Latest / 1 to 100 / E46: Acquiring Marketing SaaS & Agencies with Chase Chappell

Transcript

E46: Acquiring Marketing SaaS & Agencies with Chase Chappell · 1 to 100

• 0:00
  I'm a new fair. I haven't heard of Chase yet, so
• 0:03
  that Chase doesn't hurt to me either.
• 0:04
  We must run in different circles.
• 0:07
  You know what, dude? You know what's funny?
• 0:09
  Back when I was an agency, Chase was a villain.
• 0:13
  He was a villain to agency, right?
• 0:15
  Agencies used to. Where's all this villain
• 0:17
  language coming from? Right, because clients used to
• 0:20
  send videos of Chase and say, why don't we try something like
• 0:24
  this? And the agencies would panic.
• 0:26
  They'd be like, Oh, no, this, this person is making us do some
• 0:31
  work, creating an expectation of us to like, I don't know, run
• 0:34
  better ads. So yeah, Chase was a villain a
• 0:37
  couple of agencies that I touched.
• 0:41
  Nice. That's awesome.
• 0:42
  Well, cool, yeah. I'd love to learn more Chase,
• 0:43
  about what you've been up to. And yeah, your your villain era
• 0:48
  now I. Guess so.
• 0:52
  OK, yeah, I mean, origin story have an agency.
• 0:57
  We did a merger and still media now and then ads mastery.
• 1:01
  It's like a consulting firm where we essentially train
• 1:03
  people on like TikTok shop meta ads, shop by e-commerce and then
• 1:08
  surges where we do like web web dev for shop by stores like
• 1:12
  building out their actual websites CRO.
• 1:14
  We use a lot of AI there. We have AI audit tools
• 1:17
  essentially scanning the website because we all know like, you
• 1:20
  know, creatives are one part of the equation for ads and
• 1:23
  marketing and media buying and all that stuff.
• 1:25
  But after the click, like a lot of it really comes down to how
• 1:28
  good the Lt. VS are the AL VS conversion rate
• 1:30
  on the website, how scalable it is.
• 1:33
  So we really focus on like full funnel.
• 1:35
  So we're vertically integrated across all the companies.
• 1:38
  We also have success AI which I'm a partner in which is like B
• 1:42
  to. B SAS.
• 1:45
  But yeah, that's essentially. What we do?
• 1:46
  I'm not sure. No success, AI.
• 1:50
  What's the core product there? Yeah, it's like cold BDB
• 1:54
  outreach, like finding people's name infos.
• 1:57
  It's kind of like Zoom info. Cool.
• 2:02
  Yeah, we talked too. I forget who it was.
• 2:05
  There was somebody else. We talked.
• 2:06
  What's up? Ramsey or Nick?
• 2:09
  No, I'm sorry. I was thinking of somebody
• 2:11
  that's that's also kind of trying to compete against Zoom
• 2:14
  info. It seems like that's.
• 2:15
  Kind of, yeah, Steven. Steven OK, Yeah, yeah, it's an
• 2:18
  interesting, it's an interesting space and it is I'm curious like
• 2:23
  what what the angle is there and like what the like what your
• 2:26
  thought is in terms of like why that is a better approach than
• 2:29
  than Zoom info and like why this is the time to compete against
• 2:33
  somebody like a Zoom info. Yeah, for that specific software
• 2:37
  that we have like right now, like we're since we have so many
• 2:39
  e-commerce clients, a lot of them are using it for like
• 2:41
  wholesale and like doing e-mail automation outreach for like
• 2:44
  getting like bulk buyers and stuff.
• 2:46
  But it's grown to, you know, hundreds of millions of sessions
• 2:49
  on there now. So people are using it for a lot
• 2:52
  of different use cases. But like the initial premise of
• 2:56
  it was initially just as like a tool we were using internally.
• 3:00
  I think the interesting thing about Zoom info in the space
• 3:03
  that they play in is just how large that market is.
• 3:06
  It seems like there's so many players and they're all crushing
• 3:08
  it. We're doing incredibly well.
• 3:10
  It's like it's everybody because every fucking company in the
• 3:12
  whole world has salespeople. That's exactly what it is.
• 3:17
  Like, you know, it's like the dear Tam is literally
• 3:18
  everything. Like it's like every B2B company
• 3:22
  ever is the Tam. So it's like you just have a big
• 3:25
  enough Tam that these things can exist.
• 3:27
  And I think most of these, the way they start in six, same with
• 3:30
  six Stevens companies, like they start with the niche and then
• 3:33
  you just like slowly build it up under the hood.
• 3:35
  Like most of the data quality is sort of commoditized across the
• 3:39
  entire stack. And so you kind of like compete
• 3:41
  on pricing and a little bit of products and then try to
• 3:45
  optimize like you're like the data that you're buying or the
• 3:48
  data that you're collecting. Yeah.
• 3:50
  I mean, it's like one of these, like I feel like it's like a
• 3:51
  niche that just that continues to reinvent itself as the same
• 3:54
  thing over and over again. It's interesting too, because as
• 4:00
  AI also changes the rate at which you can start to send
• 4:04
  these kinds of messages to prospects too.
• 4:07
  So it's like it almost feels like it's creating more
• 4:09
  opportunity for the space and more demand for this kind of
• 4:14
  data access. I mean, general data demand is
• 4:18
  going to go up, right? It's like if you look at even
• 4:22
  just like super relevant right now, it's like people are
• 4:25
  setting up claw bots and stuff like that.
• 4:27
  So OK, now you have like another person.
• 4:30
  So you just multiply it a certain number of people by two
• 4:34
  in which you're saying like, hey, go and find this thing for
• 4:36
  me or go do this thing for me. And it will need data from
• 4:40
  somewhere to do XY or Z. So yeah, I think data demand is
• 4:44
  across the board. It's just going to go up.
• 4:46
  Some of it will get commoditized, but like overall
• 4:48
  demand, is simply going to increase.
• 4:51
  Did you guys see this from Nikita?
• 4:53
  The lobsters that you get is just fun.
• 4:54
  But he said something the other day about how like a lot of the
• 4:57
  tools that we use, this is it. In less than 90 days, all
• 5:01
  channels we thought were safe from spam and automation will be
• 5:04
  so flooded that they'll no longer be usable in any
• 5:06
  functional sense and we'll have no way to stop it.
• 5:09
  I mean, this is a little bit Duma.
• 5:11
  I think it's kind of fun. What do you guys think about
• 5:14
  this? I mean, I already feel that way.
• 5:17
  I just, I go through periods where I just don't even check my
• 5:19
  e-mail. Like what if you need me?
• 5:23
  You can call me if I don't have your number, then text me.
• 5:28
  And if I don't know who you are, I'm not going to save you as a
• 5:31
  contact to my phone and I won't talk to you.
• 5:32
  Like it's just like it's so I've really restricted like who I
• 5:36
  will do if you're in Slack with me, obviously I'm going to
• 5:38
  communicate with you, but just like I, I just ignore most
• 5:42
  communications who you are because it's just not worth.
• 5:44
  Slack an iMessage emails on LinkedIn, good luck.
• 5:50
  The only, I think that the only arbitrage that's like
• 5:54
  potentially left is like on any of these platforms having some
• 5:57
  sort of brand. Like for some reason I see
• 5:59
  notifications when like real verify profiles message me, you
• 6:03
  know, So it's like, I don't think this is like and like my
• 6:07
  take on this, it's just like it just, it doesn't really matter.
• 6:10
  We're kind of already there. Yeah, that's how I feel too.
• 6:12
  Like it's not going to that getting worse doesn't change how
• 6:15
  I operate really, right? 100%, you're already doing the
• 6:19
  filtering. So Chase, I kind of see you as
• 6:21
  like a polymath of ECOM. You've got the consulting,
• 6:24
  you've got agency, you've got all these different things going
• 6:27
  on. What are you like most excited
• 6:29
  about and what actually gets most of your time?
• 6:31
  Like when I think about people like Elon, think Elon was
• 6:34
  probably one of the first to sound the alarm about bots and
• 6:37
  getting a little bit of authenticity in place.
• 6:39
  So it's just not noise. But when I think about him, I'm
• 6:42
  like, oh man, how does he do all these different things so
• 6:46
  incredibly simultaneously? And, you know, in a much smaller
• 6:49
  way, you're kind of like doing this in E com.
• 6:52
  In every space that you can operate in E com, you're in it
• 6:56
  and doing fairly well. So yeah.
• 6:58
  What are you most excited about? Well, we're doing a lot with the
• 7:02
  personal brand. We're expanding our content team
• 7:04
  like crazy. We've been like staffing up
• 7:06
  editors, and we're building out like a bigger content house.
• 7:09
  We're just literally looking at another space to expand the
• 7:11
  content. Feel like personal brands, like
• 7:13
  the biggest differentiator right now for most people because
• 7:16
  there is a lot of noise going on.
• 7:18
  Yeah, like Elon and people like those are huge personalities,
• 7:21
  but it also helps with attracting talent, like the
• 7:24
  amount of intros I get or going into a conversation, somebody's
• 7:27
  already seen the content, like the doors already open and
• 7:31
  already familiar with everything.
• 7:32
  So I mean, I think personal brands are like the biggest play
• 7:35
  right now. And it's like the one note that
• 7:37
  really holds the whole thing together.
• 7:40
  It's why influencers and all that stuff is like really still
• 7:43
  big thing right now. That's how you know it's a
• 7:45
  banger at silence afterwards. I will, thank you.
• 7:48
  But yeah, yeah, I mean, we're doing a lot of content.
• 7:51
  Yeah. But I mean, outside of that,
• 7:53
  like we're doing more M and A deals.
• 7:55
  So we're acquiring. We have like a couple of Lois
• 7:57
  out there. So we're doing roll ups right
• 7:58
  now or you know, bringing in a lot of like power teams and just
• 8:04
  putting them in place with our company.
• 8:06
  So yeah, we just have a lot of LO is what kind.
• 8:10
  Of companies are you buying? Like TikTok, shop agencies,
• 8:15
  e-mail marketing, smaller marketing firms, smaller sasses
• 8:21
  and rolling those up. Anything that's like can be
• 8:25
  vertically integrated or like build out like we're getting a
• 8:28
  lot of inflow. If we have a.
• 8:29
  Heavy amount of inflow for like people reaching out for a
• 8:31
  certain thing then it's easier for us to just roll in a team
• 8:36
  and integrate and like people are reaching out now like the
• 8:39
  brands are getting so big that we're like speaking with a lot
• 8:42
  of them are all like meaning they're doing distribution like
• 8:45
  30 million + a year trying to get 102 hundred 300, you know,
• 8:49
  $800 million a year and like getting a Walmart Target and all
• 8:52
  that. So we were just having a
• 8:54
  conversation, lining up partnerships for like, you know,
• 8:56
  streamlining that process because we kind of want to be
• 8:59
  the full vertical. You know, you scale on
• 9:00
  e-commerce initially, spin up your website.
• 9:03
  We have it at the early stage, start advising them, they start
• 9:06
  getting management and then now they're going retail, Amazon,
• 9:09
  TikTok, shop so. We're really trying to cover all
• 9:11
  distribution channels possible because that's like how you can
• 9:14
  control the results the best anyways.
• 9:17
  And like do it at stages because a lot of times people get
• 9:19
  focused on like 1 little thing and they're really good at it,
• 9:22
  which is great because you're solving 1 minor problem like a
• 9:24
  conversion rate issue or like an e-mail open rate or a specific
• 9:27
  tool. Like foreplay does this really
• 9:30
  well. Like it's creative inspiration
• 9:32
  all day long and people love it and it's what people are using
• 9:35
  it for and how we're using it. They'll have other use cases,
• 9:37
  right? But like, it's the ad analysis
• 9:40
  and being able to like really diagnose like what's working and
• 9:43
  what's going on in the space and how long stuff is running.
• 9:46
  But then you have like, you know, all the other stuff where
• 9:48
  people have to do it or create the content or, you know, do the
• 9:51
  media buying or, you know, come up with the messaging.
• 9:54
  So we're trying to like, you know.
• 9:56
  Be across all of those spaces at the same time to control the
• 9:59
  full narrative across the board, which helps, you know,
• 10:01
  accelerate the process for us at least.
• 10:03
  Yeah. Which of those things, like
• 10:07
  broader categories there, do you think is going to be least
• 10:09
  disrupted by AI over the next 5 to 10 years?
• 10:12
  Oh, least disrupted by AI. Yeah, he's moving into a lot of
• 10:17
  stuff right now. It's hard to say.
• 10:20
  I mean, I feel like AI is 1 shooting some of these things
• 10:22
  out there right now too with what they're releasing.
• 10:25
  I don't think it's going to touch like service based or
• 10:27
  like, you know, the client relationship aspect.
• 10:30
  But then again, I see the AI calls and all that stuff
• 10:32
  happening now with voice. But I I still think the personal
• 10:35
  relationship aspect is still going to be there to some
• 10:37
  degree. And then I don't think it's
• 10:39
  going to take over the personal brand at all.
• 10:43
  AI is not going to like fly out to meet you and take you to a
• 10:47
  nice dinner or something like. Yeah, I was also thinking.
• 10:50
  About this whole clawed bot thing and I was trying to wrap
• 10:52
  my head of or not the clawed bot, the what is it called right
• 10:56
  now? Open claw.
• 10:58
  They keep renaming it to so many different things.
• 11:01
  There's like panic of like, you know, is this going to replace
• 11:04
  like everything that everybody does, like the AI agent thing.
• 11:08
  But I think it's only going to amplify like what people are
• 11:11
  doing currently. I don't think like there's going
• 11:13
  to be like A1 tool, like one bot that just does all media buying
• 11:17
  for everyone because the way that thing works is based off of
• 11:20
  like who's training under the knowledge, even if it has
• 11:23
  general knowledge, it's not super hyper specific.
• 11:25
  I think agencies and people will have those internal tools and
• 11:28
  I'll just amplify what they do best.
• 11:30
  And you hire their agency, you get there, you know, open call
• 11:33
  and they're humans and it's just a amplified version of what
• 11:36
  they're already really good at doing.
• 11:38
  So I just think it's going to make everyone better at what
• 11:40
  they do, but you're still going to have like a choice between
• 11:43
  different companies. Somebody's going to have a
• 11:44
  really good, you know, open claw team for AI websites or, you
• 11:50
  know, Google Ads. But they're not always they're
• 11:53
  not going to have both. I just don't see them having
• 11:55
  like A1 size fits all thing. I think it's just going to be
• 11:58
  very, very, you know, specific to their current talent skill
• 12:01
  set internally. I feel like the other thing that
• 12:05
  I, we realized the other day was like thinking about it, it's
• 12:07
  like if you just look at like the viral videos that are about
• 12:12
  a lot of these AI automation stuff, it's like always about
• 12:14
  saving a little bit of money, but it's never about making an
• 12:17
  exorbitant amount of money. And so it's like, it also kind
• 12:20
  of aligns my thesis on life, which is like, you can only save
• 12:23
  to 0. Like you quite literally can
• 12:25
  only save to 0. And so if your business was
• 12:28
  like, if the biggest optimization is your business is
• 12:31
  like removing the four KA month retainer agency, it's like, what
• 12:35
  the fuck are you optimizing for anyway?
• 12:36
  Like this, You know what I'm saying?
• 12:38
  So I yet have been able to see like, oh, this is how we're
• 12:43
  going to make boatloads more money.
• 12:46
  I've seen a lot of things like, OK, we can bring a couple
• 12:49
  categories to 0, but it's like actually the impact of your
• 12:53
  business to do that is like almost nothing.
• 12:55
  And so I think the biggest distraction is figuring out how
• 12:59
  can I, I think everyone's brain right now is going to like, how
• 13:02
  can I save a bit of OpEx? But it's like, well, who gives a
• 13:06
  fuck about saving a bit of OpEx? Like you probably could have
• 13:09
  hired someone like 15 months ago to save you the same amount of
• 13:13
  OpEx that you will through like wasting your time setting up all
• 13:16
  these quad bots. Like the thing I still haven't
• 13:19
  found is like the way AI is going to hyperscale business
• 13:23
  that's not in AI, which I think is like something I, I think
• 13:27
  that's like what everyone should be thinking about.
• 13:28
  It's like, how do we make more money not save a bit more like
• 13:30
  this? It's like a stupid thing to
• 13:32
  optimize for. But that's a, that's a.
• 13:34
  It's a lot less obvious, right? Like I mean this validate your
• 13:37
  point here, but like it's a. Hard.
• 13:39
  It's a hard conversation. Yeah, hard problems as well.
• 13:41
  It's it's very true. The the framing of that is very
• 13:44
  accurate. Yeah, like reducing time,
• 13:47
  cutting talent. It's just a cost saving.
• 13:50
  That's how people are looking at it.
• 13:51
  That's a very valid point. This is it, I think.
• 13:53
  I think it's not obvious. It's not obvious.
• 13:56
  And I think that the the cleanest example of a not
• 14:00
  obvious decision was Doctor Squatch's Sydney Sweeney
• 14:04
  campaign. They spent $12 million on a
• 14:06
  campaign that could only sell a maximum of like 40-50 K worth of
• 14:10
  product. Of course, it would have sold
• 14:12
  that wider inventory right in the world.
• 14:13
  Other goals to it. But I don't think you plug in AI
• 14:16
  into that system and you say, hey, you know what we want to
• 14:18
  do? We want to sell Sydney Sweeney's
• 14:20
  bath water. We only want to sell like, I
• 14:21
  don't know, 5000 units and this is going to cost us $12 million.
• 14:25
  I don't know if that's the sort of decision that you can
• 14:27
  automate or if it actually requires ridiculous people
• 14:30
  sitting in a room going fuck it, let's just do it.
• 14:33
  You know, that's why I think it's an amplifier.
• 14:36
  It's just going to amplify things like that.
• 14:38
  It's. Yeah, Yeah, those same, those
• 14:42
  same people sitting there with AI asking like what is the most
• 14:46
  ridiculous thing I could do in this context?
• 14:48
  And poking holes and things gets you, maybe helps you actually
• 14:51
  tune that concept a little bit more in a way that is helpful,
• 14:56
  but it's not going to generate that for you.
• 14:58
  We're leaning into it though, at the same time, like we're
• 15:02
  adapting to this stuff and like integrating it.
• 15:04
  I, I like seeing a lot of the stuff that comes out.
• 15:07
  I mean, it's moving so fast at the same time.
• 15:09
  But like what I was saying earlier with the amplification,
• 15:12
  like that's what we're like kind of leaning in towards is like if
• 15:15
  we have ad copy frameworks or if we know there's like certain
• 15:19
  things that should be structured on a website.
• 15:21
  Like a lot of it is repetitive action of just like hey, you
• 15:24
  need to install this app, but a lot of that can be truly
• 15:28
  automated with the AI process, AI creative Gen.
• 15:31
  Like there's like very specific things that were like leveraging
• 15:34
  just accelerate that whole process and it is working.
• 15:37
  This conversation's an interesting segue potentially
• 15:40
  into the conversation about blocks on their layoff
• 15:44
  yesterday. Did you, I am assuming you all
• 15:46
  saw that the letting 4000 employees go, is this coming in
• 15:50
  mass in 2026? I'm curious what you all think.
• 15:52
  But I mean, at the end of the day, I think the, the reasons
• 15:57
  for this seem kind of obvious in terms of like this is a publicly
• 15:59
  traded company and the stock price hadn't been doing well.
• 16:02
  That's your point, Zach, Like it's really good at cutting,
• 16:05
  reducing costs like that, that whatever.
• 16:08
  I don't, I didn't look into what it cost him to run that
• 16:10
  business. But if you cut 40% of your
• 16:13
  workforce, it's going to cost a whole lot less to run the
• 16:15
  business. And obviously like Jack doing
• 16:19
  this and that person that he is, send signals to the market.
• 16:23
  I haven't looked at the stock price today, but at least last
• 16:25
  night when I was looking at it was up 24% after hours.
• 16:28
  I almost wonder if Jack just saw what Elon did to his company, X
• 16:32
  after cutting all the staff, and he's like, oh, maybe I'm just
• 16:35
  like, you know, have too much bloat in my companies and should
• 16:37
  just reduce it. And Twitter's like, or X is
• 16:40
  running on, you know, 40 people when he had, you know, hundreds
• 16:44
  of thousands of employees. It's like maybe you should have
• 16:48
  just done that from the beginning.
• 16:50
  Yeah, I think that on face value, when you know, you're
• 16:52
  requiring all these different teams, the tick tock shop at
• 16:54
  agency or e-mail agency can seem like you're actually doubled
• 16:58
  doubling down on labor, right. But I don't think it is like, if
• 17:01
  you think about the intricacies of a tick tock shop, agency
• 17:03
  actually have a huge amount of access.
• 17:05
  There's a huge amount of unlock there.
• 17:06
  That isn't to do with the the people on the labor.
• 17:10
  It's to do with the capital, right?
• 17:12
  The access that you've unlocked by acquiring the agency.
• 17:15
  I, I, I know a few people who are like platform experts.
• 17:20
  They're just really good at Clavio or in their mind, they
• 17:24
  think they're really good at Clavio or they're really just
• 17:26
  like, they can only really use one tool and they've built the
• 17:28
  whole career around that. I worry about those people.
• 17:31
  I don't think acquiring those people is valuable at all
• 17:35
  because I don't know. I think that those will be the
• 17:39
  people who get automated out first.
• 17:41
  I think the context of Jack is, is pretty interesting because
• 17:44
  obviously he's, I'm not sure if you guys have heard the story
• 17:48
  about how he acquired title, but I think he was on a boat
• 17:52
  partying and he got a call from Jay-Z and he was like, fuck it,
• 17:55
  let's just buy title. So he's a pretty like out there
• 17:59
  kind of guy, you know, eccentric, a founder, CEO type.
• 18:02
  Essentially what happened in the last two weeks is that software
• 18:04
  companies were priced differently in the public
• 18:07
  markets than they are previously.
• 18:09
  So you were able. So right now, the public market
• 18:11
  markets are favoring capital expenditures for some for some
• 18:16
  reason, because essentially what Anthropic did and the way that
• 18:21
  used to underwrite a software company in the public markets,
• 18:24
  which was, you know, all of these people are writing, all of
• 18:27
  this code is going to be valued at a certain amount in the
• 18:29
  future. It's going to produce this much
• 18:31
  free cash flow is no longer true anymore.
• 18:33
  And so those humans are not longer no longer worth what they
• 18:35
  were worth in the public market. That's the only thing that
• 18:38
  changed. And then the other thing that
• 18:40
  multiplied this decision to make it easy is that like, yeah, a
• 18:43
  lot of developers aren't writing all too much code anymore.
• 18:46
  Like they're just not doing as much as like they're not sitting
• 18:51
  there writing, which means that it opens up the ability to do
• 18:53
  more by having like multiple tabs open and working on
• 18:58
  multiple features at once. And now I think this is kind of
• 19:01
  an ongoing joke at this podcast. But like, we now have multiple
• 19:04
  friends that have bots that are fixing bugs and then pushing
• 19:08
  them into Slack and being like, do you want me to push this live
• 19:11
  from the moment that a bug was reported all the way through,
• 19:14
  like fixing it? So it's like you have a bunch of
• 19:17
  people that are not as valuable as they once were tactically and
• 19:21
  in the market. And so like, this was like a
• 19:23
  really good move. Obviously, yeah, I can agree
• 19:26
  more. I think that one of the that
• 19:27
  might be one of the challenges of being a dev right now in this
• 19:30
  environment is that you've moved from a state of like extreme
• 19:35
  comfort, right, for a really long time.
• 19:37
  And comfort not being challenged on a day-to-day basis is like
• 19:40
  the the best way to dull someone's edge, to make them
• 19:43
  really incapable of adapting to, you know, new situations to
• 19:48
  changing environments. So I mean, I would hate to be
• 19:52
  one of those people who who builds very slowly and is is
• 19:55
  paid very well to build very slowly because now everyone can
• 19:58
  build and they can build very quick and they can build very
• 20:00
  well. And that's only going to
• 20:01
  increase. But I also think that, you know,
• 20:04
  there's potential that the easiest problem to solve through
• 20:06
  automation and AI customer service is also one of the
• 20:10
  heaviest, you know, in terms of headcount of I would imagine in
• 20:15
  a business like blocks. Roles, right?
• 20:18
  I could see them having so much customer support and that's all
• 20:20
  automated away and also doesn't have the same risk, right?
• 20:24
  At real customer support agents at organizations like Blocks are
• 20:29
  actually at risk, right? I'm sure you guys have heard of
• 20:33
  customer support teams selling customer information or stealing
• 20:37
  from customers or businesses like Blocks.
• 20:39
  So moving to towards automation actually hardens them, right,
• 20:43
  makes them a little bit stronger while cutting a huge amount of
• 20:45
  headcount. Yeah, I wouldn't be surprised of
• 20:47
  a significant amount of these employees that were cut, this
• 20:50
  40% was customer support. The thing that they, I think it
• 20:55
  was the CFO added some comments too, so separate from that note,
• 20:58
  but said something to the effect of we believe we can move faster
• 21:04
  with a smaller team. And I think that's an underrated
• 21:06
  aspect of this too. Like there's something about
• 21:09
  there's the complexity of change, making change gets
• 21:13
  harder and harder the more people you have to get to buy
• 21:16
  into that change. And I, I don't know, I don't
• 21:20
  know, 6000 people still, that sounds crazy to me in terms of
• 21:24
  like managing change in that context, but I suppose it's
• 21:26
  easier than managing it 10,000 people.
• 21:30
  And so maybe that's, that might actually be one of the most
• 21:33
  important factors in this decision.
• 21:37
  Oh, also, sorry, one more thing I thought was really
• 21:39
  interesting. The he mentioned making one deep
• 21:42
  cut as opposed to lots of small cuts over time.
• 21:46
  I actually think that's a really important piece of us too.
• 21:48
  Like, I mean, as somebody who just had to go through layoffs.
• 21:52
  And like that was one of the things that we thought about was
• 21:54
  like, hey, we could do small cuts.
• 21:55
  We could kind of like slowly work our way towards this.
• 21:59
  But the that's extremely demoralizing and I don't want
• 22:02
  people to be sitting there six months from now saying, am I
• 22:06
  getting cut soon? That doesn't help anybody.
• 22:08
  It's better to cut deep and then fill the gaps than to have to do
• 22:15
  that many times over a period of several years.
• 22:17
  I think I know, I, I thought that was really interesting.
• 22:20
  I actually think that is the right thing.
• 22:22
  It kind of sucks. It's like it really sucks.
• 22:23
  If you're one of those people, probably you would have
• 22:26
  preferred to have a job for another year.
• 22:27
  But at the same time, like the the alternative isn't good
• 22:30
  either. There is no good scenario there.
• 22:32
  So yeah, it's it is interesting to see that.
• 22:36
  But I think like the cutting deeper, being more generous in
• 22:39
  the short term, all that stuff is the the better approach
• 22:42
  ultimately. Chase you know what's so
• 22:44
  interesting? We're just talking now about
• 22:46
  like cutting talent and cutting labor.
• 22:50
  The tech companies sort of uniformly doing this in huge
• 22:54
  swathes, right? Every you know, big tech company
• 22:57
  seems to be just head countdown, right, But you're acquiring
• 23:01
  service based businesses, you're acquiring headcount across these
• 23:04
  different businesses that you're buying.
• 23:05
  How does this make you feel? I I feel like you're just
• 23:07
  focused on building. You're like, I want to build the
• 23:10
  biggest thing, the best thing. When we, when we like, when we
• 23:13
  acquire these companies, a lot of time they have bloat.
• 23:15
  Like it's just like there's usually like a handful of
• 23:18
  talented people doing like a majority of the work and we end
• 23:22
  up cutting like a majority of the team anyways when we bring
• 23:25
  them on because we can still handle it with our teams and our
• 23:28
  capacity. And I think a lot of people
• 23:30
  think they're at capacity, but they're not like there's so much
• 23:33
  more work that can be done in a day.
• 23:35
  And if you literally just like in office, that's why we I'm
• 23:37
  like big on in office environment and making sure all
• 23:40
  of our teams. It's like a requirement.
• 23:42
  Like we're looking at another space now to like make sure the
• 23:45
  teams are in person because you just realize like there's so
• 23:48
  much more work that can be done in a given day.
• 23:51
  And with the AI advancements and all that stuff, like you can
• 23:54
  literally get a ton done. And if you're really talented,
• 23:57
  like you're like more than just a superstar, but there's a lot
• 24:00
  of people who just do maintenance work and maintenance
• 24:02
  work to me is it's bloat. Like a maintenance work is no
• 24:06
  production for us. Like maintaining stuff just
• 24:08
  doesn't mean anything to us. It's really just like, who's
• 24:11
  actually moving the ball forward across all verticals like we're.
• 24:14
  We're spinning up like, you know, we're, we're pushing
• 24:17
  sprints on like getting features launched and, you know,
• 24:20
  acquiring teams and then cutting and then like, who's really the
• 24:22
  talented? People and you'll pretty much
• 24:24
  realize if you really sit down like with people, it's pretty
• 24:27
  obvious to the spot, but I don't know if people are even
• 24:29
  comfortable with cutting. I I have no problem with it.
• 24:31
  I think people like there's management layers and then they
• 24:35
  bring on friends or they bring on you know, people give them a
• 24:38
  role and there's really not enough to do They thought they
• 24:40
  needed the role. It's like we needed this like
• 24:43
  one thing solved and then like it's solved.
• 24:45
  And it's like where they like, I really like people who can do
• 24:48
  like the whole thing extremely well and like progressively get
• 24:52
  even better over time. That's it.
• 24:54
  Hey, progressively getting better.
• 24:56
  So I think that like with a lot of the AI automation, I mean,
• 25:00
  I'm sure you saw this on LinkedIn, suddenly every post
• 25:03
  looked exactly the same, full of M dashes, full of contrast
• 25:07
  negation. It's not this, it's that and
• 25:09
  like that. Those were the early doctors.
• 25:10
  Those were the people who were trying to maximize potential.
• 25:12
  Now, actually, a lot of those people who were doubling and
• 25:15
  tripling down on X and LinkedIn with their, you know, AI slot,
• 25:20
  they were they were getting a lot of clients in the process,
• 25:23
  right? They were actually doing
• 25:25
  meaningful work, even though it was kind of the worst.
• 25:27
  I mean, I'm not sure if you tried to read these posts, but
• 25:29
  I'm like, oh man, how could you do this with your name attached
• 25:32
  to it, You know, with your, with your face and your name attached
• 25:35
  to the to the, to this post with all these M dashes and, you
• 25:39
  know, cheesy AI prose, right? But yeah, I don't know, I think
• 25:43
  that's, that's sort of the balance right now.
• 25:45
  It's like, how do you use these tools but use them adaptively to
• 25:48
  create something that's great, right?
• 25:50
  We just launched an ad. I'm not sure if you've seen it,
• 25:52
  Chase. It's like a bunch of Africans
• 25:53
  dancing around saying how much they love foreplay.
• 25:55
  It's crazy, right? We've got so many messages from
• 25:58
  our customers saying that we're lunatics for it.
• 26:01
  It's crushing. It's it's like the the least,
• 26:04
  no. Is it crushing?
• 26:05
  No, no, it's not crashing at all.
• 26:07
  It's like it's hardly getting any spend in our account.
• 26:11
  But I thought it'd be funny and I saw one of those.
• 26:13
  I'm not sure if you guys get these ads every now and then.
• 26:15
  It's like a bunch of like, you know, people, Africans dancing
• 26:18
  around saying like happy birthday or something for
• 26:20
  someone. So I was getting these ads, or
• 26:22
  at least my girlfriend was showing me them.
• 26:24
  And I was like, we should do one of these before play.
• 26:26
  Do you want me to show you it? Why does every software and
• 26:31
  agency always use memes and like gifts and funny stuff for their
• 26:35
  ads? That's a that's a good question.
• 26:37
  I think it's because it's it's easy to communicate really
• 26:40
  complicated things came to what you think, Jeremiah.
• 26:44
  But I think that like most software, and especially as it
• 26:47
  ages and it deepens, it's really hard to explain what the thing
• 26:50
  does fully. So you kind of have to just like
• 26:54
  dumb it down to the consumer level, especially when it's that
• 26:58
  person is consuming. They're just like scrolling
• 27:00
  through their feed and make it like fun to engage with and
• 27:03
  compete with, like funny cat videos.
• 27:06
  That would be my take also because it's fun.
• 27:08
  Like, I don't know, scripting that foreplay ad with the
• 27:12
  Africans was just so much fun. So I don't know, that's kind of
• 27:15
  why we do it. Yeah, I basically have done no
• 27:21
  paid ads, but I am we're going to experiment a little bit.
• 27:27
  What's interesting though, is the things that like we tried
• 27:30
  posting this organic social, the memes and things like that and
• 27:34
  it I don't know, it wasn't that effective.
• 27:38
  Just kind of like there's a hardly any engagement, but we
• 27:41
  start posting data and telling stories about like and actually
• 27:47
  case studies aren't that effective either, to be honest.
• 27:49
  I think they are for very specific use cases.
• 27:51
  But you go tell the story about how data is saying that this ad
• 27:55
  platform is really effective and here's why that becomes really
• 27:58
  compelling and people latch onto that, ask questions about it,
• 28:02
  repost all that kind of stuff. So that stuff is way harder to
• 28:05
  produce. But ultimately, I think to me,
• 28:09
  that's what I've seen be the most effective in in software
• 28:12
  is, is being able to tell stories like that as opposed to
• 28:17
  the the meme stuff. But I haven't tried running paid
• 28:21
  ads. So unfortunately I don't have a
• 28:22
  whole lot of idea of how how effective that is.
• 28:29
  How about for for you and your software companies Chase?
• 28:33
  What do you found is a my memes? We don't do any memes or like
• 28:37
  comedic stuff. No, we we flex like either
• 28:41
  talent OR like knowledge or like really complicated new anything
• 28:48
  that's like new strategies that are crushing it that have not
• 28:52
  been fully discovered. Like a big bulk of our business
• 28:56
  is Facebook media buying, but it's everyone's doing it for the
• 28:59
  last 10 years. Everybody sounds the same, like
• 29:01
  we're good at it. But yeah, a lot of other people
• 29:04
  are too. That's fine.
• 29:05
  But you know, like TikTok shop or UGC automation, like building
• 29:10
  affiliate armies, like that's around, but people don't know
• 29:13
  how to just magically spin up 2000 craters and it's extremely
• 29:16
  valuable. So like we lean into the stuff
• 29:18
  that like is like either new feature or a new idea, new
• 29:22
  strategy that is working. That's like making waves.
• 29:25
  And then people come in for that and then they find out all the
• 29:28
  other things that we offer, which everybody else also
• 29:30
  probably has. But because we're, you know, the
• 29:34
  intro for the gate point is stuff they have not done or not
• 29:38
  tried and they're curious about or have not learned.
• 29:41
  So we try to be very innovative and also very forward thinking
• 29:46
  about what's coming up like with Manus and how, you know, Meta's
• 29:50
  acquired dad or, you know, open claw, like actually knowing
• 29:53
  about it and then seeing who's doing strategies, testing it,
• 29:57
  finding out if it works. Then we have like a rule like if
• 30:00
  it works for like a handful of people, we'll go to 30.
• 30:03
  And if it works extremely well for those 30, then it'll be a
• 30:06
  mass rollout. If it doesn't make it past that
• 30:08
  point, we just like, it's just not a strategy.
• 30:10
  We're going to apply it like across all companies.
• 30:14
  But we want to do the stuff that people are interested about.
• 30:17
  Like we know like UGC automation, We know like
• 30:19
  building affiliate armies. We know like, you know, these
• 30:23
  paid collab ads. We know the AI angles, everybody
• 30:28
  AI is just a buzzword in general.
• 30:29
  He will get excited about like those are great lead magnets for
• 30:33
  pulling people in. But like your core service, your
• 30:36
  Bulg, your business may not be that thing because it is new.
• 30:39
  Like why would that already be your number one?
• 30:41
  But you're talking about that. It's, it's the entry point.
• 30:43
  So I think like when you launch new innovative things like
• 30:46
  that's like the push, right? That like attracts people, but
• 30:51
  then you have all the other stuff that comes with it.
• 30:53
  Yeah, Jack, I wonder if part of the reason that stuff works for
• 30:56
  you all is because that, that like is your product, not you
• 31:00
  specifically, but the the ad it's so it's so tightly
• 31:06
  connected to what you actually do.
• 31:08
  So and for me, I think maybe that's part of the reason why
• 31:11
  the data angle works so well is because that's so tightly
• 31:13
  connected to what we actually do.
• 31:15
  So you can actually see like the compelling work we are actually
• 31:19
  distributing the thing that you want to go and find out for
• 31:23
  yourself, so. I should say that the the the
• 31:26
  African ad, if anyone goes and watches it, it's not working for
• 31:29
  us. It's not getting any spend and
• 31:31
  we actually haven't had like a meme ad that's or a meme of any
• 31:35
  kind that brought a huge amount of new customer acquisition to
• 31:38
  us. Our best acquisition channels
• 31:43
  are definitely partnership ads with the primary identity being
• 31:47
  the creator by far right and then the network effects of it
• 31:53
  as well. We've tried to encourage our
• 31:54
  partners to use common triggers, right?
• 31:57
  Give away some sort of lead magnet, gets people engaging,
• 32:00
  gets an organic distribution even before we put a ad spend
• 32:02
  behind it generates leads for them as well as us.
• 32:07
  So it's like serves both parties.
• 32:09
  And we have a couple e-commerce slash marketing influences,
• 32:13
  people who are very well known in the space and renowned who
• 32:17
  are generating leads for their own businesses.
• 32:19
  But after like a week, two weeks, the organic distribution
• 32:23
  drops off, right? People stop commenting, people
• 32:25
  stop sharing. That doesn't happen when you run
• 32:27
  it as an ad as well. So I think it's a, you know, I
• 32:30
  think it kind of comes back to the core of what an ad is.
• 32:31
  An ad's just paid distribution on content to people who are
• 32:35
  likely to meaningfully engage with it, perform whatever the
• 32:37
  optimization event is, whereas content is, you know,
• 32:41
  distributed based on like, who's engaging with it and what's
• 32:43
  going to maximize user attention on the platform.
• 32:46
  So yeah, if you can combine the best parts of both of those
• 32:49
  things, that seems to be what works best for us.
• 32:52
  But man, it's fun to hire some Africans and get them to dance
• 32:55
  around. I think I know you're talking
• 32:58
  about now. I, I, I just looked at it on the
• 33:00
  ad library. I saw the video.
• 33:03
  I see those guys on TikTok all the time.
• 33:06
  Yeah. I mean, another thing is like
• 33:08
  the angles of like the, I wonder, like, what people wonder
• 33:12
  about, like, how is that possible or how does somebody do
• 33:14
  something or how does this get achieved?
• 33:16
  Like, like, for instance, like an Elon Musk uses like Sleep 8.
• 33:20
  It's like a sleeping mattress. And people wonder how Elon's
• 33:23
  able to operate and do all this stuff and get good sleep.
• 33:26
  He's not out there promoting AIDS sleep, but just the fact
• 33:29
  that he uses the product and then that could be framed from
• 33:31
  their company as an ad. It's like, how does he, that's
• 33:34
  not the thing that's making him do everything, but it's the
• 33:37
  question in the back of their mind of like, how is he
• 33:39
  performing at those levels? Is it sleep well?
• 33:41
  Then all of a sudden curiosity buffer hits and then now people
• 33:44
  are going and buying the product.
• 33:46
  So it's like it brings people in it like you want those like
• 33:48
  weird unique cases where if you were like, how's this company
• 33:51
  running 2000 ads? I can't find a creative team
• 33:53
  that can produce that many. And it's like, well, there's
• 33:56
  actually this other automated system in the background that
• 33:58
  you just have never heard of. And that's like, wait, is that
• 34:01
  what that is? I'm going to go get that now.
• 34:04
  So it's like, what are people doing at scale?
• 34:06
  That is like really, really hard shit to do that seems extremely
• 34:11
  challenging to solve. And people keep batting around
• 34:13
  trying to figure it out. And then you framing that as
• 34:16
  like, well, it's, you know, it could be this or this is what it
• 34:19
  is. And then people are like, wait,
• 34:20
  that's what I need. And they jump so quick.
• 34:25
  I like that. Focus on what people are
• 34:27
  actually paying attention to and what people are interested in
• 34:29
  right now. Even if it's not your core
• 34:31
  thing, it's not the bread and butter of the agency, the
• 34:32
  service based business. It's something that you, well,
• 34:36
  jump on. Find the and extract the
• 34:39
  arbitrage while it's available. Continue to do your bread and
• 34:41
  butter things. Maybe this becomes a service
• 34:43
  that lasts a long time, or maybe it becomes a service that dies
• 34:46
  off and isn't sustained. And that's fine because there
• 34:48
  isn't always arbitrage. There isn't always value in a
• 34:52
  new hack, growth hack for forever, right?
• 34:55
  Sometimes these things die off. I like that approach.
• 34:59
  I think it's I think it's smart makes sense.
• 35:01
  You're giving the market what they want or you're letting them
• 35:03
  say, hey, I'm interested in this and you're saying we're going to
• 35:06
  figure it out for you. I think by the same measure,
• 35:09
  I'm. And I'd love to hear what you
• 35:10
  think about this Chase. I'm not interested in image ad
• 35:12
  generators at all. I saw a video the other day and
• 35:15
  some guy was like, I spent 10 hours vibe coding with Claude,
• 35:18
  this image ad generator. I was like, oh, you too?
• 35:20
  Like, why is everyone doing this simultaneously?
• 35:23
  Like someone's going to solve it and it's going to be free
• 35:25
  because everyone's building it and they want the user base,
• 35:28
  right? So they can upsell something
• 35:29
  else that has, you know, is able to retain its marginal value.
• 35:33
  Right. But my fashion project is that
• 35:35
  too. Yeah.
• 35:37
  Fun. Yeah, well, that's it.
• 35:39
  Building stuff is fun. I don't know.
• 35:41
  I see it everywhere too, but I'm trying to because there's so
• 35:45
  many AI creative gentles out there and like we try them all
• 35:48
  and they're all half like they never do what you want them to
• 35:51
  do. Here's the thing.
• 35:52
  The problem with creative gentles that drives me insane.
• 35:56
  You like nano banana pro, insane gentle Higgs field, crazy stuff.
• 36:00
  Like there's so many good creative gentles out there, but
• 36:03
  when you give it a product, it'll change the text on it,
• 36:05
  it'll resize it and then you can't get like very the
• 36:09
  consistency aspect. It's like no matter how much you
• 36:12
  prompt like to get like a full blast of like 200 ads, the AI
• 36:18
  just gets off track constantly. Like you're always having to put
• 36:21
  guardrails on this stuff. Even with like Google's latest
• 36:24
  one. It's what is that new tool
• 36:26
  called? They just launched.
• 36:27
  They just just launched Nana Banana too I think.
• 36:30
  So that's the one after Nana Banana Pro.
• 36:32
  There's another one that's for like ad creative Gen.
• 36:35
  No way. Like started with like AP
• 36:38
  anyways, but they launched it and I was like, oh man, like
• 36:41
  this might have just killed my passion project.
• 36:43
  And I was like, fine, I'll just go with this one.
• 36:45
  And I go and try and it's the same thing.
• 36:46
  It still messes up the text. So like, you know, you, the AI
• 36:52
  tool has to be able to render perfect text.
• 36:55
  And by the way, AI doesn't type the text.
• 36:58
  Like when you go on a word dog and you choose Times New Roman,
• 37:00
  it types Times New Roman, you will get Times New Roman text.
• 37:04
  When you tell AI make Times New Roman text, it looks at the
• 37:07
  Times New Roman text and it draws it and it doesn't always
• 37:10
  get the exact serif and you know, stands correct, it looks
• 37:14
  similar. AI only looks similar to what
• 37:16
  you give it. It doesn't take your product and
• 37:18
  paste it in somebody's hand, it redraws it.
• 37:21
  So that's where you get inconsistencies and all AIS have
• 37:23
  that problem. So I'm like trying to figure out
• 37:27
  the layering aspect of it of like, you know, like actually
• 37:31
  being able to composite render, you know, shadow grade and do
• 37:35
  this in a layering sequence. So that way it comes out exactly
• 37:38
  how we want our formats with gridlocked positioning.
• 37:40
  Because if I say hey, make me and features and benefits add,
• 37:42
  it throws 3 icons at the bottom and then I say make me another
• 37:45
  one, it's going to throw the three icons on my product label.
• 37:48
  And that's not my product label. I don't want that.
• 37:51
  So there's no consistency there. So there has to be a gridlock
• 37:55
  system that like hard locks, what's allowed and what's not on
• 37:57
  these things. Nobody's doing that right now.
• 37:59
  And I'm just like on the side when I have one hour a day and
• 38:03
  my fiance gets mad at me on the weekend because sometimes I'll
• 38:05
  like sit 8 hours just tinkering around with this stuff, just
• 38:08
  playing with it. But it's fine.
• 38:12
  But everybody's doing it. But you know, it's, it's, it's
• 38:14
  cool to like try to figure that kind of stuff out, but it still
• 38:16
  is a major need. But there's so much
• 38:19
  inconsistencies in it. You you say, hey, make me a
• 38:22
  studio paper backdrop with my product and it puts literal
• 38:25
  studio lights in the photo. And it's like, no, no, no, I
• 38:28
  wanted the lighting of the studio, not the studio lights.
• 38:32
  Like, you know how insane that is?
• 38:33
  Like going back and forth with these prompts and you just time
• 38:35
  suck your, you know, hours trying to get one thing right
• 38:39
  when you could have just went and done the photo shoot.
• 38:42
  So the goal for us is like, if since I've been working on a
• 38:46
  little bit of this stuff, it's like, how do you gridlock this
• 38:47
  thing and like actually force it to always put out perfect
• 38:50
  consistency? Yeah, yeah, yeah.
• 38:53
  We need that for a software product, right?
• 38:56
  This is one of the main reasons we don't have image Gen. or any
• 38:58
  sort of genin foreplay. And we thought about like
• 39:01
  imagine you have a board, a foreplay board open and you
• 39:03
  could drag your product onto an image and then you would
• 39:06
  generate like X number of variations.
• 39:09
  The problem is the consistency. And that's why I would never
• 39:12
  build in the the the creative space anyway.
• 39:14
  You're completely reliable, reliant on the model for the
• 39:17
  upper end of what it can do and then the reliability of what it
• 39:20
  can do. So even if the upper is
• 39:22
  incredible, right now, your your customers are relying on you for
• 39:27
  the reliability. But it's not in your hands, it's
• 39:30
  in the model's hands. Any AI tool that we're making
• 39:33
  right now with AI in human, you know, interaction like we're
• 39:37
  we're not we're not selling it at the moment like it's just
• 39:39
  being given away for. Free.
• 39:40
  It's like we spin it up. It's not like, you know, it's
• 39:44
  just a feature test. It's like, hey, do you like what
• 39:45
  we just built here? We're not charging for these
• 39:47
  things or look at these as like, you know, our 10 mil AR, you
• 39:50
  know, next golden goose. It's like what's sticking?
• 39:53
  How? Like, how do you actually get
• 39:54
  this to a point where you know it's worth something?
• 39:58
  Jeremiah is going to laugh at me now because we just had a call
• 40:00
  about the different things that I'm working on, but I didn't
• 40:03
  tell him about this. But I've got another brand and
• 40:05
  I've got this branding agency that I'm working on it that
• 40:08
  we're going to be launching soon.
• 40:10
  I don't know what sort of magic they're doing to generate these
• 40:13
  images. But they're doing such a good
• 40:15
  job. I'm like falling in love with
• 40:17
  our AI models. Is this a shirt that's been
• 40:20
  pasted onto AI model? Yeah, yeah, yeah.
• 40:22
  The shirts and the shorts zoom. All the way in on the font.
• 40:26
  Let's look at it. Yeah, that's it, right.
• 40:29
  Can you see it's a little bit messed up here, but.
• 40:32
  What's it actually says? Customer care right?
• 40:35
  Does the customer care? I don't know right?
• 40:37
  Is this possible for the customer?
• 40:39
  They did this like AI think two months ago, maybe a month and a
• 40:43
  half ago. So the models are getting better
• 40:45
  and I imagine the consistency is going to get better and maybe we
• 40:48
  should be delaying some of this work for a little bit later so
• 40:51
  that it can be more accurate. But I don't know if you don't
• 40:54
  zoom in and you start here. That's the thing, right, because
• 40:59
  features on a person are like you don't no 1-2 people ever are
• 41:03
  perfect clone of themselves. Forget the twin thing.
• 41:06
  But just like even a hair, like you're not going to have the
• 41:08
  exact amount of hair. So it's it's like you can have
• 41:10
  really good AI models, but if your logo is your logo, it
• 41:14
  should never change. Right.
• 41:17
  That's what drives me nuts is like when you have the AI
• 41:20
  product get changed, the AI models look in Stellar, but you
• 41:24
  just don't know because there's no reference of knowing if it's
• 41:26
  an AI at this point. Yeah, 100%.
• 41:29
  I think that like especially in a service capacity, you can't
• 41:31
  deliver that sort of thing as a deliverable to a client.
• 41:33
  Like what the hell? No, our logo is messed up.
• 41:36
  Me, I'm like, I don't think customers are going to care.
• 41:39
  Let's run it right. Hey, this looks awesome, but
• 41:41
  yeah, you know, that's. Like the Olympics thing when
• 41:44
  they like used AI for that and the logo got overlapped, but
• 41:47
  they're like the hardest sticklers about like their
• 41:49
  trademarks and like their bran kits and their styling.
• 41:51
  It's the Olympics. But then like the logo colors
• 41:56
  were wrong because the AI nobody caught it or somebody just like
• 42:00
  maybe they won't care and they just went past like creative
• 42:02
  review it. Is it is interesting, Yeah,
• 42:07
  because she was. I don't know exactly what your
• 42:08
  software products are, but the, the thing that I find really
• 42:11
  challenging right now is that a customer can now connect their
• 42:18
  open cloud bot to our API and we'll just use no commerce as an
• 42:22
  example. They can connect to the API,
• 42:23
  they can pull the data out they want, and they can build a
• 42:26
  dashboard to show that information however they want
• 42:29
  and for themselves, right? It's a, it's a single use case.
• 42:33
  They're doing it for themselves. They're doing it exactly the way
• 42:36
  they want it to be for themselves.
• 42:38
  The problem when you're building software is you're doing it for
• 42:41
  1000 people at the same time. And so you can't just let it be
• 42:46
  whatever somebody wants it to be.
• 42:49
  There's obviously a way you can kind of get to this, but the the
• 42:52
  challenge ends up being it's, it's the same thing we're
• 42:54
  talking about here. Like it's, you have to have this
• 42:56
  consistency. And then when you take it and
• 42:58
  apply it to a single use case where you are the person who is
• 43:02
  in control of it, you may not care if something's off a little
• 43:05
  bit in this data here or this chart doesn't look quite the way
• 43:08
  you want it here or whatever that is that you can make it.
• 43:10
  You can craft it to be exactly what you want it to be, or at
• 43:14
  least 95% of that. And you're OK with that 5% gap.
• 43:17
  But that doesn't work when you're trying to figure out how
• 43:19
  do you make this work across 1000 different customers at the
• 43:22
  same time. And I, I think that's the the
• 43:23
  thing that's extremely hard for all of us who are building
• 43:27
  things at scale. Whether you're serving clients
• 43:30
  or building software, whatever it is, you can't use the same
• 43:33
  methodologies that you can just go vibe code something once.
• 43:37
  That's given me so many more ideas of like, that actually was
• 43:41
  soothing for me. You saying that stuff.
• 43:44
  I'm just like thinking about all the I just have a lot of
• 43:46
  thoughts now. I like that.
• 43:48
  That's a really good way of saying it.
• 43:51
  I think that's the the the customer's main gripe with
• 43:53
  Facebook ads manager, right? It's responsible for so much of
• 43:56
  the economy as to serve so many people.
• 43:58
  So it can't be perfect for anyone.
• 44:00
  It has to be like possible, usable for so so many people,
• 44:06
  you know, income, a bunch of vibe coded ad launches.
• 44:09
  Yeah, people can find 1000 ways to spend some or 1000 different
• 44:14
  use cases for a thing. Like yeah, it could be mock ups,
• 44:18
  it could be for accompanying presentation, maybe it doesn't
• 44:20
  matter like when. But when we're looking at it,
• 44:22
  it's like we're trying to solve 1 specific problem that we know
• 44:26
  are same like, you know, catalog of thousands of clients have.
• 44:30
  And it's like we know we have these thousand clients and we
• 44:32
  know there's going to be plenty more that want to solve that
• 44:34
  exact issue. But to Jeremiah's point, like
• 44:38
  it's like, well, what may not, what may seem like an issue
• 44:41
  isn't necessarily always an issue for maybe another 50,000
• 44:46
  people and they don't care. And maybe that's the difference
• 44:48
  between adding, you know, exponential more revenue made.
• 44:53
  That's why open AI is open AI, right?
• 44:56
  So it's up for self interpretation.
• 44:59
  But I really like the Super like dialed in, you know, what's the
• 45:04
  one solve? I like the one solve things too,
• 45:07
  because that's where I can get the exponential value out.
• 45:09
  And then how do I, you know, connect that to another thing
• 45:11
  where it's like, well, this does this part, then this does this
• 45:13
  part, and layer those all in together?
• 45:16
  I like that, yeah. There's two ways of looking at
• 45:19
  it. I I like your point of view
• 45:20
  though, Jeremiah. That thought process seems
• 45:22
  bigger to me. It's like it seems more big
• 45:25
  picture thinking. Yeah.
• 45:27
  It's like how, how do you enable both?
• 45:29
  I think that's the question I'm I'm trying to ask myself.
• 45:31
  It's like, what does that look like to let a customer do it
• 45:34
  however they want to do it, but also let it work in sort of the
• 45:38
  base case for for everybody? I don't know the answer to that
• 45:42
  yet, but it's not simple. Yeah, you just rewired how I
• 45:46
  think on that. That is a balance strike, right?
• 45:51
  You know, Jeremiah was talking the other day about my Shopify
• 45:54
  app and a big part of my Shopify app is building for people who
• 46:00
  don't want unlimited options, right?
• 46:02
  Most analytic stores out there are flooding you with thousands
• 46:04
  of numbers that are like begging for your attention.
• 46:07
  And the reality is you don't need to pay attention to 1000
• 46:09
  different numbers. There's just a few that are
• 46:12
  really important to use to educate good decision making,
• 46:17
  right. So it was, it's a process of
• 46:18
  filtering down information as opposed to, you know, letting
• 46:23
  you create 1000 different types of graph with all these
• 46:26
  different customizations. You know, Shopify, any analytics
• 46:31
  tool you can think of is already doing that for you, right?
• 46:35
  And they're all in competition with each other for feeding the
• 46:38
  appetite for a never ending, you know, desire for more numbers,
• 46:43
  right? But yeah, I think, I think that
• 46:46
  sometimes making things simpler for people like you said Chase
• 46:49
  and solving a specific problem really niching down is I hope in
• 46:55
  this case helps people, right? Yeah.
• 46:58
  I, I just look at it like, because we're vertically
• 47:00
  integrated, like somebody will come to us like at 0 and then
• 47:04
  we'll be working with a company that's doing $100 million a
• 47:07
  year. They have two totally different
• 47:10
  objectives, two way different problems.
• 47:12
  It's, it's incomparable. But at every stage, we're still
• 47:16
  there all the way through. And like, a brand will come to
• 47:19
  us that's like, you know, eight years ago.
• 47:20
  That's now finally at the $30 million stage all that time
• 47:23
  later. But we don't want to like, you
• 47:26
  know, start there and then all of a sudden, like, they're not
• 47:28
  going to come back at the end. We want to be at every stage.
• 47:30
  And so it's like we're each bottleneck at every layer.
• 47:34
  What is the, like one key thing that people run into as they hit
• 47:37
  each like, you know, point? Is it inventory forecasting or
• 47:41
  is it a conversion rate? Is it a creative team issue?
• 47:43
  Is it a hiring thing? Like we want to know all those
• 47:45
  like every detail and we know the stages as people are hitting
• 47:49
  them. And then like, you know, going
• 47:51
  from 30 million to 100 million on online is a lot more
• 47:53
  challenging than going to 30 million to 100 million just
• 47:55
  penetrating retail real quickly because you already have the
• 47:58
  social virality. So it's like we're always
• 48:00
  looking for the fast pass there too.
• 48:02
  So I always look at, I was like, where's the one big problem that
• 48:05
  starts to really pile up? And then I just want to go
• 48:07
  figure that one out. I guess we can wrap up.
• 48:09
  I, I, I just have a curiosity question for you.
• 48:13
  Like what's the weirdest space, the weirdest problem you found
• 48:16
  that you guys have approached? Like have you approached how do
• 48:18
  we tackle tariffs? Like when that comes up, is that
• 48:21
  like a space that you never thought you would deserve?
• 48:23
  Policy and stuff like that. I don't try to touch anything.
• 48:27
  That's like way too dynamic that can be like messed with by
• 48:30
  policy. I, I tend to stay away from like
• 48:34
  we're, we're, we like to stay really compliant on things, but
• 48:37
  like tariffs and stuff, I don't try to enter in a space like
• 48:39
  that and say, oh, we need to solve this and find a workaround
• 48:41
  because I know that stuff flips so quickly.
• 48:44
  Anything that's like political or has like, you know,
• 48:47
  government involved or policy like that stuff is so wishy
• 48:50
  washy. Even with fulfilment.
• 48:52
  I don't like touching fulfilment.
• 48:53
  I don't ever want to get into that.
• 48:54
  Shopify did, it was a horrible mistake.
• 48:56
  They tried to, you know, build out all these warehouses and do
• 49:00
  the inventory management for their clients.
• 49:02
  That was a horrible disaster. They fired the guy who came up
• 49:04
  with the idea. They lost a ton of money doing
• 49:06
  it. TikTok just tried to do the
• 49:07
  exact same thing. Everybody's freaking out.
• 49:09
  What do we do? They're going to force everybody
• 49:11
  to then have their products sent to the TikTok warehouses.
• 49:13
  They're going to manage fulfilment because they want to
• 49:15
  compete with Amazon's warehousing because Amazon has
• 49:17
  their shipping on lock and TikTok wants to compete with
• 49:20
  them. So they're going to do it right.
• 49:21
  And I was like, it was another horrible move in my mind.
• 49:24
  I was like, there's no way this is happening.
• 49:26
  And it is and everybody's panicking and I was like, I'm
• 49:29
  not even going to go into this because I just know this is
• 49:31
  going to be a disaster. Literally a week ago there they
• 49:34
  sent out an apology letter and says after all the feedback we
• 49:36
  got, this was a horrible idea by the TikTok team.
• 49:38
  We're not going to force this anymore.
• 49:41
  It's like I anything to do with policy or like really, really
• 49:44
  big stuff around fulfilment or logistics.
• 49:45
  I don't like to get near because I know it changes at the flip of
• 49:49
  the switch. Right.
• 49:51
  And like, imagine building for something and then all of a
• 49:54
  sudden the next day they're like, never mind.
• 49:56
  Yeah, Yeah, good point. Hey, yeah, I was talking to
• 49:59
  someone recently that's building an ad launcher, you know, one of
• 50:01
  the Seas of people building a meta ads launcher to launch ads
• 50:06
  a little bit faster or not use media buyers or Vas or whatever
• 50:09
  it is. And I said that I think that the
• 50:12
  the game here in in software is you've got to build to solve a
• 50:16
  really expensive problem the best as best as you can.
• 50:20
  And then you've got to work on distribution because it's not
• 50:22
  just having the best product, it's also about like actually
• 50:25
  getting people to use it. That's half that's most of the
• 50:28
  game. But you've obviously got to
• 50:29
  solve the problem incredibly well.
• 50:31
  It has to be an expensive problem, not a really cheap
• 50:33
  problem that you know has already solved.
• 50:34
  For the most part, every organization that's got their
• 50:38
  shit together right, but we're. Taking a really fragmented space
• 50:42
  and consolidating it. Yeah, yeah.
• 50:45
  Good shot. Like, think about how many times
• 50:48
  you'll bounce in between AD libraries.
• 50:49
  Just go to four plane, it's all right there.
• 50:52
  That's a good shot here. We need to get.
• 50:53
  We need to get Google in. We need to get Google in.
• 50:56
  Yeah, where's Google? But dude, this was great.
• 50:59
  This is a good. Conversation.
• 51:00
  Well, this is exciting. Yeah, this is great.
• 51:02
  I'm sorry that Rashad can't come, but I think we had, you
• 51:05
  know, a great conversation. I think it's AI, think it'll be
• 51:08
  a good one. We do more of these.
• 51:10
  I like these. These are fun.
• 51:11
  Talk soon dude. Bye.