May 1, 2025 • 29 mins
What if AI-powered technology could not only solve your customers' problems faster but also anticipate their needs—and why do so many organizations still get it wrong?
On this episode of the Delighted Customers podcast, we unlock the real reasons why companies struggle to implement AI and automation tools effectively, even as consumer expectations for personalization and seamless experiences reach new heights. Greg Kihlstrom, martech and CX expert, demystifies the hype around chatbots and agentic AI, showing us where most companies stumble and what it really takes to reverse-engineer remarkable customer journeys. If you’re wondering why some AI initiatives delight while others fall flat (and what you can do about it), this episode pulls back the curtain.
You should tune in to hear from Greg Kihlstrom, a thought leader at the intersection of marketing, technology, and customer experience, whose podcast network and consulting insights are helping brands evolve—whether you're leading a Fortune 500 organization or an ambitious mid-market player. His deep dive into what’s really working at marketing and CX conferences—and what’s next for both large and small companies—makes this discussion especially valuable for anyone navigating the fast-changing landscape of customer engagement.
Here are three burning questions Greg answers on the show:
-
What are the real barriers that have kept companies from leveraging AI (like chatbots and agents) effectively in the customer journey?
-
How can AI empower customer service teams to pinpoint root causes and improve experiences—without losing the uniquely human touch?
-
For leaders at mid-sized companies, what are practical steps for selecting technology partners and overcoming siloed data challenges as they implement customer-centric AI?
Don’t miss this eye-opening conversation with Greg Kihlstrom! Listen and subscribe to the Delighted Customers podcast on Apple Podcasts and Spotify, or find us on your favorite podcast platform.
Meet Greg Kihlstrom
Greg Kihlstrom is a best-selling author, speaker, entrepreneur, and host of The Agile Brand Podcast. With 15+ years of experience at the crossroads of marketing, technology, and CX, Greg is the former founder and CEO of a successful digital agency in Washington, D.C., now working as an independent consultant to leading brands and organizations. He is the founder of The Agile Brand Podcast Network, a prolific writer on the evolving roles of martech and organizational change, and a key influencer in bridging the gap between digital transformation and meaningful customer outcomes. Greg is also recognized for his thought leadership on AI innovation, process improvement, and the future of customer experience.
Connect with Greg Kihlstrom on LinkedIn: Greg Kihlstrom LinkedIn
Show Notes & References
-
Subscribe to Greg’s newsletter: "AI Isn’t a Goal or a Strategy" on LinkedIn
-
Learn more about The Agile Brand Podcast Network: https://www.theagilebrand.com/
-
Catch up with the Delighted Customers podcast on your favorite platform for more episodes!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Well, I couldn't be more excited to have on the Delighted Customers
podcast this episode, Greg Kilstrom. Greg
and I go way back, actually over 15 years before
I got involved in CX at all. And he has somehow come
back into my life. And we crossed paths in Baltimore at one point
at a CXPA event that happened down there. And we
(00:22):
have developed a business partnership in that I and part
of his umbrella of podcast, a part of the Agile Brand
Network. Greg is the host of the Agile Brand
Podcast Network. Greg, welcome back to the Delighted
Customers podcast. Yeah, thanks so much. Yeah, excited to be here.
And I'm thrilled at the success of your show as well. It's been a
(00:44):
great partnership. Well, Greg, can you share? And I agree 100%,
and thank you so much for helping the work that you and your team
do to help lift up our show here. Can you share a little bit more
about what you do with our audience? Yeah, of course. Yeah. So I come originally
from a marketing background, so I owned. I say I'm a
recovering agency owner. So I started
(01:06):
sold an agency, ran it for about 14 years here in the. In the Washington,
D.C. area where I'm still at always throughout
that, you know, I was very passionate about the kind of the
intersection of customers and the technology that
serves those customers. And I found that, you know, after I sold the
agency, now work as an independent consultant for a number of
(01:28):
organizations. I also host the podcast. I've written a few books. I always find
myself kind of going back to whether it's in marketing, whether it's in CX
or other areas, just how can technology better serve not
only consumers and the end customers, but also the employees
and the consumers of that technology within the
corporation. So, yeah, looking forward to talking with you. Well, I consider
(01:50):
you a martech expert, Greg, and I know
you attend a lot of conferences in marketing and CX and other areas.
And I know recently you've been traveling. You still have some left to
do here, but you've been to places where there's a lot
of mind sharing going on around what's new and
cutting edge. And what I'd like to do is pick your brain a
(02:13):
little bit, if it's okay. Yeah, yeah. And share with our audience some of the
newer things, some of the fresher things that are out there. In particular, if we
can think through the customer life cycle and
different applications of AI and technology
that you've seen out there, that could help really nurture and
deepen relationships that brands have with their customer
(02:34):
base, whether it's when they're trying to get them to make that
initial purchase or they've already come on board and now we're servicing
them, the service operation model, even
helping the organization measure the success of
their customer base or how they're delivering the experience. Does that make
sense? Yeah, yeah, definitely. Happy to, happy to talk through. Yeah. I mean, I can
(02:57):
maybe, maybe just start at the beginning of the journey. And I think
there's, there's certainly, there's been tons of talk about AI. There's
tons of information misinformation and tons of stats that people
kind of hang on to, I think. But I think what's,
what's been really interesting over the last several months is
there's like, I think there's certain anecdotes that people kind of glom
(03:19):
onto as far as, like, consumers hate
chatbots and so therefore we should never use that. So
just to use an early example in the, in the journey, like, it's not
that consumers hate any form of AI based
interaction or something early in the journey, it's that they really don't
like dumb chat bots that aren't helpful, just like they don't like
(03:41):
dumb phone trees that send them in the endless doom loop. So
getting past some of the early preconceptions of consumers don't
like some of this stuff versus what don't they like about it? Consumers
love personalized content offers, deals, you name
it, interactions, and just that a, that a brand understands them.
And what was always, you know, the traditional way of
(04:04):
understanding a brand, you know, if you run a mom and pop store, it's like,
you know, the people that walk in every day by name and you know what
their kids are doing and blah, blah, blah. But AI and personalization
allows you to do that at scale. And so consumers really like that. What they
don't like is being shown irrelevant offers that are, you know,
quote unquote, personalized, but really just based on
(04:25):
some, you know, some irrelevant information. And so I think
it's because it's continually evolving. There's a lot of people
that are kind of thinking, I would say the thinking two years ago versus
the thinking today, because these tools have gotten a lot better at
understanding and anticipating, you know, and that's just talking about early
on in the journey. Yeah.
(04:47):
So yeah, Greg, what do you think stood in
the way of companies using AI in
the form of chatbots or something similar
effectively in the past and what is helping to
change that now? Because, you know, one of the things that
comes to mind for me is technology. In
(05:09):
search of a strategy never really works. And I believe, and by the nature
of this name of this show, delighted customers podcast, we believe it starts
with the customer and gets reverse engineered backwards. It should.
But what stood in the way and what's making it feel like maybe
there's some hope? Yeah, yeah, definitely. I do. I'm an optimist
when it comes to this stuff in general, but I do believe there is real
(05:31):
hope. Yeah. I write a newsletter for LinkedIn called AI isn't a goal or a
strategy much to what you just said. So I completely agree.
You know, it's, it's, it's easy to kind of jump on the bandwagon and
say, you know, slap AI on it and stuff. You know, I think
some of the reasons why early entrees into AI
weren't successful is because the, the tooling itself wasn't very smart. I mean
(05:53):
again, like a chat bot that is like pre programmed from
a very limited set of options. I've used those
successfully but maybe I can count on my hand the times that I've actually used
them to solve an actual problem versus one that's intelligent and
fed information and you know, and intended to be helpful.
Those didn't exist, you know, in, at, at scale
(06:15):
a number of years ago. Now they do. So the company that says
let's slap a chatbot on the site because our competitor has it,
they may be choosing the wrong one. So that's just a, that's a, that's slapping
a tactic on a, on a problem without really thinking it through. And, and to
your point, what's the customer experience? And just like with
the phone trees that still, you know, drive at least myself
(06:37):
batty but probably many others go through that process
yourself and actually, you know, actually go through and see if you can solve
anything. But the most basic inquiry, that's fine, but make it really
easy to get out of that, you know. And I think that's the other thing
where you know, as people are getting more comfortable with using AI,
they're also getting more comfortable with this idea that humans
(06:59):
have a real place here too in the, you know, it's not a
replacement, it's a, they're there need to, and there should
be human gatekeepers at multiple points in the
process. So you know, to go back to your, the, the AI chatbot example,
if you use one that's good and there's, there's plenty out there that, that
are, that are very good at what they do, always give the human an
(07:21):
out to say if you want to talk to someone here, you might have to
wait for, you know, 20 minutes or an hour. But if you really want to
do that, like, give them an out to do that and don't force people to
use your AI, but, you know, but give them better tools and
actually, again, go through the process yourself and see what it's like. And if it's,
if it's a miserable experience, then what are you really solving?
(07:42):
Yeah, and we've all been through the miserable experiences and we kind of
shake our heads and say, did you really think that was going to be helpful
for me? Right, right, right. So, yeah, you, you
100% agree with your thought about kind of being
empathetic and walking through the customer's
shoes. Walking in their shoes. I wanted to ask you
(08:03):
about specific applications. We've heard a lot about
agent AI and some of the things
that are coming out. What have you heard that's caught your attention that you
think could really improve the experience for customers? Yeah, definitely. So, yeah,
agentic AI is definitely. It's the buzzword of the,
you know, of the season, let's say. You know, I would say
(08:25):
I'm really excited about it. I think there's a lot of potential there. I've also
heard and experienced some firsthand, but, you know, I've
also heard plenty of stories where you just need, you know, for those
less familiar with it, I mean, essentially agentic AI is an AI talking to
another AI, you know, performing a multiple set of tasks
and scheduling it, and, you know, tons of potential in
(08:47):
that. And yet it relies even more on really good
data, really good connections between tools that
really understand what they're doing. And so, as you would imagine, with increased
complexity comes increased potential to just really screw things
up. So, you know, I have heard some, I wouldn't call them nightmare stories,
but like, some, some stories about, hey, we try. We started down this path and
(09:09):
now we need to, like, take a minute and think this through a little more.
So I think by the end of this year, we're going to hear some really
compelling stories about agentic and, you know, how it's,
how it's gonna, gonna play a role. I would say for those that like to
play around with stuff, there are, even if you use like Chat, GBT or
others like that, you can schedule things. Like, I would recommend playing around with some
(09:30):
of these tools just to kind of understand how they work and so you can
then apply them. I think, you know, what, what I've heard,
I guess, different kind of agent, but, you know, at some of the shows that
I've been to recently is how AI can really augment
and support the customer service agent. And in some ways,
and you, you know, in really interesting ways that also kind of reinforce
(09:53):
what we were just talking about as far as, like, humans are really needed in
this process, but humans don't always have to do all the work. And, and
also humans are really bad at certain kinds of work and machines
are really good at it. So, you know, when you pick up the phone, when
or when a customer service agent picks up the phone, having your
exact history and a summary of that history, rather than having
(10:14):
to read through, you know, pages and pages of note, like it may all be
there. Like, whatever system you're using, I'm sure it has a record of your customers,
but actually giving meaningful summaries of those things, taking really
good notes of the interaction. So the customer service
agent can spend their time focused on you and not making
sure that their records are complete before they have to rush to the next call
(10:36):
or even after the fact, helping with, like, root cause
analysis of issues. So, you know, again, managers of agents,
they don't have ton of. They don't have tons of time to. They can look
at. They're not going to read mountains of transcripts, right? They're going to look at
the biggest problems and do certain things. But
having a tool that helps mass summarize what's going on, what the
(10:57):
problems are and where to look to solve those problems, I mean, those are
three huge ways of looking at, you know, how AI
plays a very good role. And yet humans
get to do what they do best, which is form empathetic relationships
with customers. So an illustration of that might be
customers calling in. You know, you're talking, when you get a larger company, it could
(11:20):
be talking about tens of thousands, hundreds of thousands, even millions of
calls in a month. And there's a ton of data. There's. There's
really too much data almost. And so what you're describing to me
and make sure I'm getting this right, is that the root cause piece is
really appealing as a person, right? Because you want
to not keep playing whack a mole and doing heroics to try.
(11:42):
And, you know, it's great when, when Greg gets an accolade for doing
a triple backflip for a customer who wrote this beautif
testimonial letter. But you don't want to have to perform heroics
because there's something broken upstream. Right? And so I'm
curious about, I mean, the how, but if you generally talk at a high
level about how this works, how, how can it go back and find the root
(12:04):
cause. Yeah, yeah. I mean, and because these tools have access
to a lot deeper, you know, a lot deeper information than
just a call, you know, a customer, they may be frustrated, they may
be focused on one tiny asp. I mean, we all have biases, right? So it
could be the anchor bias of the first thing that happened or the recency bias,
that's the last thing, or any of the other 98 biases or
(12:26):
however many there are AI doesn't. You know, there's, there's plenty of talk
about bias in AI and that's a thing. And so I don't want to discount
that, but I will make the argument that humans have plenty of their own
biases. And so AI has a different way of looking at
data and what might seem anecdotal and to analyze that. And
it also can look at interactions in the aggregate. So it's not
(12:48):
just determining what happened to cause Greg's problem.
But okay, we're seeing, you know, this, this general trend.
Here are some of the points that you need to look at and you know,
and analyze those and then you solve problems for, you know, potentially 10,000
customers because you're, because the AI is able to analyze all of these
things and kind of point in the right direction of, of where to look.
(13:10):
So I think that's again, managers of, they're, they're looking at
statistics, they're looking at know, high level dashboards
and graphs, but no human can actually go through and do
that work unless maybe if they know the question to ask,
they can do it and they get lucky if they ask the right question. It's
kind of a needle in a haystack. Yeah. So, yeah, but it's
(13:31):
like the premise is that you've got good data,
complete data, accurate, complete, and you've got sort of access to
all that relevant data connected to each other. Right. And
then like if I could blue sky this for a minute, since you're an optimist
and I talked about hope earlier, wouldn't it be cool if we could
use AI to not only go to the root cause analysis of an issue,
(13:54):
but periodically, whether it's on demand or once a
day or it's upon request, we could ask the AI
to say, pick out the three improvements that we could
make in the experience, whatever that goes back to, and give
us a cost analysis breakdown on that. Yeah, what
would it cost and what would be the return on that investment
(14:16):
and explain to Us. Why? Because leaders for example are
often tasked with, burdened with whatever held responsible
for the return. And it's not just cx. I say this all the time. Marketing
people are, every department's trying to prove why they need funds and
resources. So. But wouldn't it be cool if
organizationally the AI could do that? Have you heard anything about that in your
(14:38):
travels? I think that's, it's already possible but
for the, you know, the disconnect between the financial data and
the, and the data or marketing data or whatever. So
it's, that's, that's a, that's a near future kind of
thing. And I think there are some organizations that are, that are able to do
that. Maybe it takes an extra hurdle, but I think the platforms that
(14:59):
I'm seeing, they have the capability to start connecting those dots.
It's just what I also see and you know, working with some very large
companies is the data, you know, we talk about people silos and you know,
there's silos of all kinds. The data silos are also
kind of what's holding that back. And you know, some of it is, I'll call
it well intentioned as far as, you know, A doesn't want to share its financial
(15:21):
data with some providers to just, you know, keep things
confidential or whatever. So I totally understand that piece. But this
stuff, you, you already said it. But like this stuff works best when it
sees everything and when it can tie all those things together. So
I would say, you know, that that scenario is, it's not far fetched at
all. I would say it's possible. It's, it just takes some integrations
(15:43):
that a lot of companies I think at this point are maybe reluctant to
do with a single platform. But I think they're, they're connecting
the dots as. So if I'm A,
if I'm A, it doesn't matter B2B or B2C. Let's just say a mid sized
company in my industry and I want to, I want to
deploy AI applications to improve the
(16:06):
experience. I'd hope that would be the top of the list. But obviously to grow
revenues and increase profits and lower costs. What
kind of a partner am I seeking? Like I assume if I'm
Citibank or Delta Airlines or I don't know, Chick Fil
A, I'm going to find a partner to help an AI that
might be, that might look like IBM or someone, you know, someone
(16:29):
like, someone like that McKinsey or PwC,
someone with technology arm. But if I'm a Middle sized
company. That might not be the price tag that's going to work for me.
As I think about other industries, there are sort of mid
range players that work with companies that are mid
size that might not have a staff of Harvard
(16:51):
MBAs on board, but they can help. So tell me, tell me your thoughts about
that. Yeah, I mean I think it's the same integrations apply. I
think, you know, the scale is certainly different. I don't know that the, I
think the challenge is that the scale doesn't necessarily make it
easier. You know, it's, I mean again for, for a, you know,
for a Delta Airlines, obviously they have a lot more data than a, than a
(17:13):
small or medium, even medium sized business. But that's that
medium sized business. It's, they still have hurdle, they still have the
silos and they still have to connect the dots. And that costs and you know,
whether the costs are relative or not, you know, it's, there's still, there's still considerable
costs. I think the, the overall trend in technology, you
know, composable, you know, talk about buzzwords. That's certainly a
(17:35):
trend in the technology world. I see it a lot in the, in the Martech
world. But this idea of products being easier and easier to
integrate with each other, I think there's potential there. I, I think
there still is going to be a hurdle on the financial side of. It depends
on how deep you want to go into integrating. You know,
whether it's CX numbers or marketing numbers or whatever with the
(17:57):
finances. I mean I, I'm seeing some really interesting things
starting to happen, but they're still kind of starting in the
enterprise and, and sometimes working their way down
to smaller companies. But I think it's a, at the very least
it's a hurdle that kind of happens outside of the platform, you know. So in
other words, you, you have the, you have the six data, you know where to
(18:18):
look. Because even, even the platforms I'm talking about, they're very large companies use
them but you know, other smaller companies can use a lot of these platforms as
well. And quite reasonably so, like you have all the data,
you just need to connect the dots kind of outside of the platforms that you
may be using. So it's a hurdle, but it's, it's still, it's still
possible. Yeah. It's so interesting that even with, maybe even, especially
(18:40):
because of all this technology and the speed at which it's becoming
more available that it just puts a spotlight on
what I would call change management efforts and that involves Human beings.
And that involves trust. Right. It involves co
creation and collaboration and building a guiding coalition.
As you, as you talk about silos, still got to bust those to make
(19:03):
all this really effective as it could be. Right? Yeah. And it's, yeah, I mean
it's, you know, you know, this better than me here, but like, you know,
I work on the process side of things and less on the behavioral side of
things. But changes is hard. You know, the, the Kubler
Ross change curve, you know, it mirrors the stages of grief.
Right. So it's like when people, when people's jobs change,
(19:25):
you should think of it in terms of that there's denial at, you know, you
go through the stages. So it's like it's real. And you know, people get used
to doing a job a certain way. And you know, the other slight
tangent, but I think the other, the other thing going on that I'm seeing
is, and this is even from the big companies
who made and still make a lot of money sending surveys out
(19:47):
are moving very quickly towards how do we get information
that isn't survey based. And so, you know, that's, that's great for
the technology. If I can interrupt for a minute? Sure, sure. Information
that's around customer sentiment, right?
Well, sentiment as well as like real time
behavior. And so by tying all of those things together,
(20:09):
I mean I think there's, I'm excited about that piece of it because
I mean, surveys aren't going to go away. But like, I think there's, there's
challenges with surveys and I think some of those challenges are growing with
like fatigue and stuff like that. But you know, you get someone whose job
it's been to, you know, one of the big things in their job has been
to send out surveys and rely on results from those. And then all of a
(20:31):
sudden you're telling them, well, yeah, we're doing that, but we're also collecting
all this other data that feels like it's, maybe that's, that
feels like marketing data or that feels like sales data or whatever. And so,
you know, you're, you're causing a lot of change by adopting these.
Whether, whether you call it a change initiative or not, just the way things are
moving, it's, it's causing people to need to. Even using AI in your
(20:54):
job. Certain people love that idea and gravitate towards it,
others are threatened by it and you know, rightfully so or not. Yeah.
So with any, with any change initiative, we've got to prepare our
team for it. We got to help them understand why this
even matters. Help them get to the point where they're on board with
wanting it. Give them what they need to know and help them feel
(21:16):
good at doing what they need to do and then reinforcing. I just went through
our ad car model. Yeah, yeah, right. But I was going to ask
you, what do you see or what have you heard
that's sort of futuristic, that says, you know, maybe, maybe not
six months from now, but down the road, what's in development? What kinds of things
can we expect down the road that could improve the experience for customers?
(21:39):
Yeah, I do think maybe to go back to the agentic side of things
is, you know, customers being able to solve very
complex problems from a single source. So, you know, some of
that, I mean, just to use a basic example of, you know, you call up
your bank and you get transferred to 20 different people to solve
the. Let alone having multiple problems to solve on the same
(22:01):
call or whatever, which I've also had. But, you know, being able to solve things.
Being able to solve things that are not. Cannot be solved
immediately, you know, that take timing and scheduling to be able
to do. You know, the. Another example would be, you know, planning a
trip with, you know, a complex. You know, you've got hotels,
you've got flights, you've got car rentals, you've got dinner reservations,
(22:23):
you've got all of these things, doing all of that based on a prompt
or maybe a couple prompts. But that's where. That's where things
are headed is, you know, how does AI or a team of
AIs really, from the consumer's perspective, who cares? You know, it
could be 100 AIs, they have a problem and a complex problem
that takes a lot of time, a lot of understanding and a lot of
(22:45):
complexity to solve. Again, not always something that can be
solved in one setting either. But that's. I mean, companies
that are going to be able to solve that, I mean, they're. That's what the
expectation is going to be, you know, and we see little twinkles of
that right now, even with, you know, some subscriptions even or some
things like that. You know, it's like we start anticipating things or being able to
(23:07):
plan ahead, but we're going to be talking about much more complicated things
and the expectations for those that are. I mean, the
nice part about it, it will save some time. The downside, some smaller
organizations are going to struggle at first to connect those dots.
But the large organizations, I know as well,
they're struggling just in different way. You know, they may have more resources to do
(23:29):
it, but they have way more technical debt and complex systems and stuff. So
it's, you know, maybe, maybe the level, the playing field will be level
enough and just everybody's going to be dealing with their own problem. But yeah, I
do. Back to your question. Like, I think it's, you know, a customer has
an issue which they would normally go to three different places to solve and they
solve it at once in a way that, you know, they, they get, they get
(23:51):
the results they need. Well, it's going to be interesting to watch and,
and excited to hear, excited to hear what you're learning
and share all that with the audience. What we have to look
forward to and, and also some of the things you kind of
pointed out as some of the, I guess, hurdles that we need to be, we're
dealing with now, we will continue to deal with and some of the thinking around
(24:13):
starting with the customer and working backwards with whatever technology
we're using. Yeah. My final question to you, Greg. In
someone who's been a marketer for so long, a marketing expert in my
world, who's, who's been involved with cx, you
personally, what delights you as a customer? I was like
feeling understood and you know, and before I say I
(24:35):
have to say something. So when I get, it can be super
simple. It can be a hotel knowing that I've stayed there before, like,
that's surprisingly a surprise, you know,
but it's just like, yeah, just feel it feeling understood so I don't have to
explain myself because usually I will do just about anything to not have to
interface with other people, I guess in
(24:57):
like customer interactions and situations. So, you know, when I do
interact and somebody anticipates something that's, you know, that
I'm a, I'm a customer for life. Yeah. And it's just, I
think to your point, it's so important for companies to understand that about
their customers in their business, in that particular journey or customer
life cycle. What is, what are the critical moments? What are the
(25:19):
points where they're going to maybe need to speak to a human being or the
next level of expertise in a human being and
you can get them there quickly and you can
know them. You know, you can know them. I heard that so much when I
was at the bank about know me and why we do business. We just spent
some time talking about, at least I did, about the challenges that
(25:41):
smaller companies with smaller budgets are going to have to keep up with the AI
race and they could feel like it's David and Goliath. But on the
other hand, they have a strategic advantage by being small
and knowing their customers. Like, that's something they're already better at
the big guys on in most cases. My, my
admonition, I know what you think about this is like, let's not
(26:04):
lose that focus. Right? Yeah, I totally agree. I
think AI is a way for the bigger companies to
mimic some of that stuff. And again, I appreciate it. You know, when I stay
at a large hotel chain and again, they know that I stay. It's like, I
don't care if that individual knows anything about me, but it's, it's. I
appreciate that the company owners. But when I walk into a small
(26:25):
business and they, and I was there yesterday and they don't
recognize me, like, shame on them. You know, it's like, that's a. But to your
point, like, they, they have a, they have a much better opportunity because they're
closer to the data, they're closer to their customers and everything like that. So,
yeah, lean into it. Every customer experience improvement
doesn't have to be a technology improvement, even though I know that's on
(26:47):
everybody's mind. But use AI and all that stuff when you can.
It's a huge time saver, efficiency saver. But people still
matter, right? People definitely matter. Greg, thank you so much for being
a guest again on the Delighted Customers podcast and sharing your
wisdom. Yeah, thanks so much for having me.
Well, I couldn't be more excited to have on the Delighted Customers
podcast this episode, Greg Kilstrom. Greg
and I go way back, actually over 15 years before
I got involved in CX at all. And he has somehow come
back into my life. And we crossed paths in Baltimore at one point
at a CXPA event that happened down there. And we
(00:22):
have developed a business partnership in that I and part
of his umbrella of podcast, a part of the Agile Brand
Network. Greg is the host of the Agile Brand
Podcast Network. Greg, welcome back to the Delighted
Customers podcast. Yeah, thanks so much. Yeah, excited to be here.
And I'm thrilled at the success of your show as well. It's been a
(00:44):
great partnership. Well, Greg, can you share? And I agree 100%,
and thank you so much for helping the work that you and your team
do to help lift up our show here. Can you share a little bit more
about what you do with our audience? Yeah, of course. Yeah. So I come originally
from a marketing background, so I owned. I say I'm a
recovering agency owner. So I started
(01:06):
sold an agency, ran it for about 14 years here in the. In the Washington,
D.C. area where I'm still at always throughout
that, you know, I was very passionate about the kind of the
intersection of customers and the technology that
serves those customers. And I found that, you know, after I sold the
agency, now work as an independent consultant for a number of
(01:28):
organizations. I also host the podcast. I've written a few books. I always find
myself kind of going back to whether it's in marketing, whether it's in CX
or other areas, just how can technology better serve not
only consumers and the end customers, but also the employees
and the consumers of that technology within the
corporation. So, yeah, looking forward to talking with you. Well, I consider
(01:50):
you a martech expert, Greg, and I know
you attend a lot of conferences in marketing and CX and other areas.
And I know recently you've been traveling. You still have some left to
do here, but you've been to places where there's a lot
of mind sharing going on around what's new and
cutting edge. And what I'd like to do is pick your brain a
(02:13):
little bit, if it's okay. Yeah, yeah. And share with our audience some of the
newer things, some of the fresher things that are out there. In particular, if we
can think through the customer life cycle and
different applications of AI and technology
that you've seen out there, that could help really nurture and
deepen relationships that brands have with their customer
(02:34):
base, whether it's when they're trying to get them to make that
initial purchase or they've already come on board and now we're servicing
them, the service operation model, even
helping the organization measure the success of
their customer base or how they're delivering the experience. Does that make
sense? Yeah, yeah, definitely. Happy to, happy to talk through. Yeah. I mean, I can
(02:57):
maybe, maybe just start at the beginning of the journey. And I think
there's, there's certainly, there's been tons of talk about AI. There's
tons of information misinformation and tons of stats that people
kind of hang on to, I think. But I think what's,
what's been really interesting over the last several months is
there's like, I think there's certain anecdotes that people kind of glom
(03:19):
onto as far as, like, consumers hate
chatbots and so therefore we should never use that. So
just to use an early example in the, in the journey, like, it's not
that consumers hate any form of AI based
interaction or something early in the journey, it's that they really don't
like dumb chat bots that aren't helpful, just like they don't like
(03:41):
dumb phone trees that send them in the endless doom loop. So
getting past some of the early preconceptions of consumers don't
like some of this stuff versus what don't they like about it? Consumers
love personalized content offers, deals, you name
it, interactions, and just that a, that a brand understands them.
And what was always, you know, the traditional way of
(04:04):
understanding a brand, you know, if you run a mom and pop store, it's like,
you know, the people that walk in every day by name and you know what
their kids are doing and blah, blah, blah. But AI and personalization
allows you to do that at scale. And so consumers really like that. What they
don't like is being shown irrelevant offers that are, you know,
quote unquote, personalized, but really just based on
(04:25):
some, you know, some irrelevant information. And so I think
it's because it's continually evolving. There's a lot of people
that are kind of thinking, I would say the thinking two years ago versus
the thinking today, because these tools have gotten a lot better at
understanding and anticipating, you know, and that's just talking about early
on in the journey. Yeah.
(04:47):
So yeah, Greg, what do you think stood in
the way of companies using AI in
the form of chatbots or something similar
effectively in the past and what is helping to
change that now? Because, you know, one of the things that
comes to mind for me is technology. In
(05:09):
search of a strategy never really works. And I believe, and by the nature
of this name of this show, delighted customers podcast, we believe it starts
with the customer and gets reverse engineered backwards. It should.
But what stood in the way and what's making it feel like maybe
there's some hope? Yeah, yeah, definitely. I do. I'm an optimist
when it comes to this stuff in general, but I do believe there is real
(05:31):
hope. Yeah. I write a newsletter for LinkedIn called AI isn't a goal or a
strategy much to what you just said. So I completely agree.
You know, it's, it's, it's easy to kind of jump on the bandwagon and
say, you know, slap AI on it and stuff. You know, I think
some of the reasons why early entrees into AI
weren't successful is because the, the tooling itself wasn't very smart. I mean
(05:53):
again, like a chat bot that is like pre programmed from
a very limited set of options. I've used those
successfully but maybe I can count on my hand the times that I've actually used
them to solve an actual problem versus one that's intelligent and
fed information and you know, and intended to be helpful.
Those didn't exist, you know, in, at, at scale
(06:15):
a number of years ago. Now they do. So the company that says
let's slap a chatbot on the site because our competitor has it,
they may be choosing the wrong one. So that's just a, that's a, that's slapping
a tactic on a, on a problem without really thinking it through. And, and to
your point, what's the customer experience? And just like with
the phone trees that still, you know, drive at least myself
(06:37):
batty but probably many others go through that process
yourself and actually, you know, actually go through and see if you can solve
anything. But the most basic inquiry, that's fine, but make it really
easy to get out of that, you know. And I think that's the other thing
where you know, as people are getting more comfortable with using AI,
they're also getting more comfortable with this idea that humans
(06:59):
have a real place here too in the, you know, it's not a
replacement, it's a, they're there need to, and there should
be human gatekeepers at multiple points in the
process. So you know, to go back to your, the, the AI chatbot example,
if you use one that's good and there's, there's plenty out there that, that
are, that are very good at what they do, always give the human an
(07:21):
out to say if you want to talk to someone here, you might have to
wait for, you know, 20 minutes or an hour. But if you really want to
do that, like, give them an out to do that and don't force people to
use your AI, but, you know, but give them better tools and
actually, again, go through the process yourself and see what it's like. And if it's,
if it's a miserable experience, then what are you really solving?
(07:42):
Yeah, and we've all been through the miserable experiences and we kind of
shake our heads and say, did you really think that was going to be helpful
for me? Right, right, right. So, yeah, you, you
100% agree with your thought about kind of being
empathetic and walking through the customer's
shoes. Walking in their shoes. I wanted to ask you
(08:03):
about specific applications. We've heard a lot about
agent AI and some of the things
that are coming out. What have you heard that's caught your attention that you
think could really improve the experience for customers? Yeah, definitely. So, yeah,
agentic AI is definitely. It's the buzzword of the,
you know, of the season, let's say. You know, I would say
(08:25):
I'm really excited about it. I think there's a lot of potential there. I've also
heard and experienced some firsthand, but, you know, I've
also heard plenty of stories where you just need, you know, for those
less familiar with it, I mean, essentially agentic AI is an AI talking to
another AI, you know, performing a multiple set of tasks
and scheduling it, and, you know, tons of potential in
(08:47):
that. And yet it relies even more on really good
data, really good connections between tools that
really understand what they're doing. And so, as you would imagine, with increased
complexity comes increased potential to just really screw things
up. So, you know, I have heard some, I wouldn't call them nightmare stories,
but like, some, some stories about, hey, we try. We started down this path and
(09:09):
now we need to, like, take a minute and think this through a little more.
So I think by the end of this year, we're going to hear some really
compelling stories about agentic and, you know, how it's,
how it's gonna, gonna play a role. I would say for those that like to
play around with stuff, there are, even if you use like Chat, GBT or
others like that, you can schedule things. Like, I would recommend playing around with some
(09:30):
of these tools just to kind of understand how they work and so you can
then apply them. I think, you know, what, what I've heard,
I guess, different kind of agent, but, you know, at some of the shows that
I've been to recently is how AI can really augment
and support the customer service agent. And in some ways,
and you, you know, in really interesting ways that also kind of reinforce
(09:53):
what we were just talking about as far as, like, humans are really needed in
this process, but humans don't always have to do all the work. And, and
also humans are really bad at certain kinds of work and machines
are really good at it. So, you know, when you pick up the phone, when
or when a customer service agent picks up the phone, having your
exact history and a summary of that history, rather than having
(10:14):
to read through, you know, pages and pages of note, like it may all be
there. Like, whatever system you're using, I'm sure it has a record of your customers,
but actually giving meaningful summaries of those things, taking really
good notes of the interaction. So the customer service
agent can spend their time focused on you and not making
sure that their records are complete before they have to rush to the next call
(10:36):
or even after the fact, helping with, like, root cause
analysis of issues. So, you know, again, managers of agents,
they don't have ton of. They don't have tons of time to. They can look
at. They're not going to read mountains of transcripts, right? They're going to look at
the biggest problems and do certain things. But
having a tool that helps mass summarize what's going on, what the
(10:57):
problems are and where to look to solve those problems, I mean, those are
three huge ways of looking at, you know, how AI
plays a very good role. And yet humans
get to do what they do best, which is form empathetic relationships
with customers. So an illustration of that might be
customers calling in. You know, you're talking, when you get a larger company, it could
(11:20):
be talking about tens of thousands, hundreds of thousands, even millions of
calls in a month. And there's a ton of data. There's. There's
really too much data almost. And so what you're describing to me
and make sure I'm getting this right, is that the root cause piece is
really appealing as a person, right? Because you want
to not keep playing whack a mole and doing heroics to try.
(11:42):
And, you know, it's great when, when Greg gets an accolade for doing
a triple backflip for a customer who wrote this beautif
testimonial letter. But you don't want to have to perform heroics
because there's something broken upstream. Right? And so I'm
curious about, I mean, the how, but if you generally talk at a high
level about how this works, how, how can it go back and find the root
(12:04):
cause. Yeah, yeah. I mean, and because these tools have access
to a lot deeper, you know, a lot deeper information than
just a call, you know, a customer, they may be frustrated, they may
be focused on one tiny asp. I mean, we all have biases, right? So it
could be the anchor bias of the first thing that happened or the recency bias,
that's the last thing, or any of the other 98 biases or
(12:26):
however many there are AI doesn't. You know, there's, there's plenty of talk
about bias in AI and that's a thing. And so I don't want to discount
that, but I will make the argument that humans have plenty of their own
biases. And so AI has a different way of looking at
data and what might seem anecdotal and to analyze that. And
it also can look at interactions in the aggregate. So it's not
(12:48):
just determining what happened to cause Greg's problem.
But okay, we're seeing, you know, this, this general trend.
Here are some of the points that you need to look at and you know,
and analyze those and then you solve problems for, you know, potentially 10,000
customers because you're, because the AI is able to analyze all of these
things and kind of point in the right direction of, of where to look.
(13:10):
So I think that's again, managers of, they're, they're looking at
statistics, they're looking at know, high level dashboards
and graphs, but no human can actually go through and do
that work unless maybe if they know the question to ask,
they can do it and they get lucky if they ask the right question. It's
kind of a needle in a haystack. Yeah. So, yeah, but it's
(13:31):
like the premise is that you've got good data,
complete data, accurate, complete, and you've got sort of access to
all that relevant data connected to each other. Right. And
then like if I could blue sky this for a minute, since you're an optimist
and I talked about hope earlier, wouldn't it be cool if we could
use AI to not only go to the root cause analysis of an issue,
(13:54):
but periodically, whether it's on demand or once a
day or it's upon request, we could ask the AI
to say, pick out the three improvements that we could
make in the experience, whatever that goes back to, and give
us a cost analysis breakdown on that. Yeah, what
would it cost and what would be the return on that investment
(14:16):
and explain to Us. Why? Because leaders for example are
often tasked with, burdened with whatever held responsible
for the return. And it's not just cx. I say this all the time. Marketing
people are, every department's trying to prove why they need funds and
resources. So. But wouldn't it be cool if
organizationally the AI could do that? Have you heard anything about that in your
(14:38):
travels? I think that's, it's already possible but
for the, you know, the disconnect between the financial data and
the, and the data or marketing data or whatever. So
it's, that's, that's a, that's a near future kind of
thing. And I think there are some organizations that are, that are able to do
that. Maybe it takes an extra hurdle, but I think the platforms that
(14:59):
I'm seeing, they have the capability to start connecting those dots.
It's just what I also see and you know, working with some very large
companies is the data, you know, we talk about people silos and you know,
there's silos of all kinds. The data silos are also
kind of what's holding that back. And you know, some of it is, I'll call
it well intentioned as far as, you know, A doesn't want to share its financial
(15:21):
data with some providers to just, you know, keep things
confidential or whatever. So I totally understand that piece. But this
stuff, you, you already said it. But like this stuff works best when it
sees everything and when it can tie all those things together. So
I would say, you know, that that scenario is, it's not far fetched at
all. I would say it's possible. It's, it just takes some integrations
(15:43):
that a lot of companies I think at this point are maybe reluctant to
do with a single platform. But I think they're, they're connecting
the dots as. So if I'm A,
if I'm A, it doesn't matter B2B or B2C. Let's just say a mid sized
company in my industry and I want to, I want to
deploy AI applications to improve the
(16:06):
experience. I'd hope that would be the top of the list. But obviously to grow
revenues and increase profits and lower costs. What
kind of a partner am I seeking? Like I assume if I'm
Citibank or Delta Airlines or I don't know, Chick Fil
A, I'm going to find a partner to help an AI that
might be, that might look like IBM or someone, you know, someone
(16:29):
like, someone like that McKinsey or PwC,
someone with technology arm. But if I'm a Middle sized
company. That might not be the price tag that's going to work for me.
As I think about other industries, there are sort of mid
range players that work with companies that are mid
size that might not have a staff of Harvard
(16:51):
MBAs on board, but they can help. So tell me, tell me your thoughts about
that. Yeah, I mean I think it's the same integrations apply. I
think, you know, the scale is certainly different. I don't know that the, I
think the challenge is that the scale doesn't necessarily make it
easier. You know, it's, I mean again for, for a, you know,
for a Delta Airlines, obviously they have a lot more data than a, than a
(17:13):
small or medium, even medium sized business. But that's that
medium sized business. It's, they still have hurdle, they still have the
silos and they still have to connect the dots. And that costs and you know,
whether the costs are relative or not, you know, it's, there's still, there's still considerable
costs. I think the, the overall trend in technology, you
know, composable, you know, talk about buzzwords. That's certainly a
(17:35):
trend in the technology world. I see it a lot in the, in the Martech
world. But this idea of products being easier and easier to
integrate with each other, I think there's potential there. I, I think
there still is going to be a hurdle on the financial side of. It depends
on how deep you want to go into integrating. You know,
whether it's CX numbers or marketing numbers or whatever with the
(17:57):
finances. I mean I, I'm seeing some really interesting things
starting to happen, but they're still kind of starting in the
enterprise and, and sometimes working their way down
to smaller companies. But I think it's a, at the very least
it's a hurdle that kind of happens outside of the platform, you know. So in
other words, you, you have the, you have the six data, you know where to
(18:18):
look. Because even, even the platforms I'm talking about, they're very large companies use
them but you know, other smaller companies can use a lot of these platforms as
well. And quite reasonably so, like you have all the data,
you just need to connect the dots kind of outside of the platforms that you
may be using. So it's a hurdle, but it's, it's still, it's still
possible. Yeah. It's so interesting that even with, maybe even, especially
(18:40):
because of all this technology and the speed at which it's becoming
more available that it just puts a spotlight on
what I would call change management efforts and that involves Human beings.
And that involves trust. Right. It involves co
creation and collaboration and building a guiding coalition.
As you, as you talk about silos, still got to bust those to make
(19:03):
all this really effective as it could be. Right? Yeah. And it's, yeah, I mean
it's, you know, you know, this better than me here, but like, you know,
I work on the process side of things and less on the behavioral side of
things. But changes is hard. You know, the, the Kubler
Ross change curve, you know, it mirrors the stages of grief.
Right. So it's like when people, when people's jobs change,
(19:25):
you should think of it in terms of that there's denial at, you know, you
go through the stages. So it's like it's real. And you know, people get used
to doing a job a certain way. And you know, the other slight
tangent, but I think the other, the other thing going on that I'm seeing
is, and this is even from the big companies
who made and still make a lot of money sending surveys out
(19:47):
are moving very quickly towards how do we get information
that isn't survey based. And so, you know, that's, that's great for
the technology. If I can interrupt for a minute? Sure, sure. Information
that's around customer sentiment, right?
Well, sentiment as well as like real time
behavior. And so by tying all of those things together,
(20:09):
I mean I think there's, I'm excited about that piece of it because
I mean, surveys aren't going to go away. But like, I think there's, there's
challenges with surveys and I think some of those challenges are growing with
like fatigue and stuff like that. But you know, you get someone whose job
it's been to, you know, one of the big things in their job has been
to send out surveys and rely on results from those. And then all of a
(20:31):
sudden you're telling them, well, yeah, we're doing that, but we're also collecting
all this other data that feels like it's, maybe that's, that
feels like marketing data or that feels like sales data or whatever. And so,
you know, you're, you're causing a lot of change by adopting these.
Whether, whether you call it a change initiative or not, just the way things are
moving, it's, it's causing people to need to. Even using AI in your
(20:54):
job. Certain people love that idea and gravitate towards it,
others are threatened by it and you know, rightfully so or not. Yeah.
So with any, with any change initiative, we've got to prepare our
team for it. We got to help them understand why this
even matters. Help them get to the point where they're on board with
wanting it. Give them what they need to know and help them feel
(21:16):
good at doing what they need to do and then reinforcing. I just went through
our ad car model. Yeah, yeah, right. But I was going to ask
you, what do you see or what have you heard
that's sort of futuristic, that says, you know, maybe, maybe not
six months from now, but down the road, what's in development? What kinds of things
can we expect down the road that could improve the experience for customers?
(21:39):
Yeah, I do think maybe to go back to the agentic side of things
is, you know, customers being able to solve very
complex problems from a single source. So, you know, some of
that, I mean, just to use a basic example of, you know, you call up
your bank and you get transferred to 20 different people to solve
the. Let alone having multiple problems to solve on the same
(22:01):
call or whatever, which I've also had. But, you know, being able to solve things.
Being able to solve things that are not. Cannot be solved
immediately, you know, that take timing and scheduling to be able
to do. You know, the. Another example would be, you know, planning a
trip with, you know, a complex. You know, you've got hotels,
you've got flights, you've got car rentals, you've got dinner reservations,
(22:23):
you've got all of these things, doing all of that based on a prompt
or maybe a couple prompts. But that's where. That's where things
are headed is, you know, how does AI or a team of
AIs really, from the consumer's perspective, who cares? You know, it
could be 100 AIs, they have a problem and a complex problem
that takes a lot of time, a lot of understanding and a lot of
(22:45):
complexity to solve. Again, not always something that can be
solved in one setting either. But that's. I mean, companies
that are going to be able to solve that, I mean, they're. That's what the
expectation is going to be, you know, and we see little twinkles of
that right now, even with, you know, some subscriptions even or some
things like that. You know, it's like we start anticipating things or being able to
(23:07):
plan ahead, but we're going to be talking about much more complicated things
and the expectations for those that are. I mean, the
nice part about it, it will save some time. The downside, some smaller
organizations are going to struggle at first to connect those dots.
But the large organizations, I know as well,
they're struggling just in different way. You know, they may have more resources to do
(23:29):
it, but they have way more technical debt and complex systems and stuff. So
it's, you know, maybe, maybe the level, the playing field will be level
enough and just everybody's going to be dealing with their own problem. But yeah, I
do. Back to your question. Like, I think it's, you know, a customer has
an issue which they would normally go to three different places to solve and they
solve it at once in a way that, you know, they, they get, they get
(23:51):
the results they need. Well, it's going to be interesting to watch and,
and excited to hear, excited to hear what you're learning
and share all that with the audience. What we have to look
forward to and, and also some of the things you kind of
pointed out as some of the, I guess, hurdles that we need to be, we're
dealing with now, we will continue to deal with and some of the thinking around
(24:13):
starting with the customer and working backwards with whatever technology
we're using. Yeah. My final question to you, Greg. In
someone who's been a marketer for so long, a marketing expert in my
world, who's, who's been involved with cx, you
personally, what delights you as a customer? I was like
feeling understood and you know, and before I say I
(24:35):
have to say something. So when I get, it can be super
simple. It can be a hotel knowing that I've stayed there before, like,
that's surprisingly a surprise, you know,
but it's just like, yeah, just feel it feeling understood so I don't have to
explain myself because usually I will do just about anything to not have to
interface with other people, I guess in
(24:57):
like customer interactions and situations. So, you know, when I do
interact and somebody anticipates something that's, you know, that
I'm a, I'm a customer for life. Yeah. And it's just, I
think to your point, it's so important for companies to understand that about
their customers in their business, in that particular journey or customer
life cycle. What is, what are the critical moments? What are the
(25:19):
points where they're going to maybe need to speak to a human being or the
next level of expertise in a human being and
you can get them there quickly and you can
know them. You know, you can know them. I heard that so much when I
was at the bank about know me and why we do business. We just spent
some time talking about, at least I did, about the challenges that
(25:41):
smaller companies with smaller budgets are going to have to keep up with the AI
race and they could feel like it's David and Goliath. But on the
other hand, they have a strategic advantage by being small
and knowing their customers. Like, that's something they're already better at
the big guys on in most cases. My, my
admonition, I know what you think about this is like, let's not
(26:04):
lose that focus. Right? Yeah, I totally agree. I
think AI is a way for the bigger companies to
mimic some of that stuff. And again, I appreciate it. You know, when I stay
at a large hotel chain and again, they know that I stay. It's like, I
don't care if that individual knows anything about me, but it's, it's. I
appreciate that the company owners. But when I walk into a small
(26:25):
business and they, and I was there yesterday and they don't
recognize me, like, shame on them. You know, it's like, that's a. But to your
point, like, they, they have a, they have a much better opportunity because they're
closer to the data, they're closer to their customers and everything like that. So,
yeah, lean into it. Every customer experience improvement
doesn't have to be a technology improvement, even though I know that's on
(26:47):
everybody's mind. But use AI and all that stuff when you can.
It's a huge time saver, efficiency saver. But people still
matter, right? People definitely matter. Greg, thank you so much for being
a guest again on the Delighted Customers podcast and sharing your
wisdom. Yeah, thanks so much for having me.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.