The Future of Retail is Here: A Deep Dive into POS, AI, and Running a Successful Agency

Today, I spoke with Jared Smith, CEO of Kitestring Technical Services in Bentonville, Arkansas. Jared, who has a PhD in physical chemistry from UC Berkeley, has led Kitestring, a 150-person firm specializing in retail technology, for the past 12 years. Our discussion ranged from the evolution of point-of-sale (POS) systems and the potential of AI in retail to the importance of self-care for entrepreneurs.

Evolution of POS Technology
We started by discussing the evolution of POS systems, which handle transactions, inventory management, and customer engagement. Jared’s company has a strong history in POS development. We reminisced about early Scan and Go technology at Walmart and its improvements. Jared believes the future of in-store checkout will involve both self-checkout kiosks and traditional cashier lanes, providing a seamless customer experience. He highlighted technologies like RFID and computer vision that could revolutionize checkout processes, allowing customers to walk through a checkout area while the system automatically identifies and rings up each item.

Challenges remain, such as RFID signals being disrupted by liquids and metals and computer vision struggling with obscured items in a cart. Jared suggested a hybrid approach, combining these technologies with sophisticated algorithms, could overcome these issues.

Omnichannel Retail and Blurring Lines of Commerce
Our conversation then shifted to omnichannel retail, merging online and offline shopping experiences, such as buying online and picking up in-store (BOPUS). Jared mentioned Walmart’s early BOPUS experiments, like the Alaskan Bush Pilot Project in the 90s, which predated widespread internet use.

Integrating online and offline channels poses unique challenges in inventory management. Questions arise about when an online sale becomes a store sale and how such transactions impact inventory tracking and sales figures. Retailers must navigate these complexities as they adapt to the evolving landscape.

The Promise and Peril of AI in Retail
We discussed AI’s potential to transform retail, offering personalized shopping experiences and optimized supply chains. However, AI is also surrounded by hype and uncertainty. Jared stressed the importance of caution when implementing AI, especially in mission-critical systems like POS. Retailers should avoid jumping on AI trends without thorough evaluation, as untested technologies can be disruptive and costly.

Nonetheless, AI can enhance operational efficiency. Jared shared a study where consultants using AI tools saw increased productivity and quality. While general-purpose AI tools like ChatGPT are popular, Jared sees more value in specialized AI models trained on specific industry datasets. For instance, an AI assistant trained on a retailer’s historical sales data could provide valuable insights for demand forecasting, pricing optimization, and personalized marketing.

Importance of Self-Care for Leaders
As our conversation neared its end, we discussed the importance of self-care for entrepreneurs. Running a business with 150 employees can be incredibly stressful. Jared and I bonded over our commitment to physical and mental well-being, frequenting Blake Street House, a gym where we engage in cardio, strength training, heat therapy, and cold plunges.

These practices profoundly impact mental clarity, energy levels, and overall well-being. Jared emphasized the need to prioritize sleep, exercise, and a healthy diet to manage stress and make sound decisions. He believes that taking care of himself makes him a better leader, enhancing his ability to navigate the complexities of running a successful business.

See more about my discussion with Jared Smith over on our website

Episode Transcript

Dr. Matt Waller (00:10):
Welcome to the Matt Waller podcast, where we
look at success at theintersection of technology,
logistics, supply chain, retailand CPG, also known as the
retail value chain.
I want to clarify that thispodcast is distinct from my
responsibilities as a professorin the Sam M Walton College of
Business.
Nonetheless, it aligns with myaspiration to provide practical

(00:31):
insights to professionals andbusiness by showcasing companies
and people that can enhanceyour ability to manage, lead and
strategize and marketeffectively in the retail value
chain.
Before we dive into today'sexciting episode, I'd like to
thank our sponsor, new RoadCapital Partners.
New Road invests in proventechnologies, services and

(00:53):
products that serve unmet needsin the marketplace.
They look for companies insupply chain and logistics, as
well as consumer-orientedcompanies.
For more information, go tonewroadcpcom.
I would also like to disclosethat I am a strategic advisor to
New Road.
I'd also like to recognizepodcastvideoscom for the

(01:18):
services they provide for thesepodcasts.
I'm very pleased with theirservices and now, without
further ado, let's get into theexciting episode.
I have with me today Jared Smith, who is the CEO of KiteString
Technical Services inBentonville, arkansas.
He's an expert on POS systemsand payment systems and his

(01:43):
company is very interesting.
He's been the CEO for 12 years.
Earlier in his career, he got aPhD in physical chemistry from
the University of California,berkeley, but he now runs an
agency of 150 people here inNorthwest Arkansas and it's a

(02:06):
very successful professionalservices firm.
I think you're really going tolike this discussion.
We actually talk a lot about POS, so if you're interested in
retail or technology at all, Ithink you'll be interested.
We talk a little bit about thehistory of retail and the future
, especially from a POSperspective, and a little bit

(02:27):
about omni-channel, and we talkabout AI.
Of course, every conversationhas to include AI to some degree
, but we even talk about thingslike certain aspects of running
a professional services firmwhat do you need to be good at,
what do you need to do and wetalk a little bit also about

(02:51):
strategy as well, and so I thinkyou're really going to enjoy
this.
Thank you, jared.
Thank you so much for joiningme today.
Thanks, matt.
I'm glad to be here.
Jared, I wanted to start byjust talking a little bit about
your firm, kitestring.
What do you do, how do you work, those kind of things.

(03:15):
But first just describe alittle bit about your company.

Jared Smith (03:18):
All right.
Well, kitestring TechnicalServices is our full name.
Everybody just calls usKiteString.
We are a technology consultingfirm specializing in retail
technology.
We're 25 years old, about 150people or so strong.
We have a few different linesof business.
So part of our business isfocused on talent solutions.

(03:40):
I can get into the details ofthat, part of it is strategic
consulting and then part of itis custom software development.
Those are kind of the threethings we do, all of it focused
in the retail space.
So our main historical businesshas been point of sale software
.
We've done a lot forintegration and customization of
existing point of sale softwareand also building you know

(04:02):
Greenfield brand new point ofsale software of existing
point-of-sale software and alsobuilding you know Greenfield
brand new point-of-sale software.
So point-of-sale paymentsystems, e-commerce, asset
protection.
A lot of back office softwareis what we've worked on Pharmacy
.
Another big one for us isretail pharmacy.

Dr. Matt Waller (04:15):
Regarding those three areas of your business,
which one did you start in?

Jared Smith (04:20):
We started well, you know kind of, we sort of
fell into it.
Actually, my father started thecompany in 1998.
He had been even when I was injunior high although junior high
and high school he would workin point-of-sale.
He actually built a custompoint-of-sale solution and sold
it to very small independentgrocers I think a cruise company

(04:44):
out of Hawaii, so smallcompanies like this were using
his solution.
But it wasn't really taking offor anything and he took a
contract with NCR or thenultimately later IBM to work on
Walmart's point-of-sale systemin the late 90s and incorporated
the company and that was kindof a route there and he was.

(05:06):
He spent years as anindependent contractor working
under either an ncr, I think ncrin the beginning, and then
later ibm, who owned the 4690operating system that was later
sold to toshiba.
So now you think of toshibathat used to be ibm's retail
division that owned the 4690operating system, which is still
, I think today, the biggestoperating system for gas
registers.

(05:26):
So he was doing that for maybetwo or three years on his own
and then they needed anotherperson to help some big um I
don't know exactly what theywere doing back then to um I, I
believe what they does purchasethe software from toshiba or
from ibm at the time, and theywere going to revamp it in a
newer language At the time Ithink it was C or C++ so they

(05:49):
needed a couple more people.
They let my dad do that, so hebrought in a couple.
So I think he was the firstperson in 98.
The next person, I think, was2001.
And then two people, one personright after that.
So when I got here so he wasjust doing that, he was doing
work on behalf of Walmart, butfor IBM for 12 years and I

(06:12):
showed up in 2010, and that'swhat I did.
So our roots in retail and itkind of came from this this is
what we're doing.
I was doing mobile apps in thePOS space is what I're doing.
I was doing mobile apps, um inthe pos space is what I started
doing in 2010.
In 2010, yeah, so we built, uh,the first point of sale app.

(06:32):
You know, hand customer heldpoint of sale app, um that later
became scan ago.
Wow, it was scanning.
So we did a.
We called it an executivecheckout because it was only for
executives and so I got here in2010, got to Northwest Arkansas
, joined the company, spent Idon't know six months building
the software and then I had togo around to different

(06:54):
executives' offices at Walmartand deploy that onto their
iPhone.
So I brought my laptop and wentto go in there and put it on
there.
So everybody got this on theiriPhone and we were using it at
three stores here locally.
So that was kind of our roots.
That was my introduction toretail technology was building
mobile apps for Walmart and thiskind of precursor to Scaniga

(07:15):
called Executive Checkout.
So did that and ultimately thenatural place for us to keep
going was in retail technology.
It was when I'd kind of learnedto.
I worked on that for a while.
Then I got into internationalintegration.
So as there were differentinternational acquisitions I
could, I was kind of mergingthose technologies, especially

(07:38):
in the POS space, writingmiddleware to integrate that
technology into the existingback office systems.
So learned a lot over probablyabout a six-year period of
working in the kind of guts ofwant to sell payment systems and
that was just a natural sort ofjumping off point for the
company in terms of thetechnology.

(07:58):
Maybe we'll back up here for asecond.

Dr. Matt Waller (08:02):
Yes.

Jared Smith (08:02):
I don't know if I addressed your question.

Dr. Matt Waller (08:05):
I asked of the three areas.

Jared Smith (08:08):
Okay, yeah, so what ?
It was us just working on thething and we needed people right
to help us work.
There were a lot of differentthings going on.
In the POS space, it was stillall entirely Walmart through the
90s, through the early 2000s,and then I came in 2010 and it
was again just individual.

(08:29):
So I would call this.
In the talent solution space,we were finding people that
could help build software.
At the time, our one customerwas IBM really.
So I mean, we were second tierto Walmart and then we really
began to branch out in 2012 orso, but we were just finding
people.
We did we have a business model, I think at the time it was

(08:50):
just we write code and we getsomeone to pay us for it.
That was it.
And then we started reallythinking about.
You know, my dad left thebusiness in 2012.
Um, he retired and, uh, I kindof carried the carried the mess
from there and was looking forwhat is the model, what are we
doing?
And, obviously, retailtechnologies.
What he'd been working on, whatI'd been working on, seemed

(09:11):
like a natural place and it wasfinding people to help us build
these systems and I wanted to doit was easier to do something
that was around, kind of anagency model find people to work
on these things, build fortheir time.
And you know we were.
I wasn't looking to buildcustom software, you know, to
sell.
I was wanting to find, you know, I was wanting to sell services

(09:33):
which were building customsoftware.

Dr. Matt Waller (09:34):
So the when you became CEO in 2012,.
Is that right?

Jared Smith (09:40):
Yeah, beginning of 2012.

Dr. Matt Waller (09:43):
And most of the history of the company has been
POS and payment-type systemsand I was an early adopter of
Scan and Go and I know itchanged a lot over the years and
Walmart's integrated all kindsof their—quite a few different
apps, it seems like, into oneapp.

(10:06):
But what do you think is thefuture of you know?
If you look at Amazon, forexample, they used to have the
no checkout right.
People would just put thethings in their cart and leave.
They dropped that, I believe.

Jared Smith (10:26):
Is that right?
Yes, well, I don't know if theydropped it.
There's still a few stores thatare doing that but I don't know
if it's feasible for a largestore, especially for thinking
about a super center like that.
So the places where it hasworked are convenience stores
and smaller.
I believe.
We have a limited selection ofitems and a limited square

(10:48):
footage for those items.
So I think that's where thosekind of you know, completely
frictionless, you knowgrab-and-go sort of technology
will make sense.
But in bigger stores I don'tthink that's where people are
headed.

Dr. Matt Waller (11:04):
And then there's I mean, the self-check
stands have become so mucheasier to use than they used to
be.
Again, I was an early adopterthere.
I've always liked it Personally.
I was curious as to how itworked, but it used to be.
There were always problems whenyou tried to do it.
And I know some of theinnovation wasn't necessarily

(11:27):
with the software Some of it was, but some of it was with, you
know, putting barcodes onproduce, for example, things
like that.
Is that right?

Jared Smith (11:38):
That's some of it, I think users just also.
Just so, I think it was acombination of users getting
better at the technology and thesoftware getting better.
Um, because it you know, Ithink most people did start to
figure out that.
You know, like in the old daysI don't know if this is the case
anymore there was a littlebitty barcode on, say, on
weighable items like fruit right, that was actually a plu, and

(12:00):
if you'd scan it would alwaysjust trip up like that would
cause a problem because it wouldgo to look for the what's a PLU
.
I forget what PLU stands for,but it's essentially how you
would identify weighable items.
Oh, okay, got it.
So limes, bananas, tomatoes,whatever, will have a PLU number
associated with them, and youcould I used to know them Like a

(12:20):
few of them that I would knowby hand you could just key them
in.
It would bring up bananas andyou'd weigh them.
So that's kind of what a youknow.
You'll see usually a chartwhere they're looking to see
bananas and looking for the PLUsor the lemons, right, yeah, not
the same as a bark.
So if you scan that they wouldoften get mistaken and go look
for the UBC and there wasn't aUBC and it would then require

(12:48):
associated intervention, whichmakes the whole experience
really not worthwhile any timean associate has to come over
there and get it restarted foryou.
So I think people have gottenbetter about what they can scan,
what they can't scan and youdon't need associated
intervention near as often.
Things like even restricteditems like alcohol Scan alcohol,
it would freeze and you had towait for somebody to come over
and check your ID.
Now there's a socialintervention will be delayed and

(13:10):
it can come later.
So you scan the item, you cankeep going and then at some
point they'll come over andcheck your ID.
So those little things like Iremember when I first started
using it I had to sit there andwait to show my ID to someone
before I can scan my second itemor my next item and now you can
just keep going and keep goingand they'll just show up during
that experience.
So it's gotten better.
But I mean I read an article, Ibelieve, yesterday, about a

(13:34):
store I think it was a Walmarttaking a bunch of self-checkouts
out in a.
I forgot the state, but it wasa maybe I want to say
Mississippi or somewhere, but itmade the retail news that they
were removing self-checkouts.
So I think it's still a work inprogress of whether customers
like self-checkouts and whetherthey're something a positive

(13:54):
impact on retailers in general.
It's probably a mix.
I would guess it is a mix and Ithink it's really hard if you're
a retailer trying to figure out.
I mean, your biggest thing isyou want to get customers
through checkout quickly, taketheir money quickly.
That's you know kind of themodel of point of sales, take
their money quickly and it's notclear where the right place to

(14:15):
do that is.
You know people don't like tocheck out lines but honestly
that's usually the fastest wayif there's not a line is to go
through to a checker who knowswhat to do, knows to grab your
items and how to bag those items, and it's a much quicker
process than you doing ityourself.
But if you only have a smallnumber of items, it's much
faster to go.
So it kind of depends on thespecific cart, that particular

(14:37):
shopping trip and things likethat whether you want to use a
self-checkout or use atraditional checkout.
So I think they'll probablyalways have a mix where you can
depend on that particularshopper and experience whether
you want to go to aself-checkout route, or you want
to go to the man lane.

Dr. Matt Waller (14:52):
Well, I noticed that.
You know, on the self-checkoutstands now they're having
advertisements on them.
It's part of the retail medianetwork strategy, so you know,
if you're a supplier, you canadvertise on their app, on their

(15:15):
website.
Audio in the store, otherplaces, screens in stores what's
that?
Gas pumps Gas pumps, yeah, butalso the self-check stand.
I've not seen that one bitch.

Jared Smith (15:29):
Absolutely makes sense.

Dr. Matt Waller (15:30):
So it's interesting how the marketing
and business processesthemselves are really colliding
a little bit.
It makes the trade-offs and thedecisions even more complicated
.
It makes the trade-offs and thedecisions even more complicated
.

Jared Smith (15:43):
Sure, but ultimately, if you're trying,
your ultimate goal is a goodcustomer experience, and a good
customer experience in aretailer is the checkout part,
is not the exciting part.

Dr. Matt Waller (15:57):
No.

Jared Smith (15:57):
You want to get that done with scanning your
items and get out.
So it's really hard to saybecause the handheld we've gone
through many iterationsdifferent here whether you have
a retailer-owned scanner on the,say, the cart yeah, I mean,
I've seen those you pick thoseup and scan yourself.
Well, that's not a greatsolution because you often don't

(16:18):
know you don't shop in the sameorder you'd want to bag.
So then you just put all theitems in there and then you
gotta, you gotta stop and bagthem all up at the end.
Or you can carry your own bagsand kind of bag as you go.
It's not ideal.
We haven't really found thesolution for you know, grab and
go, where you're just gettingusually three or four items, it
doesn't really matter, right,how you bag it.

(16:39):
You can put them in yourbackpack or whatever.
But when you're doing a fullgrocery shopping you generally
want to just put it in your cartand then bag it afterwards in
the proper way, where the eggsare going to be on top or the
bread's going to be on top andyou're going to have the meat
and heavier things on the bottom.
Well, that's a role and it'seven to have.
Any customer-held device willbe challenging.
They'll be challenging, so it'shard to figure out what's the

(17:01):
best customer experience.
That's a lot of things we do inour strategic portion of our
business is help with thosethings.
What's the best customerexperience for checkout?

Dr. Matt Waller (17:12):
How do you analyze that?

Jared Smith (17:14):
We typically approach it with a kind of
framework we call we didn't nameit at this framework I think it
was Clayton Christensen,someone like this jobs to be
done, so that we use the jobs tobe done framework to kind of
approach like what are the jobsthat the customers are trying to
get done and what are the jobsthe associates are trying to get
done?
And we kind of analyze that bytalking to all the different

(17:36):
stakeholders doing store visitsto try and best understand
what's going to maximizecustomer experience and what's
going to maximize associateexperience, because you want to
optimize both of those twothings to get your best possible
shopping experience.

Dr. Matt Waller (17:50):
Clayton Christensen was a professor at
Harvard and prolific writerReally interesting.
The only book I read that hewrote was called Innovator's
Dilemma, which is a great book.

Jared Smith (18:06):
I read it, but that's the yeah, that's kind of
I didn't know.
We started doing so.
Our lead designer he, was theone who kind of brought this
jobs to be done approach todoing the work we're doing in
terms of customer experience anduser experience at retailers,
and I wasn't sure where it camefrom.
So he told me about it and Istarted looking into it and

(18:28):
discovered there that theframework is often associated
with Clayton and Christian.

Dr. Matt Waller (18:36):
So there's also people are buying online and
picking up at store.
Also, people are buying onlineand picking up at store.

Jared Smith (18:50):
And so the associates are doing some of the
in-store shopping for theshopper.
Yes, I think BOPUS, or buyonline, pick up in store, has
been a huge transformation andwill continue to accelerate over
time, because that I meanalmost anyone will tell you
that's the best experience,particularly for grocery.
Not necessarily the experienceyou want for apparel or other
categories, but for grocery.
We kind of all know what wewant and it's all kind of

(19:11):
commodity at this point.
You know what milk you want,what eggs you want, so you put
those down and having themeither deliver straight to your
house or put in the trunk ofyour car when you pull up, it's
a fantastic experience.
Straight to your house or putin the trunk of your car when
you pull up, it's a fantasticexperience.
So many of us here in NorthwestArkansas, I would say,
exclusively shop that way.
You know, I have Walmart Plusand I'm sure a lot of other
people do where you can get freedelivery when you need it or

(19:33):
you can always do pickup.

Dr. Matt Waller (19:35):
And you also get 10% or 10 cents off on your
gas.

Jared Smith (19:39):
Yeah, you get a few things like that.
So interesting story here.
Going back to the early dayroots, one of the first examples
of BOPUS was actually theAlaskan Bush Pilot Project in
the 90s at Walmart where theWalmart in Alaska.
Well, they were trying to getitems out to more rural places

(19:59):
and they kind of came up with anordering system and then they
could put the items onto a planeto fly items out to more rural
places.
And they kind of came up withan ordering system and then they
could put the items onto aplane to fly these out.
So I've had a few of my.
This obviously predates me.
I was at the company at thetime but, uh, one of our older
associates told me about thealaskan bush pilot project that
we worked on and then online.
You know different kind of, youknow buy online and pick up in
store.

(20:19):
Iterations that go back to theearly 2000s were things that
KiteStream worked on in theearly early days.
So we've been kind of in thatOpus space or buying online
since, you know, associates toWalmartcom came online in 2000.
Interesting, I did not knowthat, but some of it predates

(20:40):
2000 because I think the AlaskanBush pilot project even
predated the Internet, or atleast predated the World Wide
Web Walmartcom, which is, Ithink, 2000 or so.

Dr. Matt Waller (20:53):
Yeah, which is close to the beginning.
So the Bush pilot thing, Idon't know all the details.

Jared Smith (20:59):
Okay, that's about all the details I have.
Right there is, it was a reallyearly iteration of this idea of
buying online um and but theyweren't they weren't directly
picking up on store.
It was kind of just how do you,how does a customer get their
items without um, without goinginto the store?
right, it's kind of earlyiteration and then it kind of

(21:21):
gained.
I forgot what we we called this.
The second iteration, which wasmore in the kind of mid-2000s,
was you would get on walmartcomand you could get it shipped.
I think it was called ship tostore and you shipped to store.
So that was a big project weworked on and my father worked
on.
It was ship to store and you'dgo back to the customer service
area at the back of the storeand pick up your items there, as

(21:44):
opposed to having a ship toyour house.
I believe this is beforeshipping was free.
You know everything was freeshipping now, but it didn't
start that way and so you couldhave a ship to the store, I
think for free.
So these kind of iterationshave been going.
I mean, things like this havebeen going around since the 90s,
and where it is today it's moreand more convenient and it's

(22:05):
accelerating all the time.

Dr. Matt Waller (22:12):
Pick up at the store obviously has to deduct
the unit from the perpetualinventory system.
I would think Mm-hmm Would itdeduct it when they pick it or

(22:33):
once it's bought online, or doyou know?

Jared Smith (22:35):
I don't know for sure.
I would imagine it's once it'spicked.
Once it's picked, becausethey're scanning the items.
I've seen them there or another.
It's not just at Meijer orother grocery chains.
You've seen them doing thiswhere they pick up the items,
scan them and create the cardthen send it out to your car.
I would imagine they woulddeduct that there.
But it's an interestingquestion of when you think about

(22:57):
online sales versus store sales.
You know Omnichannel reallyblends those two.
It does store sales.
You know your omni channelreally blends those two.
Where does that online?
So you know, I buy things allthe time.
Where get online, buy it on thewebsite, and it shows up on my
porch like the same day,something within like an hour,
because I happen to be in thestore, the local store, and
someone can bring it over.
Well, who's you know, from anaccounting standpoint, is that

(23:19):
an online purchase or is that astore-based purchase?
You know, I think those arekind of interesting ways to.
You know, omnichannel is justit's a whole new world, I think.

Dr. Matt Waller (23:30):
It is, and you know I've heard people say you
know, right now, Omnichannel isa new thing and it's different
than just brick and mortar orjust e-commerce.
It's integrating than justbrick and mortar or just
e-commerce.
It's integrating the two.
But you know a lot of peopleare saying, well, it's not going
to be long before we're justsaying commerce, because these
are just different channels ofdistribution, but they're not

(23:54):
easy to execute and they alsorequire customer education.
You know, I think about whereyou can buy online and have a
drone delivered as well, andthat's getting rolled out bigger
in the Dallas area soon and I'msure there's other kinds of
complications there in terms ofcustomer, and not only with the

(24:18):
customer but of course with theFAA and many other things.

Jared Smith (24:24):
Yeah, I mean, walmart's good at this, amazon's
good at this of trying a lot ofnew things, and some of this, I
would probably argue, isinnovation theater and not
necessarily like practicalsolutions for retailers going
forward.
And a lot of people don't havethat, even if it is practical
solutions for retailers.
A lot of retailers don't havethe capabilities right, the the

(24:44):
financial capital to do dronedelivery and, and you know, grab
and go, sort of just grabyourself and walk out kind of
things, because it's such a hugetechnical investment.
So I think, realistically, it'sgoing to be.
You know, the future of retailis not, I don't think, going to
be necessarily drones flying allover the place.
I think it will cause so manyproblems and so many retailers

(25:07):
won't be able to, wouldn't beable to do that.
But you know, getting an itemjust one of the things I heard I
mean this is just reading inthe news from a few years back
and I see this occasionallywhere Walmart associates, when
they're getting off work, candrop an item off on their way.
So they plug, you know, they, Ithink, first enlist and they

(25:29):
say here's where I live.
You know this is my home.
I'm going to drive from thestore I'm working at to there
and they'll say well, on yourroute home, there's potential
for dropping some items off.
I don't know if they're stilldoing that program, but things
like that, where it's veryconfusing, makes sense.
The lift is pretty low, theassociate's already driving home
.
You're not buying theinfrastructure of a drone or

(25:50):
anything like that.
So I think those kind ofsolutions that are accessible to
almost any retailer are goingto be the ones that win out.

Dr. Matt Waller (25:58):
What are some of the advancements you see
coming in POS over the next fewyears?

Jared Smith (26:07):
I think what people want is to scan their items
quickly, right?
So you know, putting that onthe customer side you know like
self-checkout does isn't a greatsolution, especially when you
have a large cart.
You know, when you have have,let's say, 20 items or more cart
, that's going to be challengingsolution.
You got to bag them yourselvesand you might not want to bag

(26:27):
them yourself.
And you might say, well, why amI having to do this?
And so you know, is theassociate or is the retailer
going to pay me ten dollars anhour to back my own stuff while
I do that?
So I think what you want is tobe able to get through really
quickly, and I thinktechnologies like RFID and
computer vision, where you canidentify the item and ring it up
very, very quickly, as opposedto the current of putting a

(26:51):
linear belt on each item on itsown, would be some nice ways of
getting.
We're still going to go to thegrocery store, I think, and
order online and get themshipped for a long, long time.
These are going to be thesolutions.
But I believe that checking outquickly by using either
computer vision or RFID, I thinksolutions like this where you

(27:12):
can quickly identify the itemand charge the customer without
having to have all theintervention of grabbing and
scanning the UPC.

Dr. Matt Waller (27:21):
With RFID.
I don't know if this is stillan issue, but it used to be that
if you had something that hadwater in it, or metal which is
polar, it would mess up thesignal from the move.
It's still an issue.
It's still a challenge.

Jared Smith (27:39):
But computer vision is getting really it is as we
all know, remarkable, gettingremarkable.
So the ability to identify items.
Now the problem with computervision is that you could
obviously disguise.
You know you can hide things inthe shopping cart.
So if you can't see it, youcan't.
But with rfid you couldpotentially identify items.
So I think there's still somechallenges, even with computer

(28:01):
vision.
And if you have a very fullshopping cart, what's in the
middle of that shopping cart?
Even if you have 360 degrees ofcamera, there's still going to
be items you won't be able tosee.
And if you can't see it, youcan't bring it up.

Dr. Matt Waller (28:13):
Do you think there might be some sort of
triangulation, if you will,between RFID and vision?

Jared Smith (28:19):
Yeah, I think.
So.
It feels like you can runstatistical models here.
To me you kind of know what'sin the shopping cart.
Now this seems weird to say I'mgoing to take an average of
what should be in your shoppingcart, based on what I see from
computer vision or maybe withRFID, and then figure out what
the typical cart would look likeand compare the prices I'm

(28:40):
getting from a computer visionto what I expect weight, you
know things.
I've always wondered if you,you know you should, from a
categorization model, be able tofigure out what all your items
should weigh.
And then if you say, well, thisis what computer vision is
telling me, it weighs and thisis what it actually weighs,
because you maybe have a scaleon the way out.
So I think there's some thingsyou could do to augment computer

(29:01):
vision.
You see certain things.
Then you use some kind ofstatistical model or
categorization model to tell youwhat should be there or what
you might be missing and give itlike an average price or
something like that.

Dr. Matt Waller (29:14):
I'd like to talk a little bit about your
agency.
Okay, and you know, running abusiness like you're running,
you've got 150 employees.
Is that right Somewhere aroundthere?
Yeah, and it's hard to findpeople, it's hard to retain

(29:35):
people.
Have you found that to be true?

Jared Smith (29:39):
For sure.
That got much more acutestarting in 2020.
Starting to have a pandemic.
We saw, I mean, the greatresignation was real and it's
really related to almost mosteconomists, I think, already
understand this.
When you have a lot of jobopenings relative to job seekers
and you get an imbalance wherethere's many, many job openings

(30:02):
for each job seeker say thatthere's three, four, five, even
six job openings per job seekeryou'll start getting I mean,
this is natural consequences,start seeing more and more
turnover, and so the greatresignation that we saw really
accelerate in 2021 was a resultof having just a lot of options,
particularly for technologyfolks.
There were a lot of jobs outthere, so we saw the same thing.

(30:23):
So retaining people is justchallenging.
And then a remote world.
I used to compete with firms inArkansas and then start
starting in 2020, I'm competingwith Google because they can
hire.
You know, I've had people go toTesla, go to Google, go to
Microsoft, and they didn't.
They stayed in their houseliving in Northwest Arkansas,
just took a new job.
So that completely changed thecompetitive landscape, at least

(30:45):
for talent.
Good for them, harder for us,you know, because now I was
losing employees who didn't wantto move to California, didn't
want to move to Washington, butnow could stay right where they
were and just take a differentjob.
So it just changed thelandscape considerably in terms
of retaining talent.

Dr. Matt Waller (31:04):
When you, as CEO of this agency, you've got a
vision as a company and amission and a strategic
direction.
But we're also in a time ofgreat flux, especially in
technology, because of differenttypes of artificial

(31:29):
intelligence and machinelearning and many other things
that are out there generative AIhow do you you know, I know
you're working, you have havecustomers that clients that come
to you and say we want to buildthis, but then it's part of

(31:50):
your responsibility, I wouldimagine, to say, well, here are
some new technologies we mightwant to consider as we build
this or just is this the rightthing to build?
how do you work with clients onthose kinds of things as we
build this?
Or just is this the right thingto build?
Is it the right thing to build?
Yeah, how do you work withclients on those kinds of things
?

Jared Smith (32:07):
I mean generally.
I mean it's not especially inretail.
It's not always going to be thelatest and greatest.
You don't necessarily want tobe implementing every new trend.

Dr. Matt Waller (32:16):
You don't want bleeding edge technology.

Jared Smith (32:19):
Ai is a great example.
I mean AI, I believe, is atransformative technology.
Ai is a great example.
I mean AI, I believe is atransformative technology.
It's going to transform thehuman experience, but it's also
a bubble, total bubble right now, and you never know.
Go back to the dot-com bubbleright, it transformed the human
experience.
It did the internet did, but itwas still a bubble.
So we have a bubble right nowand no one really knows what's

(32:40):
going to shake out, what's goingto be the right solutions and
what aren't.
And I don't think retailers,especially in the point of sale
space, are going to be jumpinginto the latest and greatest
trends just to find out that'snot really the direction it's
going.
So they're going to be morecautious and they should be, in
my estimation, a lot of times.
Pos technology will last for 15and 20.

(33:01):
You implement a POS, at leastfrom a software standpoint.
Once you implement something,you want to keep it there for a
while.
You don't want to be changingout your point-of-sale software
all the time.
That's just too disruptive.
You can easily brick a store Ifyour POS goes down.
You have a store that can donothing.
Your inventory system can haveproblems.
Your tax system can haveproblems.
There's a lot of systems godown and you can still sell a

(33:22):
line, but if POS goes down,you're just a very expensive
brick.
So I think they don't want tobe always on the bleeding edge
with this, or I would suggestthey shouldn't be on the
bleeding edge with this stuff.

Dr. Matt Waller (33:37):
So you're right , pos is critical for the
retailer, it's also critical forthe suppliers, it's also
critical for the suppliers andit's critical for the shoppers.
So it's a really importantsystem.
How?

Jared Smith (33:55):
safe or robust are they to cybersecurity issues
Generally ritzy safe.
I mean, there was the target.
That's what I was thinking.
The biggest one in that waswhen was that?
Probably 2014, maybe I can'tremember.

Dr. Matt Waller (34:06):
It's been a while it was a big news item.

Jared Smith (34:08):
Probably 10 years ago it was big news.
My understanding is that itcame into an HVAC system.

Dr. Matt Waller (34:14):
Yeah, that's what I heard, but you know.
So retailers got rid of alltheir HVACs.

Jared Smith (34:23):
So I don't know the details of exactly the hack,
except that it came in throughsome kind of HVAC software
system.
Why a one-cell or paymentsystem had any connection to
theirs.
So generally you know it's allcordoned off with the PCI
environment.

Dr. Matt Waller (34:38):
I think the HVAC connection is.
They want cold cash Sorry.

Jared Smith (34:47):
It's cordoned off in a PCI environment and
typically, at least historically, the software is kind of like
security by obfuscation we'reusing.
The number one operating systemfor point of sale is the 4690
operating system owned byToshiba.
I think number two would be theNCR's operating system, like

(35:08):
it's.
You know, the reason whyMicrosoft is the most hacked
system is it's the most widelyknown system.
People can, you know not manypeople have 4690 to sit around
and figure out how to hack.
So I think there's some certainthings that we use, what we call
retail-hardened operatingsystems, that are going to be
very challenging to hack and arenot prevalent anywhere in the

(35:29):
world outside of point-of-sale,like through specific
point-of-sale operating systems.
Now that's kind of changing andpeople are starting to use, you
know, typical Android operatingsystems and Linux operating
systems and there's a few usingeven Microsoft's operating
system in the POS case.
But they have to be.
You know, they're pretty lockeddown and this is not especially
with things like point-to-pointencryption, which I think when

(35:52):
the Walmart hack or, sorry, theTarget hack happened,
point-to-point encryption wasnot, you know.
So P2P has really, you knowit's widespread and has been
since I think we worked on thisstuff back in, let's say, 2015,
2016.

Dr. Matt Waller (36:08):
What does point-to-point encryption mean?

Jared Smith (36:10):
When you're at the pin pad, you're going to
interact with the pin pad.
You're usually going to swipeyour card or things like that.
Point-to-point encryption meansit's encrypted at that point
and going to be transferred overto the register.
At all points it's going toremain encrypted.
That was not always thestandard.
So point-to-point encryptionmeans you're encrypting the

(36:31):
credit card information rightthere at the pin pad, got it and
that's made it more secure.
So point-to-point encryption,pci standards I mean PCI
standards are about to changeagain, so I mean there's a lot
of stuff to keep Remind me of.
PCI standards For solve creditcard information or something
like this.
Don't quote me on that, butit's something like that.
It's essentially talking aboutcredit card information, got it?
When you're giving your creditcard to a retailer, they're

(36:55):
storing that.
They're storing all thetransaction information, but it
has to be encrypted and storedin a safe way and they have to
be kept in a cordoned offnetwork called the PCI
environment.
And these kind of things havemostly kept Windows L safe where
it's not as vulnerable tomalicious activity as like

(37:16):
especially e-commerce and theinternet.

Dr. Matt Waller (37:18):
Yeah, so okay, but in some ways the e-commerce
purchasing is also a key part ofthis.
Is it as encrypted and safe asPOS?

Jared Smith (37:34):
Well, I believe.
So I mean, I'm certainlygetting outside of you know,
over my skis, out over my skishere in terms of my
understanding of encryption.
So typically I think SSLencryption or whatever the
standard for encryption is whenyou put your credit card
information into a retailer,would be similar to the
encryption level that you wouldget at a pen patent.

(37:57):
You know there's some kind of Idon't know whether it's 16-bit
or 32-bit or some kind of youknow sophisticated encryption
algorithm that encrypts that andthen it still has to send it
over the Internet, but it's gotto send it over the Internet
encrypted.
So I think even the onlinepurchases are relatively safe,
but certainly in the store youknow how they avoid hacks in the

(38:19):
store is pretty much a PCIenvironment, clearly, if you're
the PCI environment in the store.
So clearly if you're buying itfrom home, you're outside the
PCI environment when you putyour credit card information in.
But that's encrypted there onyour computer and then sent over
a secure socket to get to theretailer.

Dr. Matt Waller (38:40):
So one other thing that this brings up.

Jared Smith (38:42):
So one other thing that this brings up, so one
thing that we're discussing hereis just how important the point
of sale system is in so manyways right, even the
most—Walmart's done a great jobof diversifying into com and so
on but it's still, I believe,going to be 90 plus percent of
their revenue.

(39:02):
I don't know, but mostretailers are going to be 90 to
95% of all of their revenuesthat come through a cash
register.
So it's still a huge deal.
It's a critical thing and willbe for a long time.
It's hard to imagine a worldwhere I mean, let's brick and
mortar if you go the Amazonroute and go completely

(39:30):
e-commerce, but you know there'sa lot of things that won't work
, particularly with, you know,soft items and grocery and
apparel.
I think it's still.
You know, most people stillwant to go do apparel in person.
That's going to come to a pointof sell the waste.

Dr. Matt Waller (39:38):
So here's a thought.
So here's a thought.
You know, we have a situationin the United States where we
have something like 60% of thegrid is past its 50-year life
expectancy.
We have growing demand forelectricity.
It's growing rapidly, and AI is, and other technologies are

(40:04):
causing a lot of this growth,but some of it's being
stimulated by governmentpolicies to electrify
transportation, et cetera, etcetera.
The point is, and at the sametime there's policies that are
coming down the line that aregoing to reduce some of our

(40:25):
current sources of electricity.
So you have two things going onthat hurt us from an electric
system perspective, and some ofthe things I've read are
predicting more and more there'salready some of it going on in
some states more and morerolling blackouts and these
kinds of things.
So my question is blackouts andthese kinds of things.

(40:48):
So my question is if you know,if point of sale is so important
and it is you want to protectit from cyber attacks, you want
to make sure it's up and runningand doesn't crash and that it's
easy to use, et cetera, etcetera.
How much thought is given byretailers to the energy for it?

Jared Smith (41:07):
Well, I think that would be mostly like can your
store run during a blackout,when powered up and most can,
they're going to have generatorsto run and you know you might
not be connected to the network,but can you still sell an item
when your store has gone dark?
You'll have internal power fromgenerators generators and can
you still sell items.
That would be a concern with afully cloud-based point of sale

(41:30):
system, for example.
You don't.
You know, a lot of people arekind of would back off the idea
of doing a fully cloud-basedthing because if their store is
down they can't.
So if the network's down, theycan't sell anything.
So infrastructure problem.
So in terms of the electricgrid, I've not heard many
retailers concerned aboutproblems with the electric grid.
I do know, and I've been instores, when the power goes down

(41:50):
, usually from a storm more thanfrom rolling.
But I lived in California foryears where we had rolling
blackouts or rolling brownouts Ithink they called them, and you
know you have generators thatyou have to keep, you know,
powered up and that runs theregisters and the lights and
keeps things going.
Now I think they're going to goto a.

(42:11):
Usually you won't necessarilyhave if the whole item file is
not stored locally.
You won't necessarily.
There could be potentially someitems you can't sell.
But if you can sell the vastmajority of your items, you
generally are going to be okaywith that.

Dr. Matt Waller (42:25):
The other thing I was wondering about is you
know, sometimes a supplier willsell two different bags of
potato chips.
I'm just using this as anexample.
They're different bags.
They have the same size, exactsame contents, but the bags are
different because it's maybe apart of a promotion or whatever.

(42:47):
The barcodes don't distinguishbetween the two of them, and so
it's difficult for the supplierto understand.
You know what was the impact ofthe different package, for
example?
But I've heard and I want toknow what you think about this.
Will I think you were?
Did you go to the Supply ChainTrends Conference that Plug and

(43:10):
Play had?
No, I did not, so they had.
The senior VP of customersupply chain for PepsiCo was a
speaker speaker, and one uh.
One of the things that hetalked about was he thinks the
upc is going to go away and bereplaced by something else.

(43:31):
I can't remember what it'scalled now, uh, but it allows
for greater um, uh,specialization like this.
Do you know what I'm talking?

Jared Smith (43:40):
yeah, I've run into the same thing when we were
building apps.
You know, know we had to buildactually because in the early
days we actually had to build aUPC scanner because they didn't
exist back then, this kind ofearly enough days that you
couldn't just download one offthe app store.
So we had to build one customto scan a UPC and then go hit a

(44:01):
UPC service and you wouldrealize you'd get multiple items
back.
So I exactly remember thisproblem of you having to
distinguish which item is theone you're talking about,
because sometimes UPC is notenough.
Now I didn't know if we had asolution.
I'm familiar with the problemof UPC not being granular enough
to tell you exactly what'shappening in terms of an item

(44:23):
level.
I don't know if I can offer anymore.

Dr. Matt Waller (44:27):
No, no, that's good.
I know he showed us someexamples where you could.
It's like a QR code so youcould scan it and you could see
what ingredients were in it, whythey were in there, et cetera,
et cetera, but you could alsodistinguish between very
granular detail.

Jared Smith (44:48):
Sure About the product.
Yeah, I mean you can encodesignificantly more information
in a two-dimensional barcodethan you can in a
one-dimensional barcode, so Iwould imagine that could be a
way of getting more granularinformation, but it's not
something I'm really familiarwith.

Dr. Matt Waller (45:04):
So back to the agency discussion.
We got back into the POSdiscussion because it's
interesting.

Jared Smith (45:11):
Yeah, more POS in retail than I've talked about in
a long time.

Dr. Matt Waller (45:15):
So you know, with running an agency we were
talking about, you know,retailers don't necessarily want
to be on the bleeding edge.
They want good people.

Jared Smith (45:29):
Reliability, I think, is just really important.

Dr. Matt Waller (45:30):
Reliability is very important and we talked
about that.
Reliability is a complicatedidea because it not only is
based on the technology, but thebusiness process and the users
and those and probably otherthings I'm not thinking of.

(45:51):
How do you think through thosekinds of?

Jared Smith (45:54):
we.
We have a whole team that doesexactly this.
They go in.
We'll have a team of probablyfour or five people and where,
virtually when a retail retaileris wanting to do some kind of
technology transformation ormodernization in their POS space
, we have a team that will.
A consulting team where this ison the you know I talked about

(46:14):
the three lines of business.
This is on the strategyconsulting side.
We'll go work with the retailerto help them think through
these problems about.
You know, talking to all thedifferent stakeholders walking
around inside of a store,understanding exactly what they
want to accomplish and helpingthem sort of think through that.
Because a lot of times ifyou're a retailer, and even if
you're in the technologydivision of a retailer, you've

(46:37):
probably had the same point ofsale system since you arrived or
got your job there.
You know, and ultimately youdon't necessarily know what you
don't know about how to go get anew one and what.
What's the the greatest, latestand greatest in terms of
technology?
And you know you might not wantto go to the latest and
greatest, but what's the, what'sthe best solution out there to
do what I need?

(46:57):
And so you need to go kind ofask them what do you need?
What specific are you want andwhat kind of retailer are you?
Are you a convenience store?
Are you a department store?
Are you a club format orsomething like that?
And each one of those are goingto have different needs and
we'll talk to the differentstakeholders users, customers do
store visits to understand andhopefully lay out a plan for

(47:17):
them to help them in a reallyregimented way to get from point
A to point Z eventually.
This is one of the problemswe've seen is, you know,
especially in this space,because it's such a critical
piece of software, that analysis, paralysis can kind of take
over for a reason where theydon't you know, there's a lot of
things they want to do andthere's this could solve it,

(47:39):
this could solve it and maybethis will solve it.
So we've been able to reallyaccelerate things by coming in
and saying we have a very clearprocess you do this, do this, do
step one, step two, step threeto get you from where you are to
where you have actuallyselected your new software and
then hopefully, we're going tobe the ones to come in and
integrate that software intoyour system.

Dr. Matt Waller (48:01):
So you build software for other companies,
and have you ever been temptedto build software for?

Jared Smith (48:09):
yourself?
Not really A little bit, butnot really.
The product space has neverbeen all that attractive to me
as just a business person,entrepreneur, I sit on the board
of a product company, softwarecompany, so I understand the
space and nothing wrong with it.
And there's, you know,everybody's kind of always talks

(48:32):
about this idea of build itonce and software build it once
and sell it over and over andover.
It's never that simple.
No it's not even close and Iliked the simplicity of the
agency model, where you'reselling essentially.
Time is simple and lower riskin a way that's really
attractive.

Dr. Matt Waller (48:53):
Okay, now I'm going to come back to AI again.
Okay, we were talking about AIearlier in the context of your
customers adopt, right, butthere's also as an agency, you
can adopt it to become moreefficient.
Do you think that's very viable?

Jared Smith (49:15):
Absolutely.
I mean, did I send you?
I think I sent you this HBRreview, you did Okay.

Dr. Matt Waller (49:20):
It was great.
I loved it.
It was the consulting study.
Yeah, where?

Jared Smith (49:24):
they, you know, see , consultants get 25% more
efficient.
So just for the listener, itwas a study put out by HBR, I
think it was Boston ConsultingGroup working on sort of you
know, it wasn't with a realcustomer, it was sort of meetup
problems and they used one biggroup without um, a ai and one

(49:47):
group with ai, and those thatwere using ai saw, uh, across
the board, 25 percent increasein efficiency and productivity
and, as you pointed out, I thinkthe more junior people had
slightly a bigger increase inproductivity.
Um, the senior folks saw only amarginal, it was like 17.
Yes, a marginal uptick, butstill they saw a productivity

(50:09):
increase.
So, absolutely that service.
You know agencies can servicemodels or agencies or whatever
you want to call them, canleverage AI to deliver value to
their customers in new andbetter ways.
And also just how we run ourown business.
You know I still have financeand operations and all kinds of

(50:30):
internal aspects to my businessand if I can leverage AI there,
I can potentially reduce costsand do things internal to my own
business to make it better,make it more productive, make it
more profitable.
And how can I leverage it inproviding service to customers?
So I think there's twodifferent ways I could use.
We Can Leverage AI.

Dr. Matt Waller (50:49):
So this is interesting.
That was a great article, veryinteresting and, yeah, I felt
like the thing that reallystruck me was you can take,
because they randomly sampledpeople to be in the test group
and the control group and youknow.
So it allowed them to have adistribution of performance

(51:10):
levels nominal performancelevels and you know, the group
in the bottom percentile couldbecome as productive as the
group in the top percentiles.
That's quite remarkable.
Absolutely, I love that.
But we're talking about thebubble, too, in AI.
That's very interesting, Ithink, of AI in terms of, you

(51:31):
know, there's the hardware to dothe training of the AI.
There's the cloud services,large language model companies
that you know.
Some are doing open AI and someare doing closed AI, although

(51:55):
open AI is doing closed AI,which is good, but then, on top
of that, there's the realspecialized AI, where they're
training it on specialized datasets rather than just general
data sets.

Jared Smith (52:07):
Yeah, so those are the four areas, in some ways,
which I think is key having thesubset of large language models
that are trained for specificproblems.
So if you're a company, let'ssay's say, retailer, and you
want to have you know using youryou're already probably weary

(52:30):
of using a chat, gpt or or oranthropic or something like that
but if you could build one thatwas you know specifically for
you, that's, I think, reallymeaningful, where it knew you
could trade.
You know you have a lot of dataas a retailer and you could
feed all that data into the yourown alum or a subset alum as a
retailer.
And you can feed all that datainto your own LLM or a subset
LLM that knew all about and youcould ask it questions,

(52:51):
financial questions, forecastingquestions and it can tell you
and it's essentially just foryou.
I think that's the direction alot of people are trying to go
is a subset LLMs for particular,for specific companies or for
specific sectors?

Dr. Matt Waller (53:10):
Yeah, the, the general LLMs.
Apparently that might be wherethe big bubble is, because so
much was poured into them andthere's no way to put a moat
around it, if you will.

Jared Smith (53:23):
I mean a lot of our customers will.
A lot of our customers willhave sent out alerts.
Do not have any of ourdevelopers ever, at least while
they're at work, using any ofthese LLMs.
They don't want to be uploadingcode because they're worried
about the security risk ofuploading code or uploading
their IP to an LLM that's goingto train on it or predictably

(53:46):
get hacked.
No one really knows how securethe data is and some of this is
a little bit opaque, so there'ssome fear about the security
aspect of using the big open LLM, the big large-language models
like OpenAI.

Dr. Matt Waller (54:03):
Switching gears a little bit.
So you're a very successful CEOof KiteString and the company's
doing really well and you have150 people.
It's a lot to manage.
You're in an environment whereit's hard to hire people, hard

(54:26):
to retain people.
Technology is moving quicklyand you know, as a leader really
I think this is true foreverybody, but particularly for
a leader it's important for youto take care of yourself and I
know you and I see each other atthe gym every day at Blake

(54:47):
Street House and I know that younot only do cardio exercise and
you do strength exercises andyou do heat treatment and you do
the cold plunge.
I knew you'd bring up the coldplunge.
I had to right, I had to.
This is something'd bring upthe cold plunge.
I had to right.
This is something, for thoselistening, we have in common.

Jared Smith (55:11):
We like to— the trendiest thing in the—.
It is trendy, we're a littleashamed, but I love it.

Dr. Matt Waller (55:15):
I do too.
I mean, I think—I'd like toknow your thoughts, but—and you
also monitor your sleep and soforth?
These are things that I do aswell.
Monitor your sleep and so forththese are things that I do as
well.
But you know the days if I'mexercising regularly and
sleeping well at night,especially when I'm not

(55:44):
traveling and then I do the heattreatment and the cold plunge.
The cold plunge in particular,for some reason has a short-term
effect on me.
It makes me very alert.
I could open my head.

Jared Smith (55:53):
I think, does it you too?
Yeah, me too.
So I mean, all of this, I'llsay, stems from at least I'll
speak for me personally I don'talways deal with stress the best
in my job, and I think this is,you know, anyone's job.
There's nothing special aboutmy job versus anyone else's.
But job can be stressful.
But, you know, feelingresponsible for the whole

(56:15):
company can be stressful.
So it's to me, you know,exercise is such a great outlet
and then the pull plunging, youknow, all those sort of things
are a good outlet where I feelI'm taking care of myself and
I'm sleeping better.
It all kind of, I think, makesme better at my job.
That I'm taking care of myself,that I'm getting good sleep try

(56:36):
to eat right, all those thingsmakes me a little bit, you know,
a little bit more alert, alittle bit better at what I'm
doing, a little more thoughtful,I think.
It's kind of where I'm tryingto get to.

Dr. Matt Waller (56:49):
Well, there's so much research to back that up
.

Jared Smith (56:52):
I just like it.
I'm not even reading, spendingthat much time.
I know you mentioned AndrewKuberman.
I'm not spending that much timelooking into what are the
benefits and I just kind of dowhat feels nice.
And going to the gym every day,like I almost get to where I
don't feel great, like I cantell if I miss a day me too,
like I don't.
I kind of just want to go, andeven if it's just go and do a

(57:13):
little bit of yoga or so I did aPilates class the other day.

Dr. Matt Waller (57:16):
um, doing something makes me feel like my
day is going to be better yeahjust you know well, there is
research that shows like forsleep, if you get short on REM
sleep which typically happensbefore you wake up, and so to
get it you need to make sure yousleep plenty long it increases

(57:39):
your perception of anxiety andmakes it harder for you to
control yourself from reacting.
You know and for those of youthat don't listen that haven't
been the CEO of an agency, butyou know Jared just said that
it's stressful, but all jobs arestressful.

(58:02):
I think there's truth to that,but I've run organizations
before and there's nothing thatcompares to the stress of
running the organization.

Jared Smith (58:14):
I have more stress doing this.
I'm not working on an oilplatform or something.
There's certainly harder jobsthan what I have and I love what
I do but it can weigh on you alittle bit when you're you know
got, you know you're that's 150families that are, you know you

(58:34):
got to make good decisions tokind of keep the keep things
going, and I think that can beparticularly difficult and has
led to many sleepless nights.
So things like like you knowstay, you know not getting
enough sleep, like I can tellthe next day I'm just not you
know I'm not the same.
I'm not going to be as good of adecision maker.
I might be more irritated abouthow you know, and if I approach

(58:57):
things from an irritated orannoyed standpoint just because
I didn't have enough sleep, I'mprobably going to make worse
decisions.
So I think getting my myexercise and getting my sleep
just makes me probably a betterdecision maker and it makes life
more enjoyable too.
For sure, there's also a reallyI mean, you're a great example
of this.
There's a good social aspectthere is.

(59:19):
Yeah, I see a lot of people, Imeet a lot of people.
I kind of see the same cast ofcharacters at the gym every
morning.
So I kind of like that too.

Dr. Matt Waller (59:26):
Yeah, our friendship grew out of running
into each other there which weprobably wouldn't have met
otherwise, absolutely you knowwhat advice?
So if you take someone, maybethey have experience working in
a professional services firm ofsome sort and they're thinking
about starting an agency oftheir own, in whatever area it

(59:48):
could be marketing, could belogistics, it could be anything.
But based on your experience inrunning a professional services
firm, what advice would you?

Jared Smith (01:00:03):
give them.
First of all, especially in theservices space, the barrier to
entry is lower.
You can obviously often do thison the side of your desk.
You don't have to necessarilyquit your job to go do this With
a product company, especially asoftware product, where you're
going to have to stop and spendmonths, if not years, building a

(01:00:24):
ready-for-prime-etime piece ofsoftware that somebody will pay
for.
That's going to take a lot oftime, effort and you know.
But a service you can just sortof figure out what, and I think
what's important here is it'ssurprising what people will pay
for.
There's a lot of things, youknow.
There's a business model that Ithink is really important in
the service space, just the RTFMI don't know if you're familiar

(01:00:46):
with it no.
Read the freaking manual.
So you read the manual.
You understand how, let's say,openai APIs Everybody's wanting
to use.
You know there's the interfacethat you've all used, where you
just type it directly in there,but using the APIs and then
agents and things like that.
There's a lot of stuff you cando here and there's so much
information on the internet ofhow to you know, say, do you

(01:01:08):
know, uh, agent based workflowand use these APIs?
Um, you can learn all of thatjust by reading the manual.
Hey, this might be YouTubethese days.
It's not a physical me, right,you read these things and a it,
but they haven't had the time togo figure this out.
You can go figure out thingsthat are meaningful, especially
in with ai, and just sellservices around that I'll help

(01:01:31):
you build an agent-basedworkflow or something like that,
and and that's a service, andyou got it.
You just you just had to spendnights and weekends learning by
watching some youtube videos orreading.
I mean this.
This, I believe, was Mark Cuban.
This is kind of he was readingthe manual and going and doing
the things that he read themanual for companies that didn't

(01:01:52):
want to take the time to readthe manual Interesting.
So I think the idea that youhave to have some brilliant
thing that no one's ever thoughtof is, I'd say, not true at all
.
I agree with you on that.
Just go find something that isneeded, and sometimes just
reading because reading themanual is harder than it sounds,
so it can take many hours andpeople have busy days and it's
not clear, especially in a largeorganization whose job it is to

(01:02:14):
go read the manual, figure outhow to do whatever interact with
a, you know, an llm through itsapi or what you know.
That's just one example of manywho's supposed to go do that.
And if you go do that and say Ican, I can help you, um, for an
hourly rate, and then you'redoing it.
And now you have.
You have a business and it'sjust one, and you can do a
business with just one personand services and that's really

(01:02:37):
meaningful.
You can do it, and then you canadd a second person.
When it's time to add a secondperson, you can add a third
person.
You never.
You don't need funding.
You don't need much.
At least, the only reason you'dever need funding at all is
because you wanted to maybeleave your job a little bit
earlier.
You weren't quite comfortablewith the salary you could make
through your services.
You wanted to go get a few morecustomers.
You could use a little bit offunding to go a little bit

(01:02:57):
earlier, but in general, that ummeans I'll pick on the venture
world for a second here.
80% to 90% of I don't know thenumber and it's over a time
period but 80% to 90% fail,don't make it and that success
rate is really challenging To dosomething that's going to only

(01:03:19):
succeed maybe 10% of the time,and I think that 20% would be
where you essentially get theinvestors back their money.
You're not really you're notgoing to the moon or anything
like that.
So if you can do somethingthat's lower risk unlike you
know a single person servicecompany that you add one person
at a time I think your successrate will suddenly go up by.

(01:03:41):
You know many multiples andsuddenly you're talking about
something that you have a 50, 60, 70% chance of being successful
.
This is why I mean the I can'tremember the name high step.
I'm really excited about thisbecause it's going to be really
different when you have a modelthat has a high rate of success,
in my opinion, especially onceyou give them support like what

(01:04:03):
you guys are planning on doing.
So supporting them with bothdollars and advice.
You're going to take this to.
You know I wouldn't besurprised if you know literally
70%, 80% can be successful.
I totally agree Versus, youknow, especially like the SaaS
world or typical software world.
You know you're honestly in the10%.

Dr. Matt Waller (01:04:20):
I mean we want to continue to accelerate our
entrepreneurial ecosystem innorthwest Arkansas.
If we've been talking about thecommunity's been talking about
ramping up the entrepreneurialecosystem in northwest Arkansas,
this seems like a great way todo it and, for those listening

(01:04:40):
who might want to know moreabout it, I interviewed Josh
Stanley about High Step and youcan find more information there.
I also wrote a LinkedIn articlethat's posted on my LinkedIn
profile that talks about it in alittle more detail.
But yeah, that's one of theneat things about professional

(01:05:03):
services firms you don't have toinvest so much like you would
in a SaaS model to get it up andgoing.

Jared Smith (01:05:12):
Yeah, and I think people are the way.
Especially venture investmentswork where you need.
I mean, if you look at like theTager sort of approach I talk
about this a lot so you knowyour exit, you know how many
multiples are you going to do 10times what you put in.
That's one part of it, but it'salso Kager is also to the one

(01:05:35):
over T.
So time as T is here, time inthe export it means time is
really really important and tohave good returns you need to
have a very short time to getgood CAGR.
So if people want a 10 timesexit, 10 times exit is good.
If you do it in four years, 10times exit is on 12 years.
It's probably minimal returns.

(01:05:55):
So you need to do it quickly.
And I think those couple things,particularly wanting really
rapid growth you know growth atall costs or whatever you want
to call that model can reallyrapid growth you know growth at
all costs or whatever you wantto call that model can be can
really challenge success as well.
So not only are you having toinvest a lot of dollars before
you really test out the market,you also have to.

(01:06:16):
It has to take off like arocket ship, for the investors
are ever going to get a reallygood return, and I think a good
return might be 10, x and somenumber of years.
So I think that's just really,really challenging.
If you can start with a smalleramount of invested dollars, the
multiples are already easierand then have a much higher
probability of success.
I think the math works outalmost better, because when

(01:06:40):
you'd say I've got, 10% aregoing to be amazing, 90% are
going to be duds, but it's goingto be maybe more.
Let's say, they trade onmultiples of revenue versus
multiples of EBITDA, so you getmaybe slightly less smaller
exits, many, many more of them.

Dr. Matt Waller (01:06:58):
And they're higher probability of success.

Jared Smith (01:07:00):
Yeah, that's what I'm saying.
Many, many more of them, interms of the probability of
success, goes from, say, 10% topotentially 60%, and so I think
it's quite okay for yourmultiples to go down a little
bit because they're trade ornevada, a cohort of agencies
that are get some investmentdollars will return probably
just as well as any other um youknow group whether it's a sass

(01:07:31):
investment just because you'llhave more successes than you
otherwise would.
I don't think people will.
Really.
I was listening to a podcastthe other day where they're like
oh yeah, the agent you know,where they were kind of talking
negatively about the agencymodel while they were
interviewing an agency owner whodid, I think, $200 million in
revenue a year and they're like,well, that's the worst model

(01:07:51):
you can look at.
And he was like what are youguys talking about?
So in general, I think,especially in the software world
and the venture world, agencieskind of get a bad rep as not
the best business model becauseit doesn't scale the right way
or something like that, and Ijust I don't agree.
I think there's so many greatagencies out there doing
hundreds of millions or billionsof dollars, um, and then

(01:08:14):
there's great software companiesand there's no reason I would
argue that one is better thanthe other.
Just somehow I think ventureand you know pretty in the
software space is alwaysattached to products and sass.
Those are the kind of that'sthe kind of golden goose these
days for venture returns.
But agencies, I think just ascompelling.

Dr. Matt Waller (01:08:34):
Well, Jared, thank you so much for taking
time to visit with me todayabout this.
It's been really interesting.
I appreciate it.

Jared Smith (01:08:42):
No problem, my pleasure.
Good to get to talk about thisstuff.

Dr. Matt Waller (01:08:47):
If you're finding value in this podcast.
We greatly appreciate yoursupport by subscribing to our
YouTube channel.
Additionally, following us onApple and Spotify and leaving up
to a five-star review would beimmensely helpful.
We welcome any feedback orquestions related to the podcast
, as well as suggestions forfurther topics and guests.

(01:09:07):
You can leave your comments onour YouTube channel and rest
assured that I will read eachand every one of them.
Please also take a moment tocheck out our podcast sponsors,
as they play a critical role inkeeping this podcast running.
For more information onspecific topics, timestamps or
links to articles mentionedduring the podcast, head over to

(01:09:29):
mattwallerpodcastcom.