January 28, 2025 • 42 mins
Customer Success Summit in Austin on February 11th & 12th: https://events.customersuccesscollective.com/location/austin
This is an episode of "MondAI CS", a new podcast by our CS friend Michael Forney that we recorded last week.
I felt that the content would be fantastic to post here as well.
This episode delves into the intersection of AI and customer success, discussing how technology can both enhance and disrupt customer relationships. We explore the balance between automation and personal interaction, the guilt associated with not using AI, and practical applications for CSMs, illustrating a balanced approach that integrates AI while emphasizing the necessity of human connection.
• The importance of balancing automation with human interaction
• Guilt surrounding the adoption of AI tools
• Practical AI applications in customer success
• The value of combining technology with personalized engagement
• The need for checks and balances in automation strategies
• Future considerations for integrating AI in customer journeys
Enjoy! I know I sure did...
Follow Michael's Podcast MondAI CS here: https://open.spotify.com/show/497aLePSmsVJs1mMwn5dSL
Michael's LinkedIn: https://www.linkedin.com/in/michael-forney/
+++++++++++++++++
Like/Subscribe/Review:
If you are getting value from the show, please follow/subscribe so that you don't miss an episode and consider leaving us a review.
Website:
For more information about the show or to get in touch, visit DigitalCustomerSuccess.com.
Buy Alex a Cup of Coffee:
This show runs exclusively on caffeine - and lots of it. If you like what we're, consider supporting our habit by buying us a cup of coffee: https://bmc.link/dcsp
Thank you for all of your support!
The Digital Customer Success Podcast is hosted by Alex Turkovic
🎬 This content was edited by Lifetime Value Media.
Learn more at: https://www.lifetimevaluemedia.com
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:03):
if you over automate the like automated messages that
are being sent on behalf of thecsm, at that point you kind of
lose control of the narrative.
Thing that I preach over andover again is you have to insert
yourself into your customer'sjourney on a regular basis as a
leader.
The more automation you put inplace without checks and
(00:23):
balances, the more likely you'rejust confusing your customer
with stuff coming at them allover the place.
Once again, welcome to theDigital Customer Experience
Podcast with me, alex Turkovich,so glad you could join us here
today and every week as weexplore how digital can help
enhance the customer andemployee experience.
My goal is to share what myguests and I have learned over
(00:45):
the years so that you can getthe insights that you need to
evolve your own digital programs.
If you'd like more info, needto get in touch or sign up for
the weekly companion newsletterthat has additional articles and
resources in it.
Go to digitalcustomersuccesscom.
For now, let's get started.
Hello and welcome back to theDigital CX podcast, the show
(01:05):
where we talk about all thingsdigital in CX.
I kind of mash those wordstogether, but you get what I
mean.
This is episode 89.
89 for you today, so I'mcreeping on towards 100 episodes
, which I'm super excited about,and have a couple surprises in
store for you for that episode.
Regardless, today is going tobe a little bit unconventional,
(01:29):
I guess, and we've done thisonce before where we've kind of
had two shows recorded at thesame time.
But my guest today is MichaelForney, who's a local Austinite.
He's a CS leader and he'sstarting his own show that is
really focused on AI and CS.
So he was kind enough to inviteme for his inaugural episode.
(01:51):
So we recorded that not longago, I think last week at some
point, and so I figured he'sreleased it on his stream.
I figured I would release it aswell on mine, because it's a
really cool conversation justabout our thoughts on current
state of AI in CS and, yeah, Ifigured you'd get a kick out of
(02:11):
the conversation as well.
So today, a shared episode withmyself and Michael Forney.
Just a quick housekeeping notebefore we get into the episode.
I will be speaking at thecustomer success summit event in
Austin February I think it'sthe 10th and the 11th, 11th,
12th, something like that.
(02:32):
I'll link the event details inthe show notes so that you can
go and have a look and if youwant to attend, I'd love to see
you there.
Without further ado, here's theepisode with Michael Forney.
Speaker 2 (02:47):
Alex, thank you so much for being on the first
episode of Monday I CS, where wetalk all things AI and CS.
Speaker 1 (02:56):
I was very curious to hear you pronounce Monday.
I.
Speaker 2 (03:02):
I was curious too.
That was the first time I hadvocalized it inside of a
microphone felt okay.
I think I can do better monday.
I think there's a.
There's an important pausebetween the end of the d and the
a, or you just call it monday Iwas thinking that too.
Yeah, but then I don't know.
Let's see what happens in thewild, let's see what other
(03:24):
people say.
Speaker 1 (03:24):
I can't wait to see where this journey takes you,
but it's great to be here.
Speaker 2 (03:28):
Thanks, thank you, dude.
Well, you've been very generous, you know, with your time and
you were just showing me areally interesting example.
Yeah, you leveraging Is it AIor is it robotics, or is it both
Bit of both?
Speaker 1 (03:40):
I guess All right.
Speaker 2 (03:41):
So tell me about this , or do you?
Do you not want to reveal?
Speaker 1 (03:44):
No, no, no no, yeah, that's fine.
Um, so I just uh, you know,joined a new company uh what?
Three months ago, november,something like that and uh, one
of the things we decided to dothis year was to send out, um
holiday cards, um, but wanted todo it handwritten.
(04:04):
So we actually used thisawesome service called
Handwritten with V-R-Y-T-T-E-N.
Speaker 2 (04:11):
So go look them up.
Speaker 1 (04:12):
I think it's just handwrittencom, where basically
all you do is you upload anaddress list and the message
that you want to send and theyhave a bunch of robots in the
background that do thehandwriting.
So I'll show you the, I'llblock my address out, but this
is kind of what it looks like onthe envelope, so it looks like
it's handwritten, and then youbasically design your card.
You can do it on the platformand then you upload it and they
(04:35):
put a little handwritten messagein there.
You can even like insert, youknow, booleans or tokens for,
like first name and all thatkind of stuff wow in the message
and they send it and so likewe're integrating that into our
tech stack too, to like sendhandwritten notes to our
customers when they graduate onboarding and all that kind of
stuff.
(04:55):
I think the ai part would wouldbe using some of the ai kind of
signals that we get to triggersome of those handwritten notes
and whatnot.
But super cool, super coolstuff because it combines that
human element with the machineelement.
Speaker 2 (05:10):
Yeah, completely, and you get to stand out in the
paper mail, which is actuallykind of a good way to market and
advertise these days, right,and it doesn't look like shit.
Speaker 1 (05:20):
Sorry, can I cuss on this show?
You can?
Do whatever you want GreatClothes are coming off.
No, it doesn't look like crap.
You know how you get thoseobviously fake handwritten.
It actually looks like somebodyput pen to paper, which is
bizarre.
This isn't a handwrittencommercial, by the way.
I promise you that, but there'sa list of 100 different writing
(05:44):
styles that you can choose from, so you can choose, like really
sloppy ones or really neat ones.
So it's cool.
Speaker 2 (05:49):
I think the fascinating part about that is
that it actually it's actuallywritten.
I mean, I've seen this kind ofthing before but it's printed
and then when you receive it ittotally takes away from it.
When you see the fact that thethe what should be handwriting
is is printed and it's like, atthat point I just kind of tune
out.
Speaker 1 (06:06):
So that's cool.
Speaker 2 (06:07):
Yeah, yeah, yeah, and I like that, yeah, I like that
you're thinking about doing thison the heels of, say,
onboarding, and so what I wasactually going to do to open
this show up is just to say youknow, hey, alex, when someone
says AI, what is coming to yourmind?
Like, when you think about AI,what is on your mind?
Speaker 1 (06:26):
Chaos and confusion.
Speaker 2 (06:28):
Chaos and confusion.
So you think we're still in,like the thrash of AI?
Yeah, I think so.
Speaker 1 (06:33):
I mean, look, you know, I think there's a lot of
just confusion about how to useit and I think, even a little
bit of guilt there's likeprofessional guilt involved.
If people, like, are using itless than others, you know, and
it's like you know I should beusing ai more than I am, or you
know they start shittingthemselves about it.
But, like, I think one of thethings that I am finding is we
(06:59):
have because the cost ofdevelopment is plummeting and we
now have all these you know,kind of startup companies that
are in the CS niche, that arekind of you know, developing
some use case centeredtechnology.
We have this kind of wide swathof different startups.
(07:19):
And then you know largercompanies, like your Gainsights
and your Tatangos and whatnot,that are all tackling AI and
integrating AI in some form orfashion, some as a kind of a
direct product and some as kindof like add-on to their direct
product, and so it's just themarket is just so wide and
desperate with like differentthings, and so I think what
(07:40):
we're going to start seeing islike a pretty rapid
consolidation of use cases andcompanies, you know, and a
swallowing up of differentvendors that are doing different
things, and hopefully that'lladd some clarity into, like, you
know, what platform should youlook for?
What technology should you lookfor?
What vendor do you need to youknow target?
(08:01):
Because you know I think rightnow it's just it's the wild,
wild west and it can getconfusing very, very fast and I
know a lot of people when theyget kind of overwhelmed and
confused with it they just kindof shut their laptop and walk
away.
Speaker 2 (08:16):
So I want to pick up on what you said about the guilt
.
Yeah, this is interesting, yeah.
So in many ways I look at AIand I sometimes I do have a
little bit of that sense ofguilt, but then I try to remind
myself.
You know, if you use acalculator today, nobody has
guilt about using a calculator,knowing that the actual math
(08:37):
behind everything would take youforever to do and if you're
like me, you probably justcouldn't even do it.
Speaker 1 (08:44):
Anyway, anyway, even if it's regular.
Speaker 2 (08:46):
It takes me an embarrassingly amount, a long
amount of time to do even justsimple addition and subtraction
so I I would I would be nowherewithout a calculator or excel so
so tell me more about what youmean by guilt with ai.
What's?
Speaker 1 (09:00):
yeah, I think we might be thinking about it in
slightly different ways, because, correct, no, it's fine.
This is important because Ithink you're thinking about
people who have guilt using AIinstead of doing it the hard way
or whatever.
I'm thinking that there's a lotof people out there who feel
guilty for not having adopted ityet, or they're not actively
(09:21):
using chat GPT as part of theirdaily life, or Gemini or
whatever you know you knowflavor you're, you're plugged
into or whatever, because thereare a lot of people who just
aren't you know,aren't using it on a regular
basis and I mean, um, you know,I think that's okay.
Yeah, I think it's okay.
(09:41):
Are you going to be better off,in the long run, knowing how to
prompt versus not knowing howto prompt?
I don't really think so,because I think prompting is a
short.
There's a finite window of timethat we're in right now where
prompting is the skill that youneed to have to effectively use
(10:01):
artificial intelligence.
But that prompting stuff is allgoing to go behind, you know,
behind the wall, so to speak.
Like I don't know.
I don't think 10, 15 years fromnow, you're going to need to be
real, real good at prompting,because, guess what?
Artificial intelligence will besentient and we'll all be out
(10:21):
of a job anyway.
So whatever.
Speaker 2 (10:24):
True, true, true, true, true.
I think about I mean, asexcited as I am about ai,
there's like at least a part ofme and anyone who grew up
watching the likes of terminator2.
It just feels like aninevitable future that we all
kind of put outside or at leastme, I put it outside my mind, my
mind, enjoying the rollercoaster until that happens, or
(10:46):
hopefully doesn't happen.
Speaker 1 (10:48):
Well, look, I mean, here's the deal.
We're all human and we'retalking about customer success,
right, and in customer success,humans are always going to want
to talk to other humans, yeah,right.
So if you're, you know, we'rein SAS, right?
So if we're speaking with acustomer or we're engaging with
a customer, exchanging actualwhat I like to call mouth words
(11:11):
with the customer is a very,very important thing, and I
don't think that's goinganywhere.
Helping you prep for thoseconversations, helping you have
those conversations, helping youlook up information in real
time while you're having thoseconversations so you can have
more productive ones and be moreefficient with your time, and
(11:31):
all that kind of stuff.
And I think that's what we'restarting to see, right, with the
chatbots and live agents andco-pilots and all that kind of
stuff that helps us have thoseconversations.
But also, like to our pointearlier, I don't think we've
quite gotten there yet, like wehaven't quite arrived at that
point yet.
It's still just like this thingwe have open chat, gpt or
(11:53):
whatever, and we ask it to dosome things for us, but it's not
like this live agent helpersituation so much to put to pull
on there.
Speaker 2 (12:03):
I think one, though one of the first things you had
said was about people maybefeeling guilty not using it.
So, in the spirit of giving youknow, give the cs world.
Let's give the cs world onepractical takeaway of like, what
is something, if people saywant to get in on the fun,
what's something that they cando or use to their advantage to
(12:26):
help them out that can be easilyadopted.
Speaker 1 (12:28):
Yeah, for sure, Well, I mean.
So there's a couple things andI did record an episode about
that specifically that you cango check out.
I'll give you the link to it.
It's basically an AI primer forCSMs and it gives you some
practical kind of use cases ofAI, also gives you a nice like
prompting framework that you canuse to kind of make your
(12:50):
prompts a little bit better andall that kind of stuff.
But I think, look, there's somesuper simple things that I think
a lot of tools are doing for usnow, like call transcripts and
things like that that I thinkare pretty easy to get at,
whether the tool that you'reusing to record your calls does
it or not, or whether youbasically feed a transcript into
(13:11):
ChatGPT to create a summary foryou.
I think that's a massive timesaver.
One little slight tweak to thatthat I would add is using a
tool like Gamma to create aquick PowerPoint deck as a
summary for your calls.
Like, I've thrown my AI callsummaries into Gamma, had it
create like a three or fourslide deck for me and sent the
(13:33):
deck out as a call recap, whichis kind of a cool thing because
it's a visual thing.
So I think there's all kinds oflittle things that you can do
to you as a csm, you know, justhelp you operate more
efficiently, more smoothly withthe stuff.
I mean, I always tell peoplethink of the stuff that you
spend at least 20 minutes typing.
(13:54):
Like, yeah, if you're spending15, 20 minutes typing something
up, I guarantee you you can helpyourself reduce that amount of
time by at least 80% by usingchat, gpt or Gemini or whatever.
Speaker 2 (14:07):
I completely agree and I think, the ground game,
the ground game, I think, whereI felt the alleviation for me,
like my little bandaid, my skinis starting to crack with this
dry Austin weather Totally.
Speaker 1 (14:19):
We're all complaining .
It's like 30 degrees, it's socold.
Speaker 2 (14:22):
Yeah, especially for those of us who moved from
California.
I'm one of those, yeah, and so,yeah, I'm like you know it's
cold.
So um anyhow, I'm losing mytrain of thought.
What was I talking about?
Oh, the, the short unlock in CS, I mean the.
I think the email is is is thelowest, the lowest hanging fruit
(14:43):
on on how to help.
Now I will say that, ratherthan just hey, draft this email
for me, where I have found helpwith things like GPT is and I'm
a terrible prompter, In factjust the other day I had GPT
write me a prompt for another AIand that was kind of a cool
hack that I had never triedbefore.
(15:04):
But I was like duh, that makesa ton of sense.
Why don't you tell me what Ishould tell you?
Totally yeah.
Speaker 1 (15:12):
That's a good thing to do, yeah, yeah.
Speaker 2 (15:13):
So that was actually a little bit of a a little bit
of a help.
But on, on, say the emails.
You know, in CS we're we'retrying to sell value, we're
trying to convince folks tobelieve in us.
We need their buy-in forcertain workflows or workarounds
, like there is one, one part ofwhat we do in CS as kind of
instructing and telling andadvising.
(15:35):
but it has made so much betterwhen you can tap into the things
that are important for yourcustomer, when you can tap into
the I'll say the the lessonsfrom say that, how to influence
and influence people book rightthat that book changed my life
Right, and when I'm writingcertain emails that I know are
going to be difficult to receiveas a customer, because I can't
(15:57):
give them what they want or Ihave to introduce them to
something else that they weren'texpecting, I use GPT in that
way and I, when I ask it to help, I ask it to leverage that book
as an example of how to how towrite the content such that it's
not just well-written in goodgrammar but that it's tapping
(16:19):
into thinking in the otherperson's shoes.
I also have the benefit ofhaving a corporate GPT account,
so I don't have to mask any sortof customer names or anything
else like that.
So that's super helpful.
But if you don't have acorporate account, you can just
double click and mask out thenames, make it something else.
Speaker 1 (16:39):
I think that's an important call out, though,
because I don't I don't think alot of people realize that, and
that being, unless you've set upyour gpt account a specific way
or like you are a part of yourcorporate entities chat gpt
instance unless you've donethose things, you're training
the algorithm and you do notwant to train your algorithm on
(17:01):
your company's sensitiveinformation, which is
interesting because you canactually use that to your favor
as well.
In fact, at my previous gig,the marketing department did
this whole campaign of promptingusing some of the marketing
material and some of the kind offeature benefits and those
kinds of things, because theywanted to train OpenAI's
(17:24):
algorithms on the benefits andfeatures of the platform right,
so they kind of on purpose didthat.
But yeah, if you're using likeyou know, if you're prompting
and you're not kind of onlockdown a little bit, that can
be tricky.
But you know to your point.
I think what you bring up isinteresting because one of the
things that AI can help unlockfor us is what we would
(17:45):
otherwise spend hours and hourssearching our internal systems
for, which is to say, like youknow historical conversation
context, you know what thepurchase history of a customer
is, what their product usage is,and I think organizations that
are using artificialintelligence in a really
effective way are doing it in away that really all the primary
(18:09):
mission for artificialintelligence is the aggregation
of that information so that you,the CSM, or you, the support
person, or you, the whoever, canquickly get answers to or or or
a historical kind of summary ofthe account that you're kind of
working with Now, you knowvendors like Gainsight are
actively trying to do this inplatform right.
(18:32):
There's others who have done itin a homebrew kind of way by
using Zapier and all kinds ofother integration fun stuff.
But I think that's ultimatelyanother huge aspect of
artificial intelligence thatmost of our industry or, I guess
, practice, the practice of CSisn't really plugged into is the
(18:52):
just availability of data andinsights and information about
our customers.
Speaker 2 (18:57):
It is incredibly helpful, I'll be honest.
So we are a gain size shop atResponsive and they had AI
already plugged into NPS andbefore they we were starting to
use their co-pilot thing now,which is kind of interesting,
(19:20):
but it's still early but the NPSdata they were able to
illuminate the some clustering.
I think that's actually one ofthe coolest things that ai does,
which the funny thing is likewith how fast ai moves,
something that was sorevolutionary even just a few
months ago and now seems likealmost like old news or table
stakes to the people who arepaying attention to it.
(19:44):
And I fully appreciate the factthat, like, some people are
deep in it and some people are,like, like you said, not at all
using it.
So there's a wide variety offolks with different comfort
levels and knowledge of whatexists.
But clustering analysis with AIhas been a huge unlock, just in
terms of like aggregating ourNPS feedback, for example, with
(20:08):
thousands of data points and thecomments that people were
leaving.
They were able to group thoseinto different clusters, which
you know a data scientist coulddo anyway.
But I feel like it didsomething different and in a
unique way that allowed us toidentify a very like, clear and
present part of our platformthat we needed to work on.
(20:28):
That just hadn't gotten kind ofthe proper light shined on it
before, so it was very helpful.
And you know we use that in ourstrategic decision making for
what we're working on next, sothat was very helpful.
The Gainsight co pilot like Isaid, it's still early, but it
is reading the Gainsighttimeline so it is interesting
(20:52):
and so there's definitely I meanfolks also say that they use in
fact I think it was you whotold me about perplexity for
researching companies that areyour customer, it can be super
helpful.
There's certain let's say likeif I'm super helpful, there's
certain let's say like if I'mresearching, you know, let's say
amazon is my customer or somesomething like that, a large
(21:13):
company, and if I'm askingperplexity to do some research
on it, I may be deployed in onetiny business unit of amazon so
yeah, it can have its ownlimitations, but certainly,
certainly also can illuminatethings that you just didn't see
before Totally yeah, I mean,that's part of the magic is like
(21:33):
it'll tell you the obviousstuff, but then it'll also
illuminate some things that youdidn't think about.
Speaker 1 (21:38):
Like, the clustering analysis that you were talking
about is prime for that, where,yeah, you can look at your NPS
responses for that, you know touse that example on a
spreadsheet and you can readthrough them all and all that
kind of stuff.
But you're likely to miss somecommonalities here and there,
you know, unless you're likespending tons of time in the
(21:59):
data and reading every singleverb and noun right, whereas AI
can spit that stuff out for youin like a second.
So even you know again thatquick hack type situation, even
if you want to download all ofyour CSAT responses into a CSV
and put that into chat GPT andask it to do some pretty pointed
analysis on, you know, on thelanguage and things like that I
(22:23):
mean that's like.
You don't need know $200,000 CSPto do that for you.
You can just use chat GPT ifyou're on lockdown.
Speaker 2 (22:31):
No, it's smart.
In fact, I haven't.
I haven't even done that yet.
I have so many plans for what Iwant to do with unlocking the
black box of support in general,because there's so many
conversations that happen insideof there that I think can be
very informative to the businessin terms of what else needs to
be, it needs to be worked on.
That can help you bring thatthe data.
So analyzing the the commentsfrom Zendesk tickets is actually
(22:54):
a really good idea that Ididn't think of that.
I'm going to try and go do now,as I appreciate that.
Thank you, sure, um, the otherthing is so there's all sorts of
little, little tactical things.
The other thing is is thestrategy.
So you, you mentioned thatpeople want to talk to another
(23:14):
human.
Now I'm going to, I'm going tosteel man that for a second,
because, number one, I believein what you're saying, like, I
believe that people want to talk.
I believe people want to buildrelationships with other humans.
When it comes to talking,though, I have actually started
to use that conversational modelwith GPT, and the only thing I
(23:36):
mind about it is that I takelong pauses and it thinks that
I'm done talking, so I shouldtell it.
You can adjust that.
Speaker 1 (23:42):
Oh, really, dude?
Yeah, absolutely, because I useit on my walks a lot and I'll
just have a conversation withchat GPT on my, you know, when
I'm walking around theneighborhood or whatever.
And uh, I it kept like thinkingI was done and you can ask it
to basically adjust the pace.
Speaker 2 (23:58):
So just top tip, anyway, not to.
I love that.
I do that each time I'm aboutto talk to it.
Is that what you mean, or isthere like some setting?
Speaker 1 (24:07):
Yeah, no, it's like, but you can commit it to memory,
I believe.
Speaker 2 (24:11):
Okay.
Speaker 1 (24:11):
You can like literally ask it to commit it to
memory.
Speaker 2 (24:15):
Okay.
So this is another reason whyhaving a corporate account, for
example, is a very good idea,one I mean the IT team of your
company should embrace it withfull.
Should embrace it mostlybecause, if you don't, you're
depending on people to scrubthings out and not put stuff
into places where they shouldn't, which people are error prone.
You know we make mistakes, andso I'd rather have a place where
(24:38):
you know people can puteverything, absolutely
everything, and then store, likeyou said, the memory that
actually comes in really handy.
I asked, asked GPT the otherday.
You know what, what were someof the moments?
You know what were some of thebig accomplishments that we made
in 2024?
That's all I asked it, and itremembered everything that I had
told it over the last severalmonths, and it was very good at
(25:02):
remembering, you know, people wehired and decisions we made and
different directions that wewent, the impact that we had.
So I think that it makes thecase for not just using it for
tactical items, but the more youhave conversations in a
protected environment like acorporate account, the more you
have conversations on thestrategy level.
(25:24):
Two things actually come out ofthat.
One is that it's as if you'retalking to a coach and
anything's on the table, whichis actually very helpful,
especially in a leadershipposition where you know so much
of your time actually does needto be spent working on your
business instead of in yourbusiness.
Right, you have to be zoomingout, because it's so easy to get
(25:47):
into a groove, you know, into away of a way of working in a
pattern, and so I I recentlyused it.
You know, I I had, we made somechanges inside of the
leadership team, we brought insome new people from the outside
and we kind of we made a coupleof new director positions, and
I asked it to help me out withthe operational cadence of, like
(26:11):
, how do I?
You know, it used to be a muchsmaller team, but now we have a
larger team with these differentremits.
So I wanted it to help me thinkthrough, like, what is the best
use of everyone's time here,and how do I, how do we do the
things that we do, to make themost out of everyone's time and
to make sure that that we'rebeing both strategic and
(26:31):
tactical in the right ways.
And a lot of times, my, myprompts were kind of like that,
like very, very rough in thebeginning and just kind of work
through it, and that has been.
That has been a huge help.
And that's just me talking toyou know in air quotes somebody
that doesn't exist, but it'sgiven me a lot of help.
Speaker 1 (26:54):
I've done like probably things you shouldn't do
with chat GPT conversationalmode.
Okay, one of them.
So you know, let's see, I got myfirst kind of sponsors for the
podcast mid last year, somethinglike that, and in preparation
for that, basically I had chatGPT with some guidelines that I
(27:16):
gave it on a walk draft, asponsorship contract, and now
you know, I did not go straightto the vendor with that.
Like I went, I did send that tomy lawyer and just had him look
over it and he gave me a coupleof notes.
But again, you know, there werelike four or five contracts
that I wanted to put in placefor you know, sponsorship
(27:37):
agreements and some other thingstoo that I basically had
ChatGPT create for me, becauseyou know, I'm not launching
rockets over here, I'm justpublishing weekly podcasts.
It's not a big deal, but Iguess the point is it saved me A
, a ton of time and probably asubstantial amount of my
lawyer's time as well, justhaving it be redlined instead of
(27:59):
created from scratch.
So I think there's all kinds ofstuff that people, when OpenAI
launched the conversational kindof feature of ChatGPT, a lot of
people were like comparing itto Her, like the movie Her or
whatever.
Speaker 2 (28:15):
Oh, completely.
Speaker 1 (28:16):
And I mean I will say that it has become a bit of a
companion for me, but reallymore of a work companion,
somebody that I can you knowwhen I'm not doing anything else
can like help me talk throughsome things that I'm stuck on as
a leader, where I don'tnecessarily want to go to my CEO
and talk through and I don'tnecessarily want to go to my
(28:39):
management team and talk throughthis stuff.
I just want to like workshopthis stuff on my own.
And so having that conversationwith ChatGPT is an amazing way
to do that and not only kind ofbrain dump and get some stuff
off of your mind and into themodel, but then also it'll come
back with a few suggestions thatare quite good that you
(29:00):
wouldn't have come up withotherwise, nor would your team,
because they're all stuck in theday to day and they're not
thinking about things from anoutsider's kind of neutral
perspective.
So I'm with you a thousandpercent on using it as a leader
to help you unlock somestrategery.
Speaker 2 (29:18):
So I'm curious why Matic didn't nominate ChatGPT as
one of the most influential CSleaders of 2024.
Speaker 1 (29:24):
ChatGPT is one of the most influential CS leaders of
2024.
Speaker 2 (29:28):
Everyone was posting all their stuff recently and I
was like I actually think I'vegotten well, maybe not more help
, but I've certainly gotten alot of help from GPT on the
strategy behind things.
Speaker 1 (29:44):
It goes back to that whole thing of people are
fundamentally afraid that ai isgoing to replace them in their
roles.
I maybe I'm being a dinosaur.
I'm totally open to the idea orthe potential of me like being
(30:04):
like, yeah, compute, you know,typewriter is never going to be
phased out by a computer, but atthe same time, I do.
We're all human.
I think we all crave humaninteraction and we all rely on
human interaction.
And when it comes to you, youknow things like.
(30:26):
You know leadership and coachingand life and customer-vendor
relationships and coaching.
You know coaching, basketballand like whatever those things
are that we do as humans, thatwe enjoy doing.
That's not going anywhere,anywhere.
(30:48):
And so you know, if you're aleader trying to figure out the
agenda for your two-day off-site, if you're not using ChatGPT to
help you with that kind ofstuff, you know, I think you
might be left behind a littlebit, like it's not so much that
AI is going to replace us, but Ithink you might be replaced by
somebody who's using AI moreeffectively than you are, right,
(31:09):
I think that's sometimes thethings that go without saying
need to be said, and that's oneof them, which is that AI won't
likely there are things that itwill replace for the better.
Sure, like data analysis andlike all that kind of stuff,
sure.
Speaker 2 (31:23):
Yes, or even driving at some point.
Speaker 1 (31:27):
God, I like driving man.
Speaker 2 (31:28):
I do too.
But you know what I like moreNot having cars crash on the
road, and there's just noquestion that those things are
way better than us at driving Enmasse.
Speaker 1 (31:40):
Not me, of course.
I'm a way better driver thanthe.
Ai.
No, no, I'm not.
Speaker 2 (31:44):
But I think that's a huge benefit.
But, yeah, true so to yourpoint.
(32:09):
Yes, I'm not, but I think thatthat's a huge benefit.
But true, so to your point.
Yes, it's not likely that we'llall just lose our jobs to it,
but if it's like a, it's like awrench, you know, if you're a
plumber and and there's no shameIn using it, just like if you,
if you would, if you wouldwillingly use a calculator or
excel to do your math andformulas for you, you, you
should also be using a, even asjust like a sounding board.
Hey, here's my two-day plan foran off-site.
This is the goal that I have.
Do you think this is gonna helpme get there?
What worst case scenario.
(32:30):
It says you thought ofeverything right.
I can't think of a single thingto help you.
Best case scenario it's saying,like, actually, here are some
different things to thinkinstead so yeah, yeah, totally,
and it's a great, you know it's.
Speaker 1 (32:48):
It's fun to have it give you ideas for stuff like
yes, you know, yes it's just fun, all right.
Speaker 2 (32:57):
You mentioned earlier in the in the beginning, around
incorporating ai into your d,into your digital strategy, say,
for triggering out a letterafter onboarding.
What else, like what else is onyour mind in terms of how AI
and the digital strategy cometogether?
Like what even if it's notsomething you're doing yet, like
what are some of your ideas onhow those things could exist?
Speaker 1 (33:19):
Yeah, I'm in a playground right now where
that's a very real thing right,because I joined Belfry in
November and we are in the midstof just rapid growth and trying
to scale stuff, and so a bigpart of my daily brain activity
is around like how can we makethings more efficient and smooth
(33:41):
?
We are actively interviewingchatbot vendors, you know, to do
some of the deflection for usand even just some of the
frontline stuff we have.
We have a customer base who isvery in tune to their uh or or
very kind of tied to theirmobile device and so looking at
ways that even incorporatingthat into an sms strategy and
(34:03):
getting some ai answers that way, like and I and I'm thinking
about it in a in a brandy kindof way- you know, you know
there's not brandy, the drink orthe person, but literally
creating an avatar around ourartificial intelligence.
You know like?
Like you know there's, there's.
It's not a new idea, right.
Belly yeah, Belly yeah yeah,yeah, exactly, or like Bell or
(34:27):
something like that.
Speaker 2 (34:28):
I don't know We'll figure it out.
Speaker 1 (34:30):
But yeah, I mean, and then incorporating Bell into
like everything you know, likehaving it be kind of a brand.
That said, we're very much goingto launch that stuff internally
before we ever go customerfacing with it, and so I want to
launch it internally as a team,as a um, a team, you know tool,
rather than it's just to, youknow, kick the tires a little
(34:52):
bit more.
Um, yeah, that makes sense.
You know, the team has chat GPT, so they all have access to it.
They all use it on a regularbasis.
We're we're using gong very,very heavily and so, um, you
know things that it's doing forus.
So an example would be if wehave there's a, if there's a
customer meeting, it's recordedin Gong, and then what it
(35:13):
actually does because Gong has avery tight integration with
Vitally is a few minutes afterthe call ends, vitally will
actually give you somesuggestions for follow-up tasks
that you can schedule foryourself in the CSP, just based
on the conversation that itpicked up.
So it does.
You know, there's cool and likehelper things that we're doing,
(35:35):
and I have to say the one thingthat I want to be very careful
of as we continue to launch thiskind of stuff is I don't want
to over-rotate on those thingsto where it becomes a nuisance
and it becomes something thatthe team ends up hating.
Speaker 2 (35:51):
And like we don't?
Speaker 1 (35:53):
you know, I don't want to, I don't want to create
burden with any of this stuff,right?
So, um, I think one of thethings that's key for me as we,
as we continue to implement AIstrategies and tools and those
kinds of things, is just beingeyes wide open in terms of, you
know, are we actually helping orare we hurting the situation?
Speaker 2 (36:13):
And by hurting do you mean like you're creating, like
too many?
Speaker 1 (36:16):
to-dos or yeah, too many tasks, too many things
being auto-generated.
Like you know, I think there isthere's this interesting kind of
inflection point we're at whereyou could theoretically
automate the entire customerjourney and have it all be like
trigger driven and all this kindof stuff, and I think those
(36:40):
that have tried it have, moretimes than not, just failed
having a completely likeautonomous customer journey.
I think that extends into yourlet's take a csm like a typical
enterprise, csm 2030 accountskind of situation as well.
If you over automate the likeautomated messages that are
(37:02):
being sent on behalf of the csm,at that point you kind of lose
control of the narrative alittle bit too, and you're
potentially inserting some dirtor some noise into an otherwise
happy path customer right.
So I think the thing that Ipreach over and over again is
you have to insert yourself intoyour customer's journey on a
(37:23):
regular basis as a leader,because if you're not doing that
, you're not actively seeingwhat emails get triggered when
and what notifications go outwhen and all that kind of stuff.
Because the more stuff, themore automation you put in place
without checks and balances,the more likely you're just
confusing your customer withstuff coming at them all over
the place.
Speaker 2 (37:43):
Completely.
Yeah, all right, that's goingto be the mic drop intro of the
episode.
It's true.
I mean, you know, the utopiascenario is that everything is
everything is digital.
You know exactly what yourcustomer needs to do next.
You know exactly what valuethey're getting, you communicate
that in the platform, etc.
Etc.
That's the utopia that we're, Ithink, is appropriate to be
(38:05):
driving towards at all times,but is unrealistic that you're
going to get there and staythere.
Speaker 1 (38:11):
And it's also, but you know not to interject.
But I think it's also veryimportant that you insert a
human like stop button, likepicture, a manufacturing.
You know a lot, you know aproduction line or whatever.
If somebody gets their handjammed in a machine, you want a
stop button.
And same thing with a customerjourney.
That may be a little gross tothink about, but you know the
(38:33):
customer journey.
If, if things are going off therails and you're actively
working the situation, one ofthe last things that you want to
have happen is you're, like youknow, your pre-renewal email
flow going out like never, neverbeen guilty of that.
Speaker 2 (38:48):
No, it's never happened before.
Speaker 1 (38:50):
But like it's stupid stuff like that, where you know
it's worth just sitting down andthinking through at least some
of the scenarios you know andbuilding some fail-safes in
there.
Speaker 2 (39:03):
Yeah, and again, this is where I think the blend of
using AI, especially, at leastyou know, look, it's today, and
so today using AI is a mix oftactical and strategic, and so
not only asking or having youknow, having AI help you
generate emails or generateoutreaches or notifications in
(39:23):
the product, you can also haveit help you on exactly that kind
of a strategy of you know,making sure that there's
sufficient places where humanscan get involved, because, let's
be honest, it's pretty easy tosniff out, you know what has
been AI generated, and notespecially the longer it is, the
easier it's.
The easier it is to know thatif it's, if it's AI generated,
(39:44):
not to say that that's a badthing and not to say that that's
like anyone should really get agold star.
It's like, well, can you tellif that number was generated by
a calculator, like what does it?
Speaker 1 (39:55):
matter.
You know what I mean.
The number is the number.
Speaker 2 (39:56):
You know, the communication is the
communication, but and and thisis why I think you know, people
don't necessarily mindinteracting with ai, because
here we are doing it all daylong, I mean you and me, not
everybody but we're interactingwith this thing, but people
don't want to build arelationship with an AI.
(40:20):
That's the part where I think itgets to the her level.
Interacting is completelydifferent.
If I have to interact with anAI all the time for a software
that I'm using that I want toget more value out of, I don't
necessarily mind that, becauseit's always on.
It's going to be very relevant,the answers are going to be
good quality, it's going to bethorough, but I'm not going to
lunch with the AI, you know, andso yet the human yeah, the
(40:42):
human part is knowing that thereare other humans inside of that
organization who know about youand who are going to advocate
for you in a meeting orsomething like that, where
you're talking about, or whenthat company is talking about,
the roadmap or the strategy ofwhat they're working on.
Speaker 1 (40:59):
So maybe one day.
Speaker 2 (41:00):
AI will get to that as well, but I think, at least
for today, interactions, I think, are fine with AI, the
relationship building, though,like no one, no one is ready to
start dating their gpt.
Actually, I should not say thatI don't know anyone who is
willing to, but there totally is.
There are for sure.
There's a cohort of people outthere that are already oh, I'm
(41:22):
sure, yeah, absolutely well,alex, this has been really fun.
Thank you so much for spendingthe time with me on this episode
uh, the first ever episode ofmonday I monday.
Monday cs.
Monday a I cs.
Speaker 1 (41:36):
I don't know, I'm gonna have to work on that one
yeah, it's a work in progress,but it's a good work in progress
, true, I appreciate that.
Speaker 2 (41:42):
Thank you, and good chat.
Speaker 1 (41:44):
Appreciate you having me.
Thank you for joining me forthis episode of the Digital CX
Podcast.
If you like what we're doing,consider leaving us a review on
your podcast platform of choice.
If you're watching on YouTube,leave a comment down below.
It really helps us to grow andprovide value to a broader
audience and get moreinformation about the show and
some of the other things thatwe're doing at
(42:05):
digitalcustomersuccesscom.
I'm Alex Tergovich.
Thanks so much for listening.
We'll talk to you next week.
if you over automate the like automated messages that
are being sent on behalf of thecsm, at that point you kind of
lose control of the narrative.
Thing that I preach over andover again is you have to insert
yourself into your customer'sjourney on a regular basis as a
leader.
The more automation you put inplace without checks and
(00:23):
balances, the more likely you'rejust confusing your customer
with stuff coming at them allover the place.
Once again, welcome to theDigital Customer Experience
Podcast with me, alex Turkovich,so glad you could join us here
today and every week as weexplore how digital can help
enhance the customer andemployee experience.
My goal is to share what myguests and I have learned over
(00:45):
the years so that you can getthe insights that you need to
evolve your own digital programs.
If you'd like more info, needto get in touch or sign up for
the weekly companion newsletterthat has additional articles and
resources in it.
Go to digitalcustomersuccesscom.
For now, let's get started.
Hello and welcome back to theDigital CX podcast, the show
(01:05):
where we talk about all thingsdigital in CX.
I kind of mash those wordstogether, but you get what I
mean.
This is episode 89.
89 for you today, so I'mcreeping on towards 100 episodes
, which I'm super excited about,and have a couple surprises in
store for you for that episode.
Regardless, today is going tobe a little bit unconventional,
(01:29):
I guess, and we've done thisonce before where we've kind of
had two shows recorded at thesame time.
But my guest today is MichaelForney, who's a local Austinite.
He's a CS leader and he'sstarting his own show that is
really focused on AI and CS.
So he was kind enough to inviteme for his inaugural episode.
(01:51):
So we recorded that not longago, I think last week at some
point, and so I figured he'sreleased it on his stream.
I figured I would release it aswell on mine, because it's a
really cool conversation justabout our thoughts on current
state of AI in CS and, yeah, Ifigured you'd get a kick out of
(02:11):
the conversation as well.
So today, a shared episode withmyself and Michael Forney.
Just a quick housekeeping notebefore we get into the episode.
I will be speaking at thecustomer success summit event in
Austin February I think it'sthe 10th and the 11th, 11th,
12th, something like that.
(02:32):
I'll link the event details inthe show notes so that you can
go and have a look and if youwant to attend, I'd love to see
you there.
Without further ado, here's theepisode with Michael Forney.
Speaker 2 (02:47):
Alex, thank you so much for being on the first
episode of Monday I CS, where wetalk all things AI and CS.
Speaker 1 (02:56):
I was very curious to hear you pronounce Monday.
I.
Speaker 2 (03:02):
I was curious too.
That was the first time I hadvocalized it inside of a
microphone felt okay.
I think I can do better monday.
I think there's a.
There's an important pausebetween the end of the d and the
a, or you just call it monday Iwas thinking that too.
Yeah, but then I don't know.
Let's see what happens in thewild, let's see what other
(03:24):
people say.
Speaker 1 (03:24):
I can't wait to see where this journey takes you,
but it's great to be here.
Speaker 2 (03:28):
Thanks, thank you, dude.
Well, you've been very generous, you know, with your time and
you were just showing me areally interesting example.
Yeah, you leveraging Is it AIor is it robotics, or is it both
Bit of both?
Speaker 1 (03:40):
I guess All right.
Speaker 2 (03:41):
So tell me about this , or do you?
Do you not want to reveal?
Speaker 1 (03:44):
No, no, no no, yeah, that's fine.
Um, so I just uh, you know,joined a new company uh what?
Three months ago, november,something like that and uh, one
of the things we decided to dothis year was to send out, um
holiday cards, um, but wanted todo it handwritten.
(04:04):
So we actually used thisawesome service called
Handwritten with V-R-Y-T-T-E-N.
Speaker 2 (04:11):
So go look them up.
Speaker 1 (04:12):
I think it's just handwrittencom, where basically
all you do is you upload anaddress list and the message
that you want to send and theyhave a bunch of robots in the
background that do thehandwriting.
So I'll show you the, I'llblock my address out, but this
is kind of what it looks like onthe envelope, so it looks like
it's handwritten, and then youbasically design your card.
You can do it on the platformand then you upload it and they
(04:35):
put a little handwritten messagein there.
You can even like insert, youknow, booleans or tokens for,
like first name and all thatkind of stuff wow in the message
and they send it and so likewe're integrating that into our
tech stack too, to like sendhandwritten notes to our
customers when they graduate onboarding and all that kind of
stuff.
(04:55):
I think the ai part would wouldbe using some of the ai kind of
signals that we get to triggersome of those handwritten notes
and whatnot.
But super cool, super coolstuff because it combines that
human element with the machineelement.
Speaker 2 (05:10):
Yeah, completely, and you get to stand out in the
paper mail, which is actuallykind of a good way to market and
advertise these days, right,and it doesn't look like shit.
Speaker 1 (05:20):
Sorry, can I cuss on this show?
You can?
Do whatever you want GreatClothes are coming off.
No, it doesn't look like crap.
You know how you get thoseobviously fake handwritten.
It actually looks like somebodyput pen to paper, which is
bizarre.
This isn't a handwrittencommercial, by the way.
I promise you that, but there'sa list of 100 different writing
(05:44):
styles that you can choose from, so you can choose, like really
sloppy ones or really neat ones.
So it's cool.
Speaker 2 (05:49):
I think the fascinating part about that is
that it actually it's actuallywritten.
I mean, I've seen this kind ofthing before but it's printed
and then when you receive it ittotally takes away from it.
When you see the fact that thethe what should be handwriting
is is printed and it's like, atthat point I just kind of tune
out.
Speaker 1 (06:06):
So that's cool.
Speaker 2 (06:07):
Yeah, yeah, yeah, and I like that, yeah, I like that
you're thinking about doing thison the heels of, say,
onboarding, and so what I wasactually going to do to open
this show up is just to say youknow, hey, alex, when someone
says AI, what is coming to yourmind?
Like, when you think about AI,what is on your mind?
Speaker 1 (06:26):
Chaos and confusion.
Speaker 2 (06:28):
Chaos and confusion.
So you think we're still in,like the thrash of AI?
Yeah, I think so.
Speaker 1 (06:33):
I mean, look, you know, I think there's a lot of
just confusion about how to useit and I think, even a little
bit of guilt there's likeprofessional guilt involved.
If people, like, are using itless than others, you know, and
it's like you know I should beusing ai more than I am, or you
know they start shittingthemselves about it.
But, like, I think one of thethings that I am finding is we
(06:59):
have because the cost ofdevelopment is plummeting and we
now have all these you know,kind of startup companies that
are in the CS niche, that arekind of you know, developing
some use case centeredtechnology.
We have this kind of wide swathof different startups.
(07:19):
And then you know largercompanies, like your Gainsights
and your Tatangos and whatnot,that are all tackling AI and
integrating AI in some form orfashion, some as a kind of a
direct product and some as kindof like add-on to their direct
product, and so it's just themarket is just so wide and
desperate with like differentthings, and so I think what
(07:40):
we're going to start seeing islike a pretty rapid
consolidation of use cases andcompanies, you know, and a
swallowing up of differentvendors that are doing different
things, and hopefully that'lladd some clarity into, like, you
know, what platform should youlook for?
What technology should you lookfor?
What vendor do you need to youknow target?
(08:01):
Because you know I think rightnow it's just it's the wild,
wild west and it can getconfusing very, very fast and I
know a lot of people when theyget kind of overwhelmed and
confused with it they just kindof shut their laptop and walk
away.
Speaker 2 (08:16):
So I want to pick up on what you said about the guilt
.
Yeah, this is interesting, yeah.
So in many ways I look at AIand I sometimes I do have a
little bit of that sense ofguilt, but then I try to remind
myself.
You know, if you use acalculator today, nobody has
guilt about using a calculator,knowing that the actual math
(08:37):
behind everything would take youforever to do and if you're
like me, you probably justcouldn't even do it.
Speaker 1 (08:44):
Anyway, anyway, even if it's regular.
Speaker 2 (08:46):
It takes me an embarrassingly amount, a long
amount of time to do even justsimple addition and subtraction
so I I would I would be nowherewithout a calculator or excel so
so tell me more about what youmean by guilt with ai.
What's?
Speaker 1 (09:00):
yeah, I think we might be thinking about it in
slightly different ways, because, correct, no, it's fine.
This is important because Ithink you're thinking about
people who have guilt using AIinstead of doing it the hard way
or whatever.
I'm thinking that there's a lotof people out there who feel
guilty for not having adopted ityet, or they're not actively
(09:21):
using chat GPT as part of theirdaily life, or Gemini or
whatever you know you knowflavor you're, you're plugged
into or whatever, because thereare a lot of people who just
aren't you know,aren't using it on a regular
basis and I mean, um, you know,I think that's okay.
Yeah, I think it's okay.
(09:41):
Are you going to be better off,in the long run, knowing how to
prompt versus not knowing howto prompt?
I don't really think so,because I think prompting is a
short.
There's a finite window of timethat we're in right now where
prompting is the skill that youneed to have to effectively use
(10:01):
artificial intelligence.
But that prompting stuff is allgoing to go behind, you know,
behind the wall, so to speak.
Like I don't know.
I don't think 10, 15 years fromnow, you're going to need to be
real, real good at prompting,because, guess what?
Artificial intelligence will besentient and we'll all be out
(10:21):
of a job anyway.
So whatever.
Speaker 2 (10:24):
True, true, true, true, true.
I think about I mean, asexcited as I am about ai,
there's like at least a part ofme and anyone who grew up
watching the likes of terminator2.
It just feels like aninevitable future that we all
kind of put outside or at leastme, I put it outside my mind, my
mind, enjoying the rollercoaster until that happens, or
(10:46):
hopefully doesn't happen.
Speaker 1 (10:48):
Well, look, I mean, here's the deal.
We're all human and we'retalking about customer success,
right, and in customer success,humans are always going to want
to talk to other humans, yeah,right.
So if you're, you know, we'rein SAS, right?
So if we're speaking with acustomer or we're engaging with
a customer, exchanging actualwhat I like to call mouth words
(11:11):
with the customer is a very,very important thing, and I
don't think that's goinganywhere.
Helping you prep for thoseconversations, helping you have
those conversations, helping youlook up information in real
time while you're having thoseconversations so you can have
more productive ones and be moreefficient with your time, and
(11:31):
all that kind of stuff.
And I think that's what we'restarting to see, right, with the
chatbots and live agents andco-pilots and all that kind of
stuff that helps us have thoseconversations.
But also, like to our pointearlier, I don't think we've
quite gotten there yet, like wehaven't quite arrived at that
point yet.
It's still just like this thingwe have open chat, gpt or
(11:53):
whatever, and we ask it to dosome things for us, but it's not
like this live agent helpersituation so much to put to pull
on there.
Speaker 2 (12:03):
I think one, though one of the first things you had
said was about people maybefeeling guilty not using it.
So, in the spirit of giving youknow, give the cs world.
Let's give the cs world onepractical takeaway of like, what
is something, if people saywant to get in on the fun,
what's something that they cando or use to their advantage to
(12:26):
help them out that can be easilyadopted.
Speaker 1 (12:28):
Yeah, for sure, Well, I mean.
So there's a couple things andI did record an episode about
that specifically that you cango check out.
I'll give you the link to it.
It's basically an AI primer forCSMs and it gives you some
practical kind of use cases ofAI, also gives you a nice like
prompting framework that you canuse to kind of make your
(12:50):
prompts a little bit better andall that kind of stuff.
But I think, look, there's somesuper simple things that I think
a lot of tools are doing for usnow, like call transcripts and
things like that that I thinkare pretty easy to get at,
whether the tool that you'reusing to record your calls does
it or not, or whether youbasically feed a transcript into
(13:11):
ChatGPT to create a summary foryou.
I think that's a massive timesaver.
One little slight tweak to thatthat I would add is using a
tool like Gamma to create aquick PowerPoint deck as a
summary for your calls.
Like, I've thrown my AI callsummaries into Gamma, had it
create like a three or fourslide deck for me and sent the
(13:33):
deck out as a call recap, whichis kind of a cool thing because
it's a visual thing.
So I think there's all kinds oflittle things that you can do
to you as a csm, you know, justhelp you operate more
efficiently, more smoothly withthe stuff.
I mean, I always tell peoplethink of the stuff that you
spend at least 20 minutes typing.
(13:54):
Like, yeah, if you're spending15, 20 minutes typing something
up, I guarantee you you can helpyourself reduce that amount of
time by at least 80% by usingchat, gpt or Gemini or whatever.
Speaker 2 (14:07):
I completely agree and I think, the ground game,
the ground game, I think, whereI felt the alleviation for me,
like my little bandaid, my skinis starting to crack with this
dry Austin weather Totally.
Speaker 1 (14:19):
We're all complaining .
It's like 30 degrees, it's socold.
Speaker 2 (14:22):
Yeah, especially for those of us who moved from
California.
I'm one of those, yeah, and so,yeah, I'm like you know it's
cold.
So um anyhow, I'm losing mytrain of thought.
What was I talking about?
Oh, the, the short unlock in CS, I mean the.
I think the email is is is thelowest, the lowest hanging fruit
(14:43):
on on how to help.
Now I will say that, ratherthan just hey, draft this email
for me, where I have found helpwith things like GPT is and I'm
a terrible prompter, In factjust the other day I had GPT
write me a prompt for another AIand that was kind of a cool
hack that I had never triedbefore.
(15:04):
But I was like duh, that makesa ton of sense.
Why don't you tell me what Ishould tell you?
Totally yeah.
Speaker 1 (15:12):
That's a good thing to do, yeah, yeah.
Speaker 2 (15:13):
So that was actually a little bit of a a little bit
of a help.
But on, on, say the emails.
You know, in CS we're we'retrying to sell value, we're
trying to convince folks tobelieve in us.
We need their buy-in forcertain workflows or workarounds
, like there is one, one part ofwhat we do in CS as kind of
instructing and telling andadvising.
(15:35):
but it has made so much betterwhen you can tap into the things
that are important for yourcustomer, when you can tap into
the I'll say the the lessonsfrom say that, how to influence
and influence people book rightthat that book changed my life
Right, and when I'm writingcertain emails that I know are
going to be difficult to receiveas a customer, because I can't
(15:57):
give them what they want or Ihave to introduce them to
something else that they weren'texpecting, I use GPT in that
way and I, when I ask it to help, I ask it to leverage that book
as an example of how to how towrite the content such that it's
not just well-written in goodgrammar but that it's tapping
(16:19):
into thinking in the otherperson's shoes.
I also have the benefit ofhaving a corporate GPT account,
so I don't have to mask any sortof customer names or anything
else like that.
So that's super helpful.
But if you don't have acorporate account, you can just
double click and mask out thenames, make it something else.
Speaker 1 (16:39):
I think that's an important call out, though,
because I don't I don't think alot of people realize that, and
that being, unless you've set upyour gpt account a specific way
or like you are a part of yourcorporate entities chat gpt
instance unless you've donethose things, you're training
the algorithm and you do notwant to train your algorithm on
(17:01):
your company's sensitiveinformation, which is
interesting because you canactually use that to your favor
as well.
In fact, at my previous gig,the marketing department did
this whole campaign of promptingusing some of the marketing
material and some of the kind offeature benefits and those
kinds of things, because theywanted to train OpenAI's
(17:24):
algorithms on the benefits andfeatures of the platform right,
so they kind of on purpose didthat.
But yeah, if you're using likeyou know, if you're prompting
and you're not kind of onlockdown a little bit, that can
be tricky.
But you know to your point.
I think what you bring up isinteresting because one of the
things that AI can help unlockfor us is what we would
(17:45):
otherwise spend hours and hourssearching our internal systems
for, which is to say, like youknow historical conversation
context, you know what thepurchase history of a customer
is, what their product usage is,and I think organizations that
are using artificialintelligence in a really
effective way are doing it in away that really all the primary
(18:09):
mission for artificialintelligence is the aggregation
of that information so that you,the CSM, or you, the support
person, or you, the whoever, canquickly get answers to or or or
a historical kind of summary ofthe account that you're kind of
working with Now, you knowvendors like Gainsight are
actively trying to do this inplatform right.
(18:32):
There's others who have done itin a homebrew kind of way by
using Zapier and all kinds ofother integration fun stuff.
But I think that's ultimatelyanother huge aspect of
artificial intelligence thatmost of our industry or, I guess
, practice, the practice of CSisn't really plugged into is the
(18:52):
just availability of data andinsights and information about
our customers.
Speaker 2 (18:57):
It is incredibly helpful, I'll be honest.
So we are a gain size shop atResponsive and they had AI
already plugged into NPS andbefore they we were starting to
use their co-pilot thing now,which is kind of interesting,
(19:20):
but it's still early but the NPSdata they were able to
illuminate the some clustering.
I think that's actually one ofthe coolest things that ai does,
which the funny thing is likewith how fast ai moves,
something that was sorevolutionary even just a few
months ago and now seems likealmost like old news or table
stakes to the people who arepaying attention to it.
(19:44):
And I fully appreciate the factthat, like, some people are
deep in it and some people are,like, like you said, not at all
using it.
So there's a wide variety offolks with different comfort
levels and knowledge of whatexists.
But clustering analysis with AIhas been a huge unlock, just in
terms of like aggregating ourNPS feedback, for example, with
(20:08):
thousands of data points and thecomments that people were
leaving.
They were able to group thoseinto different clusters, which
you know a data scientist coulddo anyway.
But I feel like it didsomething different and in a
unique way that allowed us toidentify a very like, clear and
present part of our platformthat we needed to work on.
(20:28):
That just hadn't gotten kind ofthe proper light shined on it
before, so it was very helpful.
And you know we use that in ourstrategic decision making for
what we're working on next, sothat was very helpful.
The Gainsight co pilot like Isaid, it's still early, but it
is reading the Gainsighttimeline so it is interesting
(20:52):
and so there's definitely I meanfolks also say that they use in
fact I think it was you whotold me about perplexity for
researching companies that areyour customer, it can be super
helpful.
There's certain let's say likeif I'm super helpful, there's
certain let's say like if I'mresearching, you know, let's say
amazon is my customer or somesomething like that, a large
(21:13):
company, and if I'm askingperplexity to do some research
on it, I may be deployed in onetiny business unit of amazon so
yeah, it can have its ownlimitations, but certainly,
certainly also can illuminatethings that you just didn't see
before Totally yeah, I mean,that's part of the magic is like
(21:33):
it'll tell you the obviousstuff, but then it'll also
illuminate some things that youdidn't think about.
Speaker 1 (21:38):
Like, the clustering analysis that you were talking
about is prime for that, where,yeah, you can look at your NPS
responses for that, you know touse that example on a
spreadsheet and you can readthrough them all and all that
kind of stuff.
But you're likely to miss somecommonalities here and there,
you know, unless you're likespending tons of time in the
(21:59):
data and reading every singleverb and noun right, whereas AI
can spit that stuff out for youin like a second.
So even you know again thatquick hack type situation, even
if you want to download all ofyour CSAT responses into a CSV
and put that into chat GPT andask it to do some pretty pointed
analysis on, you know, on thelanguage and things like that I
(22:23):
mean that's like.
You don't need know $200,000 CSPto do that for you.
You can just use chat GPT ifyou're on lockdown.
Speaker 2 (22:31):
No, it's smart.
In fact, I haven't.
I haven't even done that yet.
I have so many plans for what Iwant to do with unlocking the
black box of support in general,because there's so many
conversations that happen insideof there that I think can be
very informative to the businessin terms of what else needs to
be, it needs to be worked on.
That can help you bring thatthe data.
So analyzing the the commentsfrom Zendesk tickets is actually
(22:54):
a really good idea that Ididn't think of that.
I'm going to try and go do now,as I appreciate that.
Thank you, sure, um, the otherthing is so there's all sorts of
little, little tactical things.
The other thing is is thestrategy.
So you, you mentioned thatpeople want to talk to another
(23:14):
human.
Now I'm going to, I'm going tosteel man that for a second,
because, number one, I believein what you're saying, like, I
believe that people want to talk.
I believe people want to buildrelationships with other humans.
When it comes to talking,though, I have actually started
to use that conversational modelwith GPT, and the only thing I
(23:36):
mind about it is that I takelong pauses and it thinks that
I'm done talking, so I shouldtell it.
You can adjust that.
Speaker 1 (23:42):
Oh, really, dude?
Yeah, absolutely, because I useit on my walks a lot and I'll
just have a conversation withchat GPT on my, you know, when
I'm walking around theneighborhood or whatever.
And uh, I it kept like thinkingI was done and you can ask it
to basically adjust the pace.
Speaker 2 (23:58):
So just top tip, anyway, not to.
I love that.
I do that each time I'm aboutto talk to it.
Is that what you mean, or isthere like some setting?
Speaker 1 (24:07):
Yeah, no, it's like, but you can commit it to memory,
I believe.
Speaker 2 (24:11):
Okay.
Speaker 1 (24:11):
You can like literally ask it to commit it to
memory.
Speaker 2 (24:15):
Okay.
So this is another reason whyhaving a corporate account, for
example, is a very good idea,one I mean the IT team of your
company should embrace it withfull.
Should embrace it mostlybecause, if you don't, you're
depending on people to scrubthings out and not put stuff
into places where they shouldn't, which people are error prone.
You know we make mistakes, andso I'd rather have a place where
(24:38):
you know people can puteverything, absolutely
everything, and then store, likeyou said, the memory that
actually comes in really handy.
I asked, asked GPT the otherday.
You know what, what were someof the moments?
You know what were some of thebig accomplishments that we made
in 2024?
That's all I asked it, and itremembered everything that I had
told it over the last severalmonths, and it was very good at
(25:02):
remembering, you know, people wehired and decisions we made and
different directions that wewent, the impact that we had.
So I think that it makes thecase for not just using it for
tactical items, but the more youhave conversations in a
protected environment like acorporate account, the more you
have conversations on thestrategy level.
(25:24):
Two things actually come out ofthat.
One is that it's as if you'retalking to a coach and
anything's on the table, whichis actually very helpful,
especially in a leadershipposition where you know so much
of your time actually does needto be spent working on your
business instead of in yourbusiness.
Right, you have to be zoomingout, because it's so easy to get
(25:47):
into a groove, you know, into away of a way of working in a
pattern, and so I I recentlyused it.
You know, I I had, we made somechanges inside of the
leadership team, we brought insome new people from the outside
and we kind of we made a coupleof new director positions, and
I asked it to help me out withthe operational cadence of, like
(26:11):
, how do I?
You know, it used to be a muchsmaller team, but now we have a
larger team with these differentremits.
So I wanted it to help me thinkthrough, like, what is the best
use of everyone's time here,and how do I, how do we do the
things that we do, to make themost out of everyone's time and
to make sure that that we'rebeing both strategic and
(26:31):
tactical in the right ways.
And a lot of times, my, myprompts were kind of like that,
like very, very rough in thebeginning and just kind of work
through it, and that has been.
That has been a huge help.
And that's just me talking toyou know in air quotes somebody
that doesn't exist, but it'sgiven me a lot of help.
Speaker 1 (26:54):
I've done like probably things you shouldn't do
with chat GPT conversationalmode.
Okay, one of them.
So you know, let's see, I got myfirst kind of sponsors for the
podcast mid last year, somethinglike that, and in preparation
for that, basically I had chatGPT with some guidelines that I
(27:16):
gave it on a walk draft, asponsorship contract, and now
you know, I did not go straightto the vendor with that.
Like I went, I did send that tomy lawyer and just had him look
over it and he gave me a coupleof notes.
But again, you know, there werelike four or five contracts
that I wanted to put in placefor you know, sponsorship
(27:37):
agreements and some other thingstoo that I basically had
ChatGPT create for me, becauseyou know, I'm not launching
rockets over here, I'm justpublishing weekly podcasts.
It's not a big deal, but Iguess the point is it saved me A
, a ton of time and probably asubstantial amount of my
lawyer's time as well, justhaving it be redlined instead of
(27:59):
created from scratch.
So I think there's all kinds ofstuff that people, when OpenAI
launched the conversational kindof feature of ChatGPT, a lot of
people were like comparing itto Her, like the movie Her or
whatever.
Speaker 2 (28:15):
Oh, completely.
Speaker 1 (28:16):
And I mean I will say that it has become a bit of a
companion for me, but reallymore of a work companion,
somebody that I can you knowwhen I'm not doing anything else
can like help me talk throughsome things that I'm stuck on as
a leader, where I don'tnecessarily want to go to my CEO
and talk through and I don'tnecessarily want to go to my
(28:39):
management team and talk throughthis stuff.
I just want to like workshopthis stuff on my own.
And so having that conversationwith ChatGPT is an amazing way
to do that and not only kind ofbrain dump and get some stuff
off of your mind and into themodel, but then also it'll come
back with a few suggestions thatare quite good that you
(29:00):
wouldn't have come up withotherwise, nor would your team,
because they're all stuck in theday to day and they're not
thinking about things from anoutsider's kind of neutral
perspective.
So I'm with you a thousandpercent on using it as a leader
to help you unlock somestrategery.
Speaker 2 (29:18):
So I'm curious why Matic didn't nominate ChatGPT as
one of the most influential CSleaders of 2024.
Speaker 1 (29:24):
ChatGPT is one of the most influential CS leaders of
2024.
Speaker 2 (29:28):
Everyone was posting all their stuff recently and I
was like I actually think I'vegotten well, maybe not more help
, but I've certainly gotten alot of help from GPT on the
strategy behind things.
Speaker 1 (29:44):
It goes back to that whole thing of people are
fundamentally afraid that ai isgoing to replace them in their
roles.
I maybe I'm being a dinosaur.
I'm totally open to the idea orthe potential of me like being
(30:04):
like, yeah, compute, you know,typewriter is never going to be
phased out by a computer, but atthe same time, I do.
We're all human.
I think we all crave humaninteraction and we all rely on
human interaction.
And when it comes to you, youknow things like.
(30:26):
You know leadership and coachingand life and customer-vendor
relationships and coaching.
You know coaching, basketballand like whatever those things
are that we do as humans, thatwe enjoy doing.
That's not going anywhere,anywhere.
(30:48):
And so you know, if you're aleader trying to figure out the
agenda for your two-day off-site, if you're not using ChatGPT to
help you with that kind ofstuff, you know, I think you
might be left behind a littlebit, like it's not so much that
AI is going to replace us, but Ithink you might be replaced by
somebody who's using AI moreeffectively than you are, right,
(31:09):
I think that's sometimes thethings that go without saying
need to be said, and that's oneof them, which is that AI won't
likely there are things that itwill replace for the better.
Sure, like data analysis andlike all that kind of stuff,
sure.
Speaker 2 (31:23):
Yes, or even driving at some point.
Speaker 1 (31:27):
God, I like driving man.
Speaker 2 (31:28):
I do too.
But you know what I like moreNot having cars crash on the
road, and there's just noquestion that those things are
way better than us at driving Enmasse.
Speaker 1 (31:40):
Not me, of course.
I'm a way better driver thanthe.
Ai.
No, no, I'm not.
Speaker 2 (31:44):
But I think that's a huge benefit.
But, yeah, true so to yourpoint.
(32:09):
Yes, I'm not, but I think thatthat's a huge benefit.
But true, so to your point.
Yes, it's not likely that we'llall just lose our jobs to it,
but if it's like a, it's like awrench, you know, if you're a
plumber and and there's no shameIn using it, just like if you,
if you would, if you wouldwillingly use a calculator or
excel to do your math andformulas for you, you, you
should also be using a, even asjust like a sounding board.
Hey, here's my two-day plan foran off-site.
This is the goal that I have.
Do you think this is gonna helpme get there?
What worst case scenario.
(32:30):
It says you thought ofeverything right.
I can't think of a single thingto help you.
Best case scenario it's saying,like, actually, here are some
different things to thinkinstead so yeah, yeah, totally,
and it's a great, you know it's.
Speaker 1 (32:48):
It's fun to have it give you ideas for stuff like
yes, you know, yes it's just fun, all right.
Speaker 2 (32:57):
You mentioned earlier in the in the beginning, around
incorporating ai into your d,into your digital strategy, say,
for triggering out a letterafter onboarding.
What else, like what else is onyour mind in terms of how AI
and the digital strategy cometogether?
Like what even if it's notsomething you're doing yet, like
what are some of your ideas onhow those things could exist?
Speaker 1 (33:19):
Yeah, I'm in a playground right now where
that's a very real thing right,because I joined Belfry in
November and we are in the midstof just rapid growth and trying
to scale stuff, and so a bigpart of my daily brain activity
is around like how can we makethings more efficient and smooth
(33:41):
?
We are actively interviewingchatbot vendors, you know, to do
some of the deflection for usand even just some of the
frontline stuff we have.
We have a customer base who isvery in tune to their uh or or
very kind of tied to theirmobile device and so looking at
ways that even incorporatingthat into an sms strategy and
(34:03):
getting some ai answers that way, like and I and I'm thinking
about it in a in a brandy kindof way- you know, you know
there's not brandy, the drink orthe person, but literally
creating an avatar around ourartificial intelligence.
You know like?
Like you know there's, there's.
It's not a new idea, right.
Belly yeah, Belly yeah yeah,yeah, exactly, or like Bell or
(34:27):
something like that.
Speaker 2 (34:28):
I don't know We'll figure it out.
Speaker 1 (34:30):
But yeah, I mean, and then incorporating Bell into
like everything you know, likehaving it be kind of a brand.
That said, we're very much goingto launch that stuff internally
before we ever go customerfacing with it, and so I want to
launch it internally as a team,as a um, a team, you know tool,
rather than it's just to, youknow, kick the tires a little
(34:52):
bit more.
Um, yeah, that makes sense.
You know, the team has chat GPT, so they all have access to it.
They all use it on a regularbasis.
We're we're using gong very,very heavily and so, um, you
know things that it's doing forus.
So an example would be if wehave there's a, if there's a
customer meeting, it's recordedin Gong, and then what it
(35:13):
actually does because Gong has avery tight integration with
Vitally is a few minutes afterthe call ends, vitally will
actually give you somesuggestions for follow-up tasks
that you can schedule foryourself in the CSP, just based
on the conversation that itpicked up.
So it does.
You know, there's cool and likehelper things that we're doing,
(35:35):
and I have to say the one thingthat I want to be very careful
of as we continue to launch thiskind of stuff is I don't want
to over-rotate on those thingsto where it becomes a nuisance
and it becomes something thatthe team ends up hating.
Speaker 2 (35:51):
And like we don't?
Speaker 1 (35:53):
you know, I don't want to, I don't want to create
burden with any of this stuff,right?
So, um, I think one of thethings that's key for me as we,
as we continue to implement AIstrategies and tools and those
kinds of things, is just beingeyes wide open in terms of, you
know, are we actually helping orare we hurting the situation?
Speaker 2 (36:13):
And by hurting do you mean like you're creating, like
too many?
Speaker 1 (36:16):
to-dos or yeah, too many tasks, too many things
being auto-generated.
Like you know, I think there isthere's this interesting kind of
inflection point we're at whereyou could theoretically
automate the entire customerjourney and have it all be like
trigger driven and all this kindof stuff, and I think those
(36:40):
that have tried it have, moretimes than not, just failed
having a completely likeautonomous customer journey.
I think that extends into yourlet's take a csm like a typical
enterprise, csm 2030 accountskind of situation as well.
If you over automate the likeautomated messages that are
(37:02):
being sent on behalf of the csm,at that point you kind of lose
control of the narrative alittle bit too, and you're
potentially inserting some dirtor some noise into an otherwise
happy path customer right.
So I think the thing that Ipreach over and over again is
you have to insert yourself intoyour customer's journey on a
(37:23):
regular basis as a leader,because if you're not doing that
, you're not actively seeingwhat emails get triggered when
and what notifications go outwhen and all that kind of stuff.
Because the more stuff, themore automation you put in place
without checks and balances,the more likely you're just
confusing your customer withstuff coming at them all over
the place.
Speaker 2 (37:43):
Completely.
Yeah, all right, that's goingto be the mic drop intro of the
episode.
It's true.
I mean, you know, the utopiascenario is that everything is
everything is digital.
You know exactly what yourcustomer needs to do next.
You know exactly what valuethey're getting, you communicate
that in the platform, etc.
Etc.
That's the utopia that we're, Ithink, is appropriate to be
(38:05):
driving towards at all times,but is unrealistic that you're
going to get there and staythere.
Speaker 1 (38:11):
And it's also, but you know not to interject.
But I think it's also veryimportant that you insert a
human like stop button, likepicture, a manufacturing.
You know a lot, you know aproduction line or whatever.
If somebody gets their handjammed in a machine, you want a
stop button.
And same thing with a customerjourney.
That may be a little gross tothink about, but you know the
(38:33):
customer journey.
If, if things are going off therails and you're actively
working the situation, one ofthe last things that you want to
have happen is you're, like youknow, your pre-renewal email
flow going out like never, neverbeen guilty of that.
Speaker 2 (38:48):
No, it's never happened before.
Speaker 1 (38:50):
But like it's stupid stuff like that, where you know
it's worth just sitting down andthinking through at least some
of the scenarios you know andbuilding some fail-safes in
there.
Speaker 2 (39:03):
Yeah, and again, this is where I think the blend of
using AI, especially, at leastyou know, look, it's today, and
so today using AI is a mix oftactical and strategic, and so
not only asking or having youknow, having AI help you
generate emails or generateoutreaches or notifications in
(39:23):
the product, you can also haveit help you on exactly that kind
of a strategy of you know,making sure that there's
sufficient places where humanscan get involved, because, let's
be honest, it's pretty easy tosniff out, you know what has
been AI generated, and notespecially the longer it is, the
easier it's.
The easier it is to know thatif it's, if it's AI generated,
(39:44):
not to say that that's a badthing and not to say that that's
like anyone should really get agold star.
It's like, well, can you tellif that number was generated by
a calculator, like what does it?
Speaker 1 (39:55):
matter.
You know what I mean.
The number is the number.
Speaker 2 (39:56):
You know, the communication is the
communication, but and and thisis why I think you know, people
don't necessarily mindinteracting with ai, because
here we are doing it all daylong, I mean you and me, not
everybody but we're interactingwith this thing, but people
don't want to build arelationship with an AI.
(40:20):
That's the part where I think itgets to the her level.
Interacting is completelydifferent.
If I have to interact with anAI all the time for a software
that I'm using that I want toget more value out of, I don't
necessarily mind that, becauseit's always on.
It's going to be very relevant,the answers are going to be
good quality, it's going to bethorough, but I'm not going to
lunch with the AI, you know, andso yet the human yeah, the
(40:42):
human part is knowing that thereare other humans inside of that
organization who know about youand who are going to advocate
for you in a meeting orsomething like that, where
you're talking about, or whenthat company is talking about,
the roadmap or the strategy ofwhat they're working on.
Speaker 1 (40:59):
So maybe one day.
Speaker 2 (41:00):
AI will get to that as well, but I think, at least
for today, interactions, I think, are fine with AI, the
relationship building, though,like no one, no one is ready to
start dating their gpt.
Actually, I should not say thatI don't know anyone who is
willing to, but there totally is.
There are for sure.
There's a cohort of people outthere that are already oh, I'm
(41:22):
sure, yeah, absolutely well,alex, this has been really fun.
Thank you so much for spendingthe time with me on this episode
uh, the first ever episode ofmonday I monday.
Monday cs.
Monday a I cs.
Speaker 1 (41:36):
I don't know, I'm gonna have to work on that one
yeah, it's a work in progress,but it's a good work in progress
, true, I appreciate that.
Speaker 2 (41:42):
Thank you, and good chat.
Speaker 1 (41:44):
Appreciate you having me.
Thank you for joining me forthis episode of the Digital CX
Podcast.
If you like what we're doing,consider leaving us a review on
your podcast platform of choice.
If you're watching on YouTube,leave a comment down below.
It really helps us to grow andprovide value to a broader
audience and get moreinformation about the show and
some of the other things thatwe're doing at
(42:05):
digitalcustomersuccesscom.
I'm Alex Tergovich.
Thanks so much for listening.
We'll talk to you next week.
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.