Latest / 1 to 100 / E39: Jack learns about Tech Debt

Transcript

E39: Jack learns about Tech Debt · 1 to 100

0:00
We're talking about HVAC technicians or something.
0:02
Oh yeah, yeah, that's a, that's a good business.
0:05
I got a lot of that kind of stuff.
0:06
I feel like it's just going to become more and more lucrative
0:08
compared to things like tech. Yeah, we're having like
0:13
electoral issues of the house like this weekend and this like
0:17
electrician came and like I'm just renting this house right
0:20
now, but the quote they sent was like 40 grand for like 2 days of
0:24
work and it was like and it like needed to be done.
0:27
And I was just like yeah wow. Like there's certain like, like
0:32
really physical things that just need to be done.
0:35
Like it's a good business. And yeah, that's kind of what we
0:39
were talking about was just like, I started running out to
0:42
my brother's home services business and I was like, oh,
0:44
it's kind of like shooting fish in a barrel.
0:45
Like going to like a very simple business and just applying
0:49
decent digital marketing like approaches.
0:51
And it's just like, yeah, if I ever, I don't know, wanted to go
0:57
off into the forest for a while and not like, run a business, I
0:59
feel like I could still just like, you know, make a decent
1:02
living running ads. You just need to go find a
1:06
couple of guys to do the work and you start running ads and do
1:09
the business side of it and you're good to go.
1:10
Yeah, yeah. How was, how was everybody's new
1:13
year? How's everybody feeling about
1:16
going into 2026? I mean, I'm excited for 2026.
1:19
I think 2025 was like the work that I did in 2025 I think was
1:25
the most challenging I've ever done.
1:27
It wasn't the hardest year for me on a personal level because I
1:31
think two things, one, like I've had years in the past where I
1:34
didn't make any money and I had really hard work.
1:37
It's just like is working insane hours.
1:40
You're doing you're just trying to survive and you also aren't
1:43
making any money doing it. And that sucks That's you've got
1:46
this like pressure of feeling like you're not going to make
1:50
it. So I don't have like on a
1:52
personal level, not going to make it.
1:53
So 2025 wasn't like that. Like it wasn't emotionally
1:57
draining, but I think the work that I did was the hardest work
1:59
I've ever done. So I think 2026 might be that
2:02
way as well. But I'm excited to, I feel like
2:04
it's going to, I'm going to see a lot of the positive results of
2:08
that that I didn't get to see in 2025.
2:10
How about you? Good.
2:12
I mean, I feel like really, really good about 2025.
2:14
Like generally we grew like pretty well.
2:19
Like generally our profitability increased.
2:20
So those are just like two things that are great when they
2:24
go hand in hand. I feel like pretty good about
2:27
2026, mainly because I think that there's just like more of
2:33
less, like kind of like less questions with more answers.
2:35
And if you just kind of just repeat answering the questions,
2:37
is there OK, the business will grow like everything like that.
2:40
And they just have like a weird, I don't know, like esoteric
2:43
feeling that that's 2026 will probably be like a really
2:46
fruitful year for like businesses in general.
2:49
I have like a a lot of optimism for it.
2:54
I mean, basically since 20, mid 2022, it's been rough for
2:59
businesses. I mean we're, we're looking at
3:02
like 2 1/2 three years now of really challenging times for
3:06
businesses in general. So it does seem like we should
3:08
be coming out of that at some point.
3:09
You know, I always wonder like I, I think that most businesses
3:13
fail, right? So it's always been really tough
3:15
for businesses, but there are some businesses that are
3:17
crushing it and I think that, yeah, we're going to see more.
3:21
I love seeing the capital requirement come down, right?
3:25
Obviously the, the big one right now is in developers and, you
3:28
know, agentic coding AI and stuff and the idea that they
3:31
could be, you know, solo founders that have these
3:34
incredible, extremely useful businesses that they're creating
3:37
or small teams or just less of a requirement for, you know,
3:42
extremely expensive development, development teams, right?
3:47
We got Opus 4.5 now and people are blown away, right?
3:51
But something better is coming. It's going to be really fast.
3:55
So I don't know, I'm, I'm excited.
3:56
I agree with you, Zach. I think the things are going to
3:58
be great for some businesses, but yeah, also terrible for for
4:02
others. For most businesses, it's going
4:04
to be terrible. Wait, why?
4:06
Because most businesses fail. You're talking about terrible
4:08
because the 90 that like 90% fail start right?
4:12
Right. Yeah.
4:13
But if you would you have to adjust for like, that stat is
4:16
actually like really inaccurate. Right, it's a bad friend of
4:19
reference. Yeah, the better because you
4:21
have like a bunch of people that like start cupcake stores and
4:25
then you also it doesn't adjust for businesses that like
4:28
survived more than one year in which you're like a longevity is
4:32
much longer. And then it also doesn't take
4:34
into account people that just like shut down their business
4:36
for like whatever reason. And you also just like can't
4:40
really compare yourself to like if you compare yourself to
4:42
everyone, if you just like wake up and like are a moderately
4:49
decent human being if you're in the 1% like already.
4:52
So it's like if you compared yourself against people that at
4:55
least have the same level of ambition as you, it's like much
4:58
more than like 90 something percent.
5:00
You know, it's just like businesses aren't just
5:03
businesses, like it matters like who's like running them and like
5:06
stuff like that. I think of anything like
5:10
businesses might meet need, might be able to survive longer
5:14
than before because like the capital requirement to keep
5:17
something floating is like so much less than it ever has been.
5:21
And there's like more things that can just like exist without
5:24
capital requirements. And so as much as I don't think
5:28
that that's that is like a useful one, I think it would
5:30
improve like over time with technology.
5:35
But then you can also assume that with like the ease and like
5:38
the reduction in sort of barrier to entry, you're going to have
5:41
higher volume, so potentially more quote UN quote failing.
5:46
Yeah. So how do you define even like
5:48
what a business is, you know, like, you know, is it a side
5:52
project that someone just like started up and like they created
5:56
a Delaware corporation or is it like, OK, maybe it needs to like
6:01
hit a certain point to be considered a business.
6:04
My guess is if you had a business that's say at least
6:07
three to five years old with over 1,000,000 in revenue, I bet
6:11
the survival, like the annual survival rate of those
6:13
businesses are probably quite high.
6:16
You know, in any given year, I'm, I'm guessing single digit
6:20
percentage of those businesses fail, but that could be wrong
6:24
too. But that's, that's like at a
6:26
point where you're like you're doing revenue, you've been
6:28
around, got some something some momentum or at least like some
6:32
base to build off of. Right.
6:34
Talk about new businesses, Zach. What's the new business?
6:38
Not yet. Yeah, not yet.
6:40
Yeah, I'm excited though. It's not really like a business,
6:44
it's more of like an art project that I think will make money.
6:49
Hell yeah, that sounds fun. It's Bored Apes or whatever
6:54
that. Was no, no, I mean, talking
6:56
about crypto and NFTS, we can probably bring up, I thought,
7:00
one of the most interesting. We have a tweet here from Taylor
7:02
Holiday about 2026 for fictions. The one that stood out the most
7:05
of me was #1 which is just like stable coins being like a
7:08
really, really useful thing for like DTC businesses.
7:11
I think there's like a useful thing for humanity, like
7:14
generally. What do you guys think about
7:18
this tweet from Taylor? Let's see, it says it's
7:22
stablecoins. The most exciting crypto
7:23
innovation for DTC brands is incredibly boring.
7:26
That's why I love it. If you know the hyper looking PR
7:28
stable coins are just contribution margin technology
7:31
here's the math old way OK, so taxes on your revenue, new way
7:36
lower taxes on your revenue yeah, yeah, I can see it.
7:40
I think it's just. The challenge is adopt getting
7:44
people to change their behavior right?
7:45
Like that's always the that's why it's taken what, 15 years or
7:50
so for for crypto to even get aware of today.
7:54
But I, I, I guess I, I didn't look into this too much, but is
7:58
he talking about it in terms of like people using it to, to
8:01
purchase through brands and to basically exchange a monetary?
8:05
Value there's that, but I, I, I so I mean, I listened to a
8:09
podcast with the founder of stripe a long time ago, not even
8:12
a long time ago, maybe just like a year ago, and he was talking
8:14
about how they're really like interested, like building in
8:17
stable coins as like to just like the stripe ecosystem.
8:20
And the exact same thing will happen with Shopify.
8:24
I think looking at a stablecoin like adoption and speed is
8:27
happening a lot more internationally than in the USA.
8:29
Lot of that is because of the way the US banking is sort of
8:31
sort of like set up. But I do think that U.S. banks
8:35
are like, really like working at potentially like upgrading sort
8:39
of like the infrastructure. And then also, I think if you
8:41
just look at like what's happening with like World
8:43
Liberty financial and like whatever it was American Bitcoin
8:47
and stuff like that, I don't think we're like super far off
8:51
to where USD will be like digitized in some sort of way.
8:56
And if that is true, it does eliminate things like Visa,
9:00
MasterCard and things like that, because those transactions don't
9:02
need to be reconciled by them. And so the question is, I mean,
9:08
over a long enough time horizon, this just gets built into the
9:11
competitive landscape, right? Like if you have an extra 6% to
9:14
spend on acquisition, it just goes to acquisition.
9:17
So it doesn't, you know, like that.
9:18
I don't, I don't necessarily know if the if it stays on your
9:21
balance sheet forever, but if there was an opportunity to
9:26
build it into your business and just like make your business
9:28
like kind of like quote UN quote stablecoin ready whenever that
9:31
sort of shift did happen, I guess it could have like a
9:34
contribution margin sort of like season before it's just like
9:38
baked into competitive sort of like landscape.
9:41
I guess I maybe just don't understand enough about this,
9:45
but I feel like the the incentives to get there so hard
9:50
to, it's hard to get to the to the right incentive.
9:54
So Stripe obviously like they're incentivized because they're the
9:56
payment processor. And so yeah, I think it cut out
10:00
Visa, MasterCard and all those guys like that doesn't hurt them
10:04
at all. But obviously it hurts Visa and
10:07
American Express and MasterCard significantly.
10:11
And then you've got this like decades of behavior where we've
10:15
been trained into this idea of like credit card point.
10:19
And that exists because of that ecosystem, right?
10:22
The only way that the reason why there are things like cash back
10:26
on your credit cards or airline points or whatever it is, is
10:28
because of the fact that there's basically extra money being
10:31
taken in that process that doesn't need to be taken.
10:34
And that system was basically set up.
10:36
I, I mean, I'm assuming the way it started and I haven't
10:40
researched this stuff, but I'm assuming the way it started as a
10:42
credit card company said, OK, well, we can get more users by
10:45
offering this benefit. And if we offer this benefit,
10:48
we'll just pass on that fee to the merchant and the merchant's
10:50
not going to say no because they need to take the credit card.
10:52
And so that's how that stuff starts, right.
10:54
So like I just signed up for the Robin Hood credit card a recent.
10:57
It actually came out. Yeah, yeah.
11:00
So I, I just got it literally like 6 days ago.
11:03
And so now I'm putting everything on there, 3% cash
11:06
back if I drop it in my Robin Hood brokerage account.
11:09
And now I'm getting, and then I get 3 1/4% interest on that as
11:13
well. So it's just like auto funneling
11:15
my personal spend. I don't know if it's a business
11:18
card or not, but I'm just a bit experimenting with that.
11:21
So like that seems that's a great benefit to me as a
11:24
consumer on money I'm already going to spend.
11:26
So that you kind of have to like get the consumer out of that
11:28
mindset. Like I will use my credit card
11:31
to get my my rewards essentially over a debit card.
11:35
If there's not extra fees, the fees have to outweigh the
11:38
benefit that I get from that card or credit card.
11:42
Yeah. So the way that this happens is
11:44
if we're looking at this calculation that Taylor did, he
11:47
says there's 6% in totaled spend right from the consumer that
11:52
isn't baked in, which means that there's essentially 6% here that
11:59
can be built in under some sort of rewards program.
12:02
And whether that's in the shop app, whether that's from some
12:06
other sort of like credit issuer that's issuing credit on the
12:09
stablecoin level. And so like it essentially is
12:14
like someone just needs to create the incentive that
12:16
exists. I mean, also I would say like
12:18
things like credit card points and things like that are so much
12:21
less impactful for, let's call it the American average consumer
12:25
in which the main value of a credit card is the concept of
12:29
credit. And so, yeah, it's like, I think
12:34
the way that we think about credit cards and the way that
12:36
the average American consumer thinks about credit cards is
12:38
like very different. Like maybe people are like, Oh
12:40
my, I got some points. But like, realistically, most
12:42
personal spend isn't anywhere close for that to be meaningful.
12:45
Like it might have been a reason to like sign up for a card is
12:47
like, I'll get a free flight or something with like bonus miles.
12:51
But like I think the average consumer is their main value in
12:55
it is like the credit. Yeah, I know.
12:58
That's a good point. And so then you need to find a
13:00
way to replace that system if you actually want to move the
13:03
the average consumer off of using a credit card to using
13:07
stable coins effectively. So, yeah, I don't know what that
13:10
looks like, but I guess what I just think that process is so
13:13
hard. Like it, it would have happened
13:15
already if it was easy, right? But you got to get the business.
13:20
Do you know? That.
13:22
I might butcher this because I'm like, I haven't read about this
13:24
in a long time, but I read about it one time and about how Visa
13:29
started and how hard it was for like Visa to become what Visa
13:34
is. And essentially like they had to
13:37
like they went like state by state where they're reconciling
13:41
essentially credit checks. Like they were like a book of
13:45
like more like checks than actual like credit cards and
13:47
things like that. But like I think something
13:49
similar will happen here where it's like I think maybe at the
13:53
speed at which things can happen now, like they'll probably be a
13:56
lot more competition at the beginning and it will feel a lot
13:57
messy. But like someone will to kind of
13:59
like take the cake. And then someone will follow.
14:01
And I think it will just be a very, I mean, maybe they will
14:04
right like Visa very well could just like figure out like.
14:09
I mean it's. Something like the, the way
14:12
those reward systems work, it's not like visas, whatever that
14:16
credit card, you know, exchange ends up being.
14:19
Let's say it's 2% on average. I don't know if these numbers
14:22
are here, but 2.9, well, that's, that's digital.
14:25
I know it's less for in person payments, but anyway, whatever
14:28
that number ends up being that they get on average, they're not
14:33
keeping most of that. Most that's actually going back
14:36
into the ecosystem in some form or another.
14:39
So there's probably a way for them to reduce the cost of those
14:42
things while it's still maintaining their profits and
14:44
maybe even increase their profits.
14:46
Right. Yeah, I don't know much about
14:48
stable coins, but I can, I can see the incentive for a consumer
14:51
if the consumer can wrap their head around it and get credit
14:55
and still use stable coins. But.
14:59
I mean, it's, I think it's like the way it always should be.
15:02
Like like the fact that you can't send a wire on a Sunday is
15:07
fucking insane so. You can't send it after like
15:11
4:00 PM Eastern Time or? Something on a Friday, Yeah.
15:16
Yeah, I'm going to wait until. And if you send it on a Friday,
15:19
it won't it won't actually clear as fast as if you sent it like
15:22
Monday morning. And like there's like, like that
15:25
shouldn't exist in 2026. So the guy had a friend that
15:29
needed to pay me like a large amount amount of money for
15:32
something. And yeah, it was like we were
15:34
just in a car driving stable coins.
15:37
Boom. It was in my account in like 10
15:39
seconds. He didn't have to call his bank
15:42
to approve the wire. Like it, it is the way that it,
15:45
it just should be. It's going to take some time,
15:47
but for, for it to be in our everyday transactions.
15:49
But I think for, I mean, even just a replacing wires like I
15:53
recently doing our, we had to renew our Cloudflare contract
15:57
and it was like a quarter $1,000,000.
16:00
And I wasn't able to call in that wire to the bank that we
16:04
use in Canada. So like I had to go on on
16:07
Christmas Eve to the bank and just hand that wire like it's
16:12
like to get in my car and drive. Like that's fucking crazy.
16:16
And that is getting why that ability to use crypto is getting
16:22
a lot easier. I mean, now you can go and you
16:24
can set up a Coinbase account in a few minutes.
16:26
You connect it to your bank with plate or whatever they're using,
16:30
and then, yeah, you can pull that money and convert it to
16:33
USDC. And now you have a stable coin
16:37
paid at the US dollar that you can transact with.
16:39
And it's it literally takes minutes to do that.
16:42
So yeah, it's, it's easier now than it was.
16:45
I mean, 10 years ago, transferring crypto to somebody
16:49
was such a pain. And now it's yeah, you can do it
16:52
in minutes. Yeah.
16:53
And the incentives there, right, for, for like every party
16:56
involved, right? Every time you buy a stable
16:58
coin, they're buying U.S. Treasuries, right?
17:03
If you think about like international customers as well,
17:06
you live in a country like Argentina or Venezuela, or you
17:09
can have U.S. dollar instead of whatever currency they have.
17:12
That's being hyperinflated, right?
17:16
And then I guess for things like micro payments as well, you
17:22
know, you can't do that across payment gateways that have these
17:25
high fees. You have to do them.
17:28
And you can imagine there's a world where, you know, AI is
17:31
performing these, you know, micro payments as well.
17:34
And there isn't. There's no way that it can be on
17:37
these old Rails. Has to be on new Rails, right?
17:40
Well, and have you guys like it's interesting to see like
17:43
Coinbase, the one I use that I've got a little bit more
17:47
understanding of how it works, but they've got lending and
17:49
stuff built up now too. So you can click a button, put
17:53
your money up for lending. And I don't know, let's say you,
17:56
you put $1000 in to, to lend to somebody immediately you're
18:00
earning. It's like it's like a interest
18:03
every minute or something like that, Some crazy number.
18:05
You literally like put your $1000 up for somebody to borrow.
18:11
It's just going into the system and it looks like right now it's
18:13
at 5.6%. So it's not it's not crazy, but
18:16
it's basically liquid. Like you put the money in, you
18:18
get your 5.6%. There's probably some risk
18:21
associated with that. I have no idea.
18:22
I haven't looked into that too much.
18:23
And then at a certain point, you just like, OK, well, I need that
18:27
money, you click the button, you get it backed out.
18:29
Like the ability to move money so easily just creates more
18:34
opportunity for things like lending and yeah, borrowing and,
18:40
and just a, it just creates more liquidity in the system in
18:43
general. Yeah, I'd love to see it.
18:46
Be cool, be really cool. What else were what were some of
18:50
the his other predictions here? Social commerce has its killer
18:53
app. I'm not sure if you saw this
18:56
one. There's one in here that I
18:56
really liked. I think it's a great idea.
18:59
I don't know who can build it. It's very complicated.
19:04
Shopify and Uber merging is core, right?
19:07
It's the idea that deliveries from Shopify could be quicker if
19:11
there's a collaboration with Uber.
19:13
I think it's already announced actually would be very cool.
19:16
And then let's see this one. I don't see it.
19:21
I can't remember which one it is.
19:23
It's basically like incremental ad buying.
19:25
It is really interesting. I just wrote a quick note.
19:29
I didn't realize that it looks like Uber and Shopify are very
19:32
similar market caps. Yeah, yeah.
19:36
Shopify two O 6 billion, Uber 171 billion.
19:40
This is it. Look at this.
19:42
So the rise of the universal feed for the last decade
19:45
accepted. I have to log into Meta, TikTok,
19:46
Google, Snapchat to buy media using the exact same creative
19:51
right in these walled gardens. So universal feed may be
19:56
launched by North Beam or house. I could see house doing
20:00
something like this. You upload the video once and
20:02
then in real time it's being placed wherever.
20:05
I don't know. Impression is cheapest, intent
20:07
is highest, CPA is lowest, ROI is highest, right?
20:12
So if it's cheap on app, loving it buys it there.
20:13
If it's cheaper on shorts and moves the budget there, powered
20:16
by incremental optimization. How cool would this be?
20:19
This would be awesome. Right.
20:21
I mean, I don't know, there's a lot of trust in that you're your
20:24
incrementality provider, but it's a hell of a lot better
20:27
than, you know, trying to juggle 6 different marketing platforms,
20:31
right? Probably using post purchase
20:33
surveys for that. Just kidding.
20:39
Let's see what were some of these other predictions.
20:43
So we've got some from Sean as well.
20:47
Why 2026 will actually be good. This isn't much of A prediction
20:54
actually. This is more just the world is
20:56
going to be better, which I can get on board with.
21:00
But Cody had a few predictions as well.
21:05
These are more marketing related though, right?
21:09
TikTok shop continues to crush Meta goes after new customers.
21:15
This is dense. Have you guys seen much of this
21:18
AI content in in ads or actually like whenever I look at full
21:24
play, I don't see any AI ads in the big brands ads inventory.
21:30
I think it really depends on how you're defining what is an AI ad
21:38
like statics? I don't think it's I don't think
21:40
you could discern whether something was AI or not.
21:42
Now video, I think there's still like a little bit of sort of
21:49
like difference between everything.
21:50
I think the one prediction I really liked about Taylor was
21:53
that the death of the creative strategist.
21:57
I think it's a bit potentially dramatic, but it's like general
22:02
thesis here, right? Is that the pontification of
22:06
like what to create next is going to be kind of gone and
22:08
it's going to be much more like directing an orchestra.
22:10
I mean, he didn't say that. I kind of just like summarize
22:13
that. Is that where it's like you need
22:15
to create, connect the pipes, you need to essentially create
22:18
as much volume as you possibly can and a lot of like the ego is
22:23
going to be removed. I, I generally agree with this
22:25
and I think it's like really aligned with like our business,
22:27
which like I would say most of our customers are creative
22:29
strategist. I I still think that there's
22:32
going to be potentially some of these people that turn into more
22:35
like brand strategist, which is like, where should this brand
22:37
go? And then how do we make sort of
22:40
like whatever knowledge base that we're using to inform the
22:44
things that we're creating next? Like how do we make this align
22:46
with the direction we want this brand to go?
22:48
But I think like individual asset creative strategist and
22:51
like sort of like over engineering, I think it's going
22:54
to whittle away, you know, I, I don't think we're going to have
22:59
people writing scripts and sending them off.
23:01
I, I think it'd be much more of like a review unlock the
23:06
workflow type process. So.
23:09
You know, I think I agree. When I look in foreplay and I
23:13
look at like it's it's all brand is there.
23:16
I guess you could you could separate the creative strategist
23:19
from the brand strategist in this way.
23:21
But if I look at like the brands that are really ripping it,
23:23
they're growing so fast. It a lot of it is brand and
23:28
sometimes really unexceptional creative.
23:30
Like this is new project. It's just a flat layer taken
23:34
point of view with the guy's feet in the shop.
23:36
There isn't like a you don't need or maybe you do, I don't
23:40
know a creative strategist for this, but you do need somebody
23:42
who gets the brand who can like forecast culture and say this is
23:47
what people want to see, right? I think this kind of align what
23:51
you see with like. Like I don't think you can do
23:53
that with a supplement like put like a element T pack out on the
23:59
ground because but like this it's like, oh, it's AT shirt.
24:02
I like this T-shirt. Like it's like this is just the
24:03
product. The product is the ad.
24:05
Right. Yeah, I hear you.
24:08
Yeah, Yeah, you can do this with supplements.
24:10
Zach, do you do you see that happening in 2026 or kind of a
24:14
process that can carry I mean start it's already started
24:18
obviously, but like just continues to expand in 2026 and
24:23
that's years out or do you actually think it's a pretty
24:25
quick? Change the more and more brands
24:29
that I talk to they all say that they're all trying to figure out
24:33
if they really just need to always be creating new ads
24:35
consistently without any shortcuts and like I think the
24:39
answer becomes yes and yes more and so I the other true thing is
24:44
that there's like 1000 ways it's going to cat right so it's like
24:48
I, I don't really I I I honestly think that potentially people
24:52
are like way too into action Twitter about like the way in
24:55
which we're scaling on meta. It's like tries different things
24:59
with your brand and with your account, like if you 3X the
25:03
volume of creatives without reducing the quality, did you
25:06
scale more? And if the answer is yes, then
25:08
you found that to be true for you.
25:10
But then there's also companies that are like super, super brand
25:12
centric in which that like short term scale is potentially like a
25:19
long term threat. So I think it really comes down
25:23
to just defining what like matters to you like right now.
25:27
And it's like, do you want to like double the number of
25:30
customers you brought in from last year, total volume of
25:34
customers? That's true.
25:36
Then I would say increase creative diversity in whatever
25:40
way that you guys can like skin the cat if.
25:43
Quite like what you said about there's a thousand ways to skin
25:45
a cat, because I think that that kind of, from how I see it, that
25:49
is why launching so many creatives works.
25:52
I used to work with this brand and they were huge and no one
25:55
got it. And they worked with a couple of
25:56
different agencies and I'd meet with these different agencies
25:59
and they would chat and then another one come in and they'd
26:01
all be like, what's your hero product?
26:02
And this brand hit the quota on Google for the amount of
26:07
products you're allowed to have listed.
26:08
I don't know. It was something like 2 million.
26:10
They had to apply to increase the the counter products.
26:13
Then I guess version of let's launch thousands of different
26:18
creatives was let's launch thousands of different products.
26:21
That's just can keep finding new ways to provide supply to market
26:24
demands. They found new best selling
26:25
products through the process of just launching new stuff, right?
26:29
And I think that's the same thing is like you're creating
26:34
more that's in competition for the customer, right?
26:38
Like if you're if you're like a one SKU store, that product
26:41
better work because all your chips are in black.
26:43
Same thing with an ad. If you have one ad that better
26:46
work because all your chips are in black.
26:48
But as you introduce more, you give yourself more opportunities
26:50
to hit a home run, right? So yeah, I think there's, I
26:54
think I couldn't agree with you more.
26:55
There's a thousand ways to skin a cat.
26:57
But like activity moving forward is any way you're going to find
27:01
it, right? I say the only way, but I guess
27:06
you know, there's a thousand ways.
27:09
I just keep thinking with all with all things AI and the way
27:14
that we change workflows and start thinking about things
27:17
differently. I keep wondering how the
27:22
technology question is one thing and like what the technology can
27:26
do and then the human adoption of those technologies is the
27:29
other thing. And there's a 2 separate factors
27:33
and I just, I don't, I still don't know how, I don't feel
27:37
like I have a good sense of how fast humans will adopt these
27:39
technologies. And that's that's like a the
27:43
biggest variable in all of this. Also, if you had a brand that
27:47
grew 300% last year, are you going to change anything in
27:50
January of this year? Possibly fucking not, right?
27:54
Yeah. So it really depends on where
27:55
you're at and if you need to adapt versus not adapt like.
28:00
Yeah, but if you've been in slow decline for three years, you're
28:04
going to try whatever to try to get yourself.
28:07
Yeah. Is there anything that you guys
28:13
think will be drastically different about the way that you
28:17
approach things in 2026 versus the way that you approach things
28:21
in 2025? Yeah, for me personally, I think
28:24
I'm I'm more and more sold on more activity.
28:27
I think us just and you said something once in the podcast I
28:30
thought was it's really stuck with me is that most things are
28:34
our two way doors. Most things aren't one way
28:36
doors. You know, we've tried things at
28:39
foreplay that were on paper a bad idea from everyone who
28:43
pitched like an opinion said, oh, this is terrible because I
28:46
know many chat tried it and it sucks that many chat and then it
28:50
worked for us. So I think that I'm so sold on
28:54
activities, on doing as many things as possible, even if you
28:59
know, some of those things end up being a swing and a miss, it
29:02
creates more opportunity, I think.
29:04
I think the one thing that I really took away from 2025 is
29:09
that in in a business and especially like when you once
29:12
you get beyond that, I have an idea and I want to see if the
29:15
market wants this idea, you enter a different mode of
29:17
business and it becomes very much about execution.
29:21
And I think that's the thing that I took away from 2025 is
29:24
how do you become excellent at execution?
29:26
And it's like, how do you get a little bit better every day?
29:29
How do you get a little bit better every week?
29:30
How do you get a little bit better every month and then be
29:33
thinking about the big swings you're going to take.
29:35
But the reality is and in a given year, you're not taking
29:38
very many big swings, you're taking 123.
29:42
I mean it depends on the business I guess.
29:43
But like in the in my context, it's like one to three big
29:47
swings in a year and everything else is just about like doing
29:50
the little things really well. And so I think, I think I didn't
29:53
pay as much attention to that previously in 2025.
29:56
That really felt like a big learning for me.
30:00
And I think for 2026, that's something that I really am
30:03
planning on doubling down on just executing really well and
30:06
the daily things and then being and it actually executing really
30:11
well on the big swings too. But just making sure that like
30:13
you're like, what really matters is not so much the big swing,
30:16
it's like the swing. And then how do you iterate on
30:18
top of that after you make that happen?
30:21
Because it almost never works the way you want it to
30:23
initially. And yours is very similar.
30:26
Like I think the thing that potentially we lost a little bit
30:29
near the second-half of 2025 was sort of just like velocity that
30:34
like small features were built. And I think part of that was
30:39
like, I wouldn't even call it tech debt, but just like product
30:42
considerations beyond the thing that the feature was touching.
30:46
And so I think my biggest goal is to just like eliminate the
30:50
idea that because something effects something else, that the
30:55
pace needs to slow down. And like all those things need
30:58
to be considered. Like I almost just think like
31:00
that, that doesn't really matter.
31:01
Like it is, you know, a bit like you don't want to get into a
31:05
situation where where you're playing whack A mole, but like
31:09
if you need to hit a few different areas, that's not that
31:12
sort of bad. So I think I just want to get
31:14
back to consistently shipping new products because I also
31:19
think that's just like generally the best marketing as well and
31:23
then have a little more attention into sales.
31:28
And it was big, the biggest sort of like changes and differences
31:31
to 2026. He says that while our head of
31:34
sales is on our RIP right now, he's absolutely crushing, no.
31:39
I mean like more my my attention.
31:41
Right. Yeah, I get you.
31:42
Yeah, I get you. Yeah.
31:45
Did you guys see this from Meta? One SEC?
31:49
OK. I mean, the first line, I don't
31:52
know if I can even read it, but urgency forces prioritization
31:56
and stops people working on stupid things.
31:58
I think there's a idea of tech debt and like I know it's
32:00
important. I don't want to minimize it.
32:02
I'm not a deaf, so I I probably shouldn't talk about it, but
32:05
like it. It's like a red light to me,
32:07
right? It's a stop instead of build
32:09
right? And I don't know, I, I think I
32:13
have I have a bias to action. I like the idea of building
32:16
things, even if they kind of suck to begin with.
32:19
I'm not sure if you guys remember when sharp brought up
32:22
on one of our first episodes, he built a button.
32:26
I think himself worried like he made someone do it for like
32:29
Apple Pay or something that didn't even work just because he
32:33
wanted to see how many people click it, right?
32:35
I think that sort of thing, even if it sucks, even if it doesn't
32:38
work, you've learned something in the process, right?
32:42
But tech debt and a lot of the things that you hear that slow
32:46
things down, I think there's there's some validity to it, but
32:50
I think it slows things down a lot in the process.
32:53
I do think it's such a hard thing.
32:54
I mean, having experienced a lot of tech tent over the last 15
33:00
years, there's a it is a trade off.
33:03
There's a balancing act that you have to play that right because
33:06
you get certain things where it's like, OK, this thing is not
33:10
very good right now. Let's say it's a this this bit
33:12
of code is A4 out of 10 in terms of functionality.
33:15
Like it, it works, but it kind of sucks and you want to make it
33:18
better and then you start having to look at like, OK, well, if I
33:22
what is the cost of getting this to A5678, whatever it is?
33:26
And then what is the, what is the business cost of not doing
33:30
that thing? And in, in software, those that
33:33
cost, the cost of not solving it often time compounds as you
33:37
grow, right. And I think the mistakes that a
33:40
lot of people make with solving tech debt, and there's nuance to
33:43
us as well too, is Luke's taking something and saying, OK, well,
33:47
I got to get this to a nine, right?
33:48
If it's, if it's not a four, I got to make this a nine.
33:50
Well, to get it to a nine, you have to stop what you're doing
33:53
for three months or get that fixed.
33:56
But that's really not what the objective should be.
33:58
The objective should be how do I get it to a 5?
34:00
Because that is 20% better than it was previously.
34:04
And in terms of looking at like the exponential cost of this, it
34:08
maybe reduces the cost to your business by cuts that in half of
34:12
like whatever problems this is creating, right?
34:14
And so I think that's the thing that's I, I can't say I'm
34:18
extremely good at that, to be clear, like I think I'm still
34:20
learning how to navigate that. But that's the constant
34:23
conversation we're having it like, how do we how do we look
34:25
at that trade off and it it's a constantly shifting target in
34:29
that context and usually just slightly incrementally better is
34:33
the right approach as to saying no, we just got to redo this
34:35
whole thing. Right.
34:36
It reminds me of our conversation with Matt and Ryan
34:39
from Clover Labs, you know, and stamped a little bit too
34:42
actually, because stamped it seems like it's a bit of a, a
34:45
case study and just build fast and create as much tech debt as
34:49
possible with that original founder from Singapore.
34:53
Is that right? Yeah, yeah, yeah.
34:55
And it's like, OK, build, build fast.
34:57
Don't worry about the tech debt. You have to pay for that
34:59
eventually. But then I think that the then
35:03
the problem on stamped was like, OK, well, we're just going to,
35:05
we're going to do everything the right way.
35:07
But if you have to, if it takes you 2 years to do it the right
35:12
way when you could have been making it incrementally better
35:15
every month, that making it incrementally better every month
35:18
is the better approach in my opinion.
35:21
You you fix the problem in some ways the the same way that the
35:25
problem was created, right? Yeah, I could be wrong, but
35:29
that's my work in theory in terms of what the the right
35:32
approach is. I like it, yeah.
35:34
It's it's it's interesting because Stamped is a product
35:37
that so many shop, so many merchants use and get a lot of
35:42
value from. But it was built in this rush by
35:45
a savant who made a huge mess. But like, it worked, right?
35:49
Yeah. And it still works.
35:51
Like to be clear, it's like it works.
35:53
It's like it's the things on the margins that you are know like
35:57
most people aren't affected by. But it has enough impact at
36:01
scale that it just is like a massive drain.
36:03
And that's the the challenge. What are you thinking, Zach?
36:07
How do you feel about tech debt and building the offset to
36:12
speed? It just depends on what your
36:14
objective is in any given time. Like realistically, if you have
36:18
like really, really high churn, probably don't build anything
36:22
new. If you have really good
36:26
retention but your growth is slow, you probably certainly
36:30
should build something new. Yeah.
36:31
So I think it just like really depends.
36:33
But I would also say that I sometimes look at things like
36:40
how, like how densely can you like build layers and things
36:45
like that. And so if you have a bunch of
36:49
debt that every time you build a new layer becomes like more and
36:54
more unstable. Not only is it increasing the
36:58
risk for something bad to happen in which there's a fire drill
37:02
and which then all productivity goes to 0, you're also
37:06
decreasing the total quality of everything you do there going
37:11
forward. And so you somehow need to like
37:15
reconcile the idea that, yeah, we might be moving faster, but
37:20
at some level the quality of output is going to be less.
37:25
And the longer that it's put off the the, our, our ability to
37:31
achieve excellence reduces. So it's like you just it's like
37:35
a very much a case case by case basis thing.
37:38
And I actually think often times it's something that you can like
37:42
come to conclusion with based on gut.
37:45
Like I actually don't think sourcing opinions often for
37:48
these types of like decisions are like really good because
37:51
like a CTO is going to be like, probably like super, super on
37:54
board with like fixing everything and making everything
37:55
clean and like all that kind of stuff.
37:58
Versus like a new developer is going to be like, oh, I just
38:00
want to do work something something new.
38:02
I don't want to touch the code that I wrote the first week I
38:05
was ever at this company when I knew nothing.
38:07
And so it's like there's bias, like an in this area, like an
38:13
incredible amount of bias. And so I think it's best to just
38:16
like go with your gut and say like, this is what needs to be
38:19
done regardless of sort of the opinions at hand is like, I
38:24
think the best way to approach it.
38:26
I think that's well said like that, a good example of that.
38:28
I know we're rebuilding our API and the reason for that is like
38:33
it, it works fine, but it's not as fast as it should be.
38:38
And it becomes a very expensive thing for us to operate because
38:42
it's, the code isn't efficient. And the way in which people try
38:47
to get information out of our API causes a lot of inefficiency
38:51
in querying the data and all that kind of stuff, right?
38:53
So if we build it better, then it makes it everything run more
38:58
effectively. And then we are able to add more
39:01
things on top of it because we're constantly adding new
39:03
integrations and all that kind of stuff.
39:05
And each of those compounds the problem with this, right?
39:07
And so we've got this thing that works.
39:09
We could theoretically continue to scale it as is, but it just
39:12
becomes more and more fragile the the bigger it gets.
39:15
And so the better approach is to just rebuild it.
39:18
And fortunately, we've known this is going going to be a
39:20
problem for like 2 years. And I've been slowly just
39:24
working towards, OK, well, here's what we need to do in
39:26
this case. Here's what we need to do in
39:27
this case. And actually like prepping to
39:29
then say, here's the the new, much better rebuilt version of
39:34
this thing that makes everything faster and better and allows you
39:39
to build those layers on top of it that aren't going to be as
39:41
fragile. This episode's going to be
39:43
titled Jack Learns About Tech Debt.
39:46
OK, did you guys see this? Did you guys see this tweet?
39:53
I'm not even sure if I'm sharing the right tab here.
39:56
Do you see it? Yeah.
39:57
Yeah, I think you are. Is this the first time you guys
40:00
are seeing this? Yeah, I haven't seen it yet.
40:03
I mean, it went super viral, but this guy's having like if what
40:07
feels like a existential crisis over playing around with AI
40:13
development tools. I should read it out loud.
40:16
Actually. Nah, We get the sense from the
40:19
first line. Basically, you know, he's, you
40:22
know, realizing how much can be done without him and so much
40:26
faster and maybe maybe better. This last line here is roll up
40:32
your sleeves to not fall behind, I guess on this topic.
40:38
Have we converted you to the dark side yet, Jeremiah?
40:41
Are you vibe coding yet? Coding, No, I'm vibe, I'm
40:47
learning a lot. I've been, I've been doing a lot
40:51
of our DevOps work at stamped over the last couple weeks and
40:53
it's actually incredible. Like what now with AI, things
40:57
that would have taken me weeks, like we no longer have a DevOps
41:02
person. And not that I can fill all of
41:05
those gaps, but legitimately in two weeks, I learned much like
41:11
most of how the system works, how to make things better,
41:13
actually made a massive number of changes.
41:15
We actually cut our AWS cost by $300,000 a year with the work
41:19
that I started doing and getting the team involved.
41:22
So I'd go and be like, hey, here's the problem.
41:24
Identified it like figured out what was going on.
41:26
Obviously using AI to help with all this because I don't know, I
41:29
didn't have, I have contacts in knowing how software works on
41:33
that and how servers work and all that kind of stuff, but no
41:35
context for the specific system for AWS, all that kind of stuff.
41:39
And now with AI you can I can talk to it and be like, OK,
41:43
well, yeah, that makes sense. That doesn't make sense.
41:44
Go ask these things. That makes sense.
41:46
That doesn't make sense. So it's not vibe coding, but
41:48
it's effectively the same thing, just that the DevOps version
41:50
that I was doing and yeah, in two weeks we cut $300,000 a year
41:55
server cost. So I didn't write the code that
41:57
fixed those things, but I did go identify all of these different
42:00
issues. There are other people
42:01
identifying some things too. So I can't take all the credit,
42:03
but literally found hundreds of thousands of dollars of like
42:07
savings that also translate to things being done in a way that
42:11
we're slower and less efficient. And so it's not just saving
42:16
money, but it's also making the system faster and more
42:19
performing so. Dude, that's incredible.
42:23
That's awesome. I mean, it's great.
42:25
Yeah. I did something useful for a
42:28
couple weeks right before Christmas.
42:30
I saw some saw one of our posts clips from previous episode
42:35
Rashad Rashad talking about how you know a customer service
42:40
request comes in and it goes straight to PR.
42:43
I thought that was interesting. Yeah, I need to figure that out.
42:45
That's what I need. I don't fucking think that's
42:47
true. I also don't know how that
42:49
works. Like sure, but like I think it
42:55
in in in I think in concept, like yes, but like there's
43:00
definitely no way. You'd have to be like the most
43:03
basic thing ever. It depends on what it is, right?
43:09
Like we had one of the, one of these things that we found that
43:12
was an issue was like, there's this internal tracking tool that
43:16
was built years ago that nobody's using and it cost us
43:20
$5000 a month and just like data transfer stuff that it was
43:23
doing. And I was like, OK, well,
43:25
nobody's using it, let's go turn that thing off.
43:26
So the one of our engineers went into Claude and basically was
43:32
like, hey, remove this function, right?
43:34
And 10 minutes later put up APR that had 250 lines of code
43:39
changed and you just like scroll through it like yeah, that's
43:42
good, good, good, good, good. And then send it like that's
43:45
that kind of stuff can be done right now.
43:47
But to like make the strategic decision and do it, I don't
43:50
think happens yet. Next episode, we're
43:53
interrogating Rashaab on what he's doing there.
43:57
All right. Well, happy New Year, guys.
43:59
OK, catch somebody else too.