
experience when it was like,

Rachel Satow: My chatbot
is constantly apologizing.

It's always apologizing to me.

"You're right.

I'm sorry."

Brandon Giella: a, yeah.

And not attorneys.

So I, like

Rachel Satow: not attorney

Ian Andersen: but like we are getting
more and more tech-savvy as a society

as a whole just by sheer need, right?

Like it, you know, how many people…

When I joined the military, basically
it did 21 years ago, we don't

need to worry about exact numbers.

the-- I remember having guys, guy,
who checked their email, right?

Who literally wouldn't turn their computer
on, and the only reason it was in their

office is 'cause they had to, they, they
were forced to have it in their office.

now those people cannot
survive in an organization,

in virtually any organization,
whatever your industry, right?

Without being able to check an email.

AI, I think it's just gonna become
so much more ubiquitous that sort

of tech illiteracy isn't gonna
be as sharp of a ramp-up as maybe

some other forms in the past.

Brandon Giella: I, and
I think it is the…

It, it is a classic debate between,
of course, privacy and convenience.

That's been a debate in tech
for 30 years, probably longer.

And there's also a debate now between a
chat interface the proper UI to create

a good experience for, like, a customer?

Maybe there's other ways to do it.

We just haven't figured out exactly
how to use AI with, in, in some other

kind of interfaces, unless it just
builds things for you, like you're

building a website or whatever.

And so, but, but what I love about
that, the, some of these trade-offs,

it's convenience, personalization,
privacy, you know, ease of, of using

this thing, is that it flattens all
of these different, that people have

had into, like, a unifying layer.

Like what you're talking about, Rachel,
and, and Ian, both of you, where it's sort

of, collapsing a lot of different data
silos or departments within a company.

Because everybody hates, like, you
know, I call a, a credit card or,

you know, a travel firm or whatever,
and it's like, "Oh, well, that's…

We're on the booking department.

You have to call the rescheduling
department over there, and this is…

Oh, you're talking about membership?

Oh, you gotta…

I'll transfer you over to this department.

Oh, this is about your credit
card-specific billing details?

Well, let me call these folks over
here about your, you know, details."

And it's like, gosh, I've spent two
hours on the phone with something that a

chatbot could've figured out in literally
one thread But at the same time, I think

it's also not only just the different
departments, but like the journey, like

what Rachel, what you're talking about,
like, a-and, and Ian, you mentioned

like prospect to customer and beyond.

All of these debates, conversations, the
technicalities, the different departments,

sales and marketing, customer support,
whatever, all of that is collapsing.

I find it like a fascinating that
I be willing to trade everything

Claude knows about me, including
maybe some conversations I've had

with my therapist, and I'm like
really digging into something, and

I'm, I was just like talking to
it, talking to it late at night.

Do-- will I trade all of that to get way
more personalized information into offers,

loyalty, my membership status, getting
more points, more savings, helping me plan

a trip, helping me book a car, knowing
things about my family and things I should

consider that I hadn't even thought of?

Would I be willing to trade all that
and interact in a chat interface

to grow from, you know, a middling,
know, once a year user to like,

I'm talking to this thing week?

'Cause I'm thinking about so
many different things that

a company can do for me.

You know, I'm thinking of like a
larger, you know, travel company.

There's lots of things that I'm-- I
gotta plan and think about and do,

and, could it help me with that?

Like it's a, it's like a fascinating…

then thinking like in two
years, this technology is gonna

be literally ten X better.

Like every four months, there is a step
change in the model and how efficient

it is and how much context it can…

And, and how, you know, cheap it
gets, and the compute obviously

needs to keep up, and the electrical
grid needs to keep up, and so on.

But if the trajectory remains the same,
at least in the midterm, these things

are gonna be crazy good at booking and
me as a customer on all kinds of things.

And that's fascinating, and I'm
willing to trade a little bit

of my privacy maybe for that

Ian Andersen: Yeah, I think it's
gonna not be an option, right?

Like,

Brandon Giella: Yeah, right.

Yeah

Ian Andersen: if you're wanting to buy in
the market, you know, using your credit

card or whatever, at some point some of
the stuff is gonna be non, non-negotiable

Rachel Satow: I would agree with that.

And then to play devil's advocate, I think
we will see iterations of things like

GDPR come through, and say, actually,
AI needs to provide the user the ability

to control what brands are able to see
and utilize within their own platforms.

Because at the … Like, Brandon, to
your standpoint, like I'd be willing

to, to trade certain information.

Like, how much of that
information am I willing to trade?

I think we, we will find … And
again, not a lawyer, not

an attorney, not in policy.

But like, I think we will find
that it becomes so prevalent, that

question becomes so prevalent of, are
people willing to trade their data

for more personalized experiences,
and if so, what data and how much

of it are they willing to trade?

It, that question I think is going to
surface at an entirely separate level

than just conversational loyalty,
and how loyalty programs can utilize

that information they're collecting.

Because at the end of the
day, personalization is

supposed to feel helpful.

It is supposed to feel like we are
providing you with something that

we think you will like based on
your past behavior, and the other

indicators that you have provided
to us, and given us consent to use.

Not like the program is Big Brother,
and watching every move, and per- using

every single piece of detail that I could
have on you in order to provide you with

the most unique experience possible.

I think from a policy standpoint, we
will likely see some sort of, legislature

come through requiring the ability for
people to dictate how much information

is being taken, what kind of information.

If it's collecting information
that you'd say you don't want

it to use, what it does with it.

I think obviously that is, that
is well in the future, because

it's not quite yet there.

But I do think that that's going to
come into fruition here, for sure.

Brandon Giella: Yeah.

Rachel Satow: yeah

Brandon Giella: I was listening to,
Lenny's podcast recently, and, the guest

was talking about how she uses, several
OpenClaw agents to manage her family.

And so she had like her own Mac Mini
servers in her, in her home kinda that

were disconnected from her personal
life and all that kind of stuff.

But, but that they would
help manage some things.

And she was saying, you know, what a
great kind of opening UX could be is

instead of filling out a form, the chatbot
would just like introduce itself to you.

Like, "Hi, I'm Claire," you know.

"Tell me about yourself."

And then you give the information you want
to give to the chatbot to help it plan

for you, and you only give it so much.

Whatever you tell it,
it can use, you know.

That could be like a-- I, I
imagine a fairly simple policy.

Again, I'm not an expert here.

But I imagine

Ian Andersen: I think I'm

Brandon Giella: here's
some stuff about me."

Like I imagine like five questions
could probably get a good

agent tuned into what you need

Ian Andersen: Sure.

I, I am less optimistic than
you two, I think, about,

Brandon Giella: come on, Ian

Ian Andersen: regulation being introduced.

I mean, been, it was 2003
that CAN-SPAM Act came out.

It's been

Brandon Giella: don't know.

I'm not that old.

Sorry

Ian Andersen: the FCC,
Privacy Protection Act, right?

Like I th- Especially in seeing recent
developments with, with some of these AI

companies going public and just having
unimaginable levels of resources, they

are not going to want to be regulated.

Brandon Giella: No

Ian Andersen: definitely not
to any, particular degree.

So I'm, I'm for now gonna operate
under the assumption that it's gonna

go on essentially like the wild, right?

And, to the individual company's
policies, and sorry, I don't

necessarily trust that they're gonna
be super, user-forward when it comes

to, comes to data and information,

Brandon Giella: Yeah, of course.

Yeah.

Yeah, of course

Ian Andersen: so I, there is a bit
of fatalism in that, that I think

we've kind of crossed the Rubicon
on a lot of this stuff, and th- it's

just a fact of life you have to live
accept to some degree that your data

is out there and accessible, and it,
you know, a-accessible by a lot of

different organizations you never
necessarily gave it specific access to.

so why not make it work
for you a little bit and

Rachel Satow: Yeah.

Brandon Giella: Either,

Rachel Satow: Brandon

Brandon Giella: entirely or get off of it.

It's kinda

Rachel Satow: yeah

Brandon Giella: yeah.

Ian Andersen: Yeah

Rachel Satow: Brandon, you, you made
a comment before we went down the

data rabbit hole about whether or
not chat and chatbots are the channel

that this should be utilized for.

And I think in, in my opinion,
the answer is yes, it's the

only one that truly makes sense.

if you look at some research, you know,
two-thirds of Americans spend more

than four hours a day on their phone,
and around 90% of consumers have said

they want to use texts to communicate
with businesses, but when they send

a text, they never get a response.

So we're using our pocket computers
constantly, and that, it…

That's just, that's not
gonna change, in my opinion.

But, like, your s- your marketing and your
service, like, it has to evolve to remain

effective, and that includes recognizing
when there is a shift in the dynamic of

preferred channels and communication,
both from a tech standpoint of how

you're going to implement conversation,
conversational loyalty, conversation bots,

and just, like, a generational standpoint.

I mean, we, we all have family
members who just, like, don't

want to use their cell phone.

They prefer to have a landline.

Like, that still absolutely exists.

But on the other hand, you have a
generation who is perpetually online.

Brandon Giella: And

Rachel Satow: And I…

Brandon Giella: to be, to counter

Rachel Satow: And exactly

Brandon Giella: There's like folks
that are 25 and under, they're

probably not on email, but they
will text you or Instagram DM you.

Are you

Rachel Satow: Yeah.

Ian Andersen: I, I

Rachel Satow: so many businesses
that have s- sorry, Ian.

Many businesses that have started
implementing some of this technology are

using a WhatsApp, a Messenger, something
that is, is inherently baked in the way

that we as humans are communicating with
each other, and I think that, I think

that's the only avenue for it, honestly

Brandon Giella: Yeah

Ian Andersen: I think the only problem
with that, i- is so is the chatbot the

right, channel or, or platform for this?

And the answer is yes.

And is phone and voice?

Yes.

And is email?

Yes.

Like, it, it can't be one thing, right?

And with AI, with the capabilities
that it's developing, it

doesn't have to be, right?

You can have your AI chatbot, be, be
voice if somebody wants to actually

dial a phone number and talk.

It can be a chatbot, text message, email.

It can be multiple avenues, and I
think it has to be all of them, right?

If, if you wanna be successful

Rachel Satow: Yeah, I
would, I would agree.

The, the biggest tip for anybody
who is starting to develop this that

I would say is just ensure there's
a way out of the AI conversation.

Ensure that there is a way for
the user to actually get ahold

of a person on the other end.

Because, 'cause, I mean, let's,
talk about this theoretically.

If you're, if you implement something
and it doesn't have all the data, from

all of the different silos within your
company, and someone is truly trying

to utilize this to get a very specific
answer and you're, you're just not

quite there yet, the friction of someone
needing to get to an actual human to

solve that answer is going to completely
eradicate any benefit you're getting from,

from this conversation from an AI bot.

And in reality, there's about, like,
there's over, I think it's, like, over

70% of people still prefer to communicate
and get the help of an actual human rather

than solely interacting with the chatbot.

It's not that they don't want to use
a chatbot, it's that going back to how

it exists today versus the aspirational
side of things, as it exists today, most

chat doesn't have the documentation,
doesn't have the mo- the most robust

training on all aspects that someone
may be trying to utilize it for.

And so that's just my very long-winded
way of saying, if you are going to

experiment with this technology and
implement it, please make sure there

is an easy way out for users to funnel
to an actual human conversation instead

of solely relying on a may or may not
well-documented, well do- well-documented

database from an AI, bot AI

Brandon Giella: I could
not agree more strongly.

Like I, i-it's, it's hard to separate
like what chatbots are today my

frustrating experience with so many
of them so often where it gives me a

list of five options because that's
what the, the team has defined as my

only options to respond to this thing.

And it's like none of those, none
of those address my question.

I just…

Rachel Satow: Where's the other?

Brandon Giella: God, I just
wanna throw my computer through

a screen or through a window.

so I totally agree with that.

But I think like, and I, and I don't know
if this has ever been true in, in, in

history, but the pace at the advancement
of this is like, I mean, if you saw Google

I/O three weeks ago, or if you see Meta or
Apple's WWDC a couple of weeks ago, like

if you see where things are going, like
much now, like these things are happening.

They're, they're at the very, very early
stages, but like talking to somebody

in a chat, somebo-somebody, I use that
anthropological term very loosely, but

you, you are talking to a, an entity
that can sound just like a human

being and, and be as helpful as that.

That technology is like a year from
now, months from now, that is happening.

So, I mean, even ChatGPT and Claude,
their voice bot sounds incredible.

So it's here, but you know,
on an enterprise scale,

maybe a year or two from now.

So all that to say, like I totally
agree with you and it's, it like

you have to prepare for that now.

So how do you set the data?

How do you set the policies?

How do you set everything you need
to, to like collapse that conversation

to go from prospect to very loyal
customer and let somebody talk through

all the stages of their journey?

I think it's a tremendous,
tremendous advice.

Yeah.

Ian Andersen: Absolutely

Brandon Giella: Well, I guess
that's the end of the episode.

I don't know.

I don't have anything else.

That's great.

No, I just totally agree.

I love where your, where your head's
at, and I, I, I couldn't agree more.

I love it.

Yeah.

Rachel Satow: I think the only other
piece that I would say is if you are

implementing this or you're, you're
working with it, focus on natural

conversation flows, rather than just,
like repeat prompting because, I

mean, you go on LinkedIn and there's
someone ranting about all of the

ways that they can tell that your
content is written by AI, and it's

it like it … I would say just
be conscientious of, like how

natural conversations flow.

There are a lot of tangents.

This podcast is a great example of
all of the tangents that can happen

within a natural conversation.

Brandon Giella: not.

We stick to one topic.

Ian Andersen: Yeah

Rachel Satow: all the time, yeah.

We're very straightforward.

but like just be, be cognizant
that when you're developing

something, it should not be

so

What else can I do for you today?

W-

Brandon Giella: yeah

Rachel Satow: would you like this?

Would you like that?

You're right, I'm sorry.

It should not be so polished that
the conversation feels inauthentic.

it should understand when the
conversation naturally ends and take

that opportunity to kind of like close
out and then revisit with all of the

data it gathered and go from there

Brandon Giella: Yes.

And while I like niceties, it shouldn't
be sycophantic to say like, "Oh,

that is a great suggestion, Rachel.

suggestion."

You know?

Like

Ian Andersen: and that's just
one of those things that it's,

that's gonna develop naturally,

Brandon Giella: Of course, yeah.

Ian Andersen: gonna,

Brandon Giella: Yeah.

But no, it's, it's definitely
… It's, it's really thinking

about the, the experience.

Like, at what point is AI helpful
and an incredible benefit to the user

experience, the customer journey?

And at what point is it like,
"You should talk to a person.

Let me get you in touch with so and so."

You know?

yeah, I think that's a really great mark.

Well, guys, as always,
thank you for this topic.

I love the topic.

I love the direction, the way y'all are
thinking about it, and I think it's, it's

so, so important to be thinking about now
because this is dramatically changing.

I mean, just two days ago we were
seeing these product launches on

some stuff that I was like, "This is
mind-blowing what is happening right now."

Like, mind-blowing.

I've been doing this for 12, 13 years,
and it's like, gosh, that's so cool.

so yeah, it's just happening so fast.

And, the time is now.

thank you.

Rachel Satow: is now.

Brandon Giella: And if you

Rachel Satow: Thanks, Bryn

Brandon Giella: your chatbot or
your agent on wonderful resources,

please go to switchfly.com.

They have amazing resources to talk
to you all about travel and loyalty.

So as always, thank you, and
we'll see you on the next episode

Rachel Satow: Thanks, Brandon