Subscribe
Share
Share
Embed
Build a Business Worth Buying brings you candid conversations with industry leaders, M&A experts, and successful founders. Learn advanced strategies to scale, optimize, and prepare your business for an acquisition—because building a business worth buying starts with smart decisions today.
Marc (00:00)
these LLMs are just tools. At the end of the day, it's still your job to do the thing that you're supposed to do. You got to tell the story, you got to build the product, you got to make this stuff happen.
Aaron Alpeter (00:43)
We have a really special guest today. I absolutely love this episode. was supply chain focused, it was acquisition focused, and just man, the density of what we were able to get into and the nuances was amazing. My guest today is Marc Held, and he has spent his career in parts of the economy that most founders never fully understand, but almost every business depends on, that's right, supply chain. So he was trained as an engineer, but most importantly, he's a builder who's repeatedly turned
complex supply chain problems in the companies that someone else is willing to buy. From tracking global container flows to predicting freight prices to optimizing inventory decisions for some of the largest CPG companies in world, has built businesses around one core idea. If you can turn operational noise into signal, you can create real enterprise value.
He has started and sold multiple companies, operated inside acquirers, run private equity strategies, and now invest across the next generation of supply chain infrastructure. Along the way, he's helped develop a clear and sometimes uncomfortable perspective on why companies actually get acquired and why many that look promising never do.
I really enjoyed this conversation. I think you will too.
Aaron Alpeter (01:46)
Marc, thank you so much for being on Build a Business Worth Buying. We met just recently, we had a call and I was just like, my gosh, I have to have you on the podcast. Just everything you've done, you are doing now. So thank you so much for being here.
Marc (01:59)
No, my pleasure. Thanks for having me.
Aaron Alpeter (02:01)
Just to get our audience warmed up with who you are and why I got so excited about you. Just want to walk through your experience and what do you know about building businesses worth buying?
Marc (02:11)
God. Yeah, no, I've done it a few times.
to give a little bit of context, I'm an engineer by trade. I've started and sold a few businesses in the supply chain intelligence space. So the first one, we were tracking shipping containers for shipping lines. Turns out the shipping lines don't actually know where their containers are.
we got to a point where we were making predictions about commodities
Coincidentally, and then we ended up getting acquired by one of our customers. They did another one in the in transit inventory optimization space, which is nerd speak for helping basically large BCOs, large shippers like Procter and Gamble basically figure out how to move inventory through their own supply chains faster. Because if you're someone like P &G, one day of inventory time is worth 80 million bucks. So if you shave off a little bit of time here, a little bit of time there, very easy sell, they're working with a bunch of brands. We ended up being acquired by a strategic.
the logistics focused TMS called Turvo ended up running data and IoT for them for a while, did some family office work, and eventually I got connected to some folks that were trying to figure out what to do with a freight brokerage out of Chicago called Cargo. Turns out they had built some interesting tech. They built TMS, they built a visibility tool, they built some cool stuff. And we spun out one of the pieces of
tech ⁓ into a business called GreenScreens, which is a predictive freight pricing business that just had a nice little outcome. So I've done it a few times. And I'm a fan of basically understanding a market, understanding the levers that get pulled within an ecosystem and figuring out how to get from something that exists to something that someone else wants. Because at the end of the day, that's whole point of a startup. You want to build something.
so that someone else acquires it or it goes public or there's some sort of liquidity event. So I've done it a few times and yeah, excited to nerd out with you about a bunch of hypotheses I have.
Aaron Alpeter (04:04)
I love it. It sounds like you always built something with the intention to sell. And I think like sometimes you meet founders where they kind of just fall into a business or they, you know, become so much part of their identity that's hard to let go. Were you always like, hey, I'm building this because this is effectively a product that I'm selling and the goal is to exit.
Marc (04:25)
Yeah, no.
The first one I did, wanted to build, actually the first company I ever did was not in logistics at all and I wanted to Jarvis from Iron Man and I just wanted it to exist and never went anywhere. But I think eventually I kind of came to that conclusion that if you build a product and there's no one around to buy it, does it matter?
And I think the same goes from the products that you're developing to the actual companies that exist. So kind of coincidentally fell into this kind of startup acquisition world through that first logistics company, because we had an unsolicited offer from one of our customers. And that just turned into a whole entire process of, how do I best create an outcome that makes sense for me, for the team, for the investors?
Yeah, not intentional. kind of just happened. And now it's kind of the lens that I see everything through.
capitalism
Aaron Alpeter (05:22)
Yes.
So when you're evaluating an opportunity, an investment, something else you might be cooking around in your mind, like how do you go about getting to the point where you're like, okay, not only is this a good idea and I can do it, but someone is actually going to want to buy it. So I'm creating a value here that someone else is going to be interested.
Marc (05:40)
Yeah, so fortunately I come from a very specific niche. I'm in supply chain. It's a very niche, but it's broad. There's lots of different applications. And just over the years, I've touched pretty much every level in the value chain. I've dealt with procurement, to manufacturing, to inventory, to transportation, to logistics, to...
Customs, you you pick something, I probably touch something both domestically and internationally. So I think just over the years, I've kind of built up a bunch of hypotheses within the ecosystem. there's a handful of reasons why companies get bought. I mean, I think there's like five, six, seven-ish. It's really, getting bought for your customers or revenue, for cost savings, for...
you technology or the assets that you have, the team, and then there's my favorite reason, which is so that no one else can have you. And I think that like the sooner you have those levers in your head as you're navigating, is this a company worth investing in or starting or partnering with? It just helps frame a lot of the, go to market. It frames how do I position myself? Like, do I want to position myself as someone who's going to take someone else's business? Do I want to position myself as, someone who
understands what the strategic goal of this public company is, and I'm going to fit right into here. There's just a million ways that you can kind of use these levers to understand how to best get an outcome. And I think that's not only necessarily in selling a business. I think it's also with respect to go to market, with respect to sales, just understanding what are the true incentives that the companies that you're working with have.
gets you to and this even has implications for like the current business that I'm at. It's a company called HOPTEK which is effectively a freight monetization platform. Turns out in trucking, like, there's really not a whole lot of great ways to quantify the strategy of a business. So you might have two companies that have the same number of assets that have the same types of revenue. But when you try to combine them, you don't necessarily know are they good fits. So like, kind of
have taken this hypothesis to the nth degree, not that like we're core built for ⁓ &A, it just happens to be another mechanism to quantify these trucking businesses.
Aaron Alpeter (07:54)
love that because I think that that framework that you shared in terms of why businesses get acquired is really helpful. I want to double click on that a little bit. How did you develop that framework? Is this something that you had to learn the hard way as you were going through and figuring out like, okay, I'm ready to exit, but really what am I selling?
Marc (08:11)
Yeah, yeah, no, I wish there's probably a book about this somewhere. I don't know. never I'm a school dropout. So like I didn't I do everything the hard way. I need to cut my teeth on a problem in order to understand it. going into that first outcome was, I did not have this framework is just kind of something I picked up by talking with all of these corporate development guys and folks that were trying to think about things strategically and just over the years, these are
things that hear over and over and over again, are you in a creative investment? Like, do I buy you and does it increase my valuation? Can I use that stock pop to pay for the cost of the actual exit? all of these crazy financial engineering humans kind of end up coming to this very specific set of things. So yeah, I wish I had a better answer for you. It literally is just stuff that I kind of picked up over the years by accident.
Aaron Alpeter (09:06)
Yeah, so I just want to go through one more time because I love it. was someone's acquiring you for your customers, your team, your technology. What were the other ones?
Marc (09:12)
Yes.
cost savings. ⁓
And then there's my favorite one, which is the strategic like so that no one else can have you or an optionality ego, you can call it whatever, but it's like the wild card. And that one is the most interesting to me, because the valuations are never linear.
So it's, it's, there's, yeah, it's completely, you know, can you craft a narrative? Can you craft a story where like you are the thing that's holding these folks back from doing something else or you are the reason why your competitor is going to beat you. like you get this interesting FOMO story that I always find interesting.
Aaron Alpeter (09:31)
never
Yeah, I love it. if I'm a founder or somebody who's thinking about wanting to build something that'll exit one day, do you suggest that people pick one of these and kind of build around that? Or do you always want to try to be the FOMO one? Like talk to me about how you take that framework and actually apply it into the act of building the business.
Marc (10:09)
think there are different strategies for building businesses with these in mind.
let's take the team play, for example, like you can build a great, fantastic team in an industry that, is acquisitive, but unless there's other strategic value for these customers, like you still need to understand what people want. like maybe you get a billion PhDs in
rocket science and it's a rocket science, rocket space vertical, like that's a clear slam dunk. Like SpaceX only has so many people that they can hire and if they can buy your company and get a bunch of talent versus going to hire all of these other folks themselves, like that's a slam dunk. That's easy. The valuation, I would argue, is probably not going to be as high as it could be, but if you can save a bunch of money for these folks and a bunch of effort, that's easy.
If you're thinking about customers, there's really two universes. this customers that the strategic acquirer doesn't already have or are these customers that they already have and they want to expand with it? So I think from my perspective, these levers are really just like the starting point as you're trying to figure out what are the strategic things that your universe of buyers care about.
that you should absolutely keep in mind as you're figuring out how to develop a product, a roadmap, a strategy. there are a million different strategies that you can play when you understand what your customers care about or what your buyers care about and how they think.
Aaron Alpeter (11:42)
I mean, you spent a big chunk of your career in supply chain and logistics, which is obviously near and dear to my heart. When you think about those types of deals or transactions, is one of those parts of the framework more common in supply chain than supposed to compare to consumer or tech or things like that?
Marc (12:01)
I think it depends on like what part of the hype cycle you're in. think if you're like now, if we're to take the AI cycle as the thing that everyone is investing in, if you've got a bunch of PhDs in computational neuroscience, like you're gonna get the best friggin valuation on the planet. Like 10 years from now, probably gonna be a very different story. 10 years ago,
No one cared. I think it is a pure function of like for the vertical that you're touching, where are you with respect to what is important? Is this a land grab opportunity? Is this a team grab opportunity? Yeah,
I think at the end of all of this, it's just, again, deciding for the big players, for the folks that can acquire you, what are they actively?
looking at. I think I found over the years that one of the great ways to look at this is if you're a public companies, you can literally look at their investor relations reports. Like every quarter, there's a document that will get published, usually, where you can effectively see how the CEO is thinking about things, you can see how the corporate development group is thinking about things, do they need to find more logos, do they need to find more revenue, are they looking for
new technology that they can go back to their customers and sell them more shit with. But I think having an understanding of the strategic roadmap of these companies will absolutely help you figure out, know, do I invest in this? Do I invest in that? And plus, these things change over time. So when you're in one of these super frothy venture things, like in 2021, I think you couldn't walk down the street without raising money in supply chain. And
that led to a lot of FOMO from an acquisition perspective. So there were a lot of companies that raised some money that did some interesting stuff. And this became kind of some fear that the larger companies that had raised money kind of got towards. So the small guys were doing interesting things and that created a lot of fear from the large folks and turned into some exits. And I think we're now starting to see the...
the price of some of these not thought through acquisitions. you know, that is the con of this FOMO based acquisition strategy. But I think for the startups, they probably made more money than they should have.
Aaron Alpeter (14:22)
I think that's true. I definitely have been on the opposite end of that where I started at a supply chain technology company around COVID. And I'm like, no, no, I don't need to raise money. Let's just do this bootstrap. And looking back now, I'm like, I could have had like $2 million like that, you know? And maybe I should have done that. But yeah, no, it's interesting. You've got to play the thing.
Marc (14:37)
Yeah. Yeah.
all these flash
in the pan moments like 2016 there was a lot of supply chain money, COVID there's a lot of supply chain money. I'm willing to bet like now with all of the funky stuff going on in customs, there's a lot of money for customs stuff. ⁓ But I think anytime you see something on the news that investors don't normally see is generally a good opportunity to try raising cash.
Aaron Alpeter (15:07)
Yeah, I love it. ⁓ Maybe you can walk us through one of your exits and break down what a buyer actually cared about versus what you thought they were gonna care about when you first started having those conversations.
Marc (15:20)
I'll talk about the inventory optimization company. We had developed a little tracker that was super cost effective that you could put onto pallets. That's what made it so that we could actually get this in-transit inventory optimization feasible. We needed to have effectively tracer bullets inside of the supply chain so that you could track a pallet all the way from the manufacturing plant to when it was sold at a retail facility. There's a couple of hops that happen along the way.
You've got P &G making a product here, another product here that gets sent to a mixing facility. ⁓ that palettes get built, sent to Walmart, who will then distribute those products to the final retail. So again, multiple hops. ⁓ I guess the traditional way to see what's going on inside of the supply chain is to integrate all of the different players that exist. the warehouse, the WMS here, the TMS here,
we should screw it, we're just going to track the pallets and see the stuff from end to end. And so we got to a point where I think this was 2016, 2017, and there are all of these supply chain startups that were raising huge amounts of money. And our investors just didn't want to play that game. They wanted us to have an outcome. So we ran a process, we met a bunch of folks. And for each one of the folks that we were talking with, know, supply chain visibility was the buzzword of the day.
So for us, was, we lean on that specific thing so that we can create a little bit of FOMO? And we did, and it ended up that the company that acquired us didn't actually, I mean, they cared in theory around what we were trying to do internationally and the customers that we had, but I mean, they really bought us for the tech. They didn't really care much about the team. They didn't care much about the customers. In fact, that was actually a really painful thing. I had to go call
⁓ our customer and cancel a very large contract, was very, you know, not
Aaron Alpeter (17:21)
That seems like counterintuitive for someone to buy something and then say, yeah we don't want, we don't like money.
Marc (17:27)
mixture of the customers that they were currently serving. They were mostly working with trucking folks. it was a pure FOMO. So they just wanted the other folks that we were talking with to not have us.
Aaron Alpeter (17:39)
Got it, interesting, okay. That's so fascinating, because I think
when I see deals that come across my desk, the first thing look at is, okay, how am gonna pay for this? How is this gonna result in more profit, more revenue, things like that? I did have a deal recently where I was like, all right, you know what? Maybe we're not gonna do this. Maybe we're gonna say no to these opportunities in order to focus on something else. And so it really is interesting, because...
Most people will look at this and say, well, your valuation should be a multiple revenue or multiple EBIT. But when you're looking at actively subtracting something from what you're presenting, how do you bridge that gap? Because from your perspective, you may say, hey, I want a multiple revenue because I've got this big chunk here. But they may look at this and say, yeah, that $2 million your client that you have is nice, but we're going to get rid that on day one. And so we're going to discount that, even though you are expected to be compensated for it.
Marc (18:30)
Yeah, no, that's a tough one. And this is again, goes back to the strategic kind of why are you being acquired? So I mean, if you care about having that revenue in your thing, and the revenue is a reason why you're being bought, like absolutely make sure that your buyer cares about it.
Aaron Alpeter (18:45)
You know, one of the things that I hear a lot from founders or executives is that data is their moat, right? Everybody just believes that just by having bits on a server somewhere that that's really good. But you've worked with supply chain data, you're an engineer. What actually turns data into something useful, something that a buyer would value versus just a bunch of storage and things that nobody sees?
Marc (18:53)
huh.
Yeah, so that's an excellent question. I mean, having built a few data businesses over the years, there are lots of moats that can be built. think if you have, I think, was it? magic phrase is now dark data. I don't know if you've heard this term. Now that we've got all of these LLMs that are doing stuff and being trained on stuff, anytime you have access to data that no one else has,
I think that's when you actually get a moat. I think there's not a whole lot of moating that exists when you're talking about large amounts of data or anything that you can get when you're talking about scraping or gathering ⁓ somewhat public data. And I think it is also a blurry line in logistics where some of the data is kind of public, but also kind of not. And do you have the rights to do anything with it? So all things to kind of keep in mind as you're figuring out the data strategy.
I think data for data sake is not super useful, again, unless it fits directly into someone else's strategy. Like if you're doing something in domestic transportation and you're working with an international business, probably not super relevant unless they're actively trying to get into domestic transportation. So again, I think a lot of this fits around where do you fit in the ecosystem and...
who can leverage the stuff that you've got in the best possible way. And maybe it's one of these like one plus one equals 11 scenarios where like the stuff that you have and the stuff that your buyer has turns into this brand new thing that would not traditionally be feasible. Does that make sense?
Aaron Alpeter (20:45)
It does, it does, and it kind of prompts a couple of their interesting questions because I feel like there must be a hierarchy to data, right? And, you know, there's certain data that's gonna be more valuable for a business on its own as well as to an acquirer. And I'm wondering, how do you think about maybe those spheres of influence in terms of, yeah, you know, nobody really cares about what you had for lunch three weeks ago, but they do care about pricing data or sales data or technical drawings or things like that.
Marc (21:14)
I would argue...
It's going to be weird. I would argue that all data is valuable to somebody. I think you just got to figure out, you know, what is the signal that you're able to pluck out of that noise and who is it worth anything for? So like your example, like what did I eat for lunch three weeks ago, might not be useful to a financial services company, but might be useful to, you know, wonder or one of these companies that are trying to figure out what are the tastes of people and how do I...
develop new menus such that the stuff that we're creating gives me more market share. So I think it's just about shifting the framing of what data do you have and what is the signal that can be extracted from that. mean, even this, again, this very specific use case of what did I eat for three weeks ago, I think that is not only a signal of what are you actively trying to eat, but
That's also a byproduct of what was available. I did you go out and purchase things? Like, I think if you kind of trace back the things that exist within the data that aren't actually in the data, that's when you start to get into this really interesting, greater than the sum of its parts universe.
Aaron Alpeter (22:31)
In your experience, do companies kind of fall into the signals and the data they have, or are they very intentional that we're going to collect this data because it gives us this signal?
Marc (22:42)
I think it's a byproduct of the type of company that is being built. If it's like a workflow automation tool, I think data is something that generally is on the back burner. If it's a little bit more of a data forward product, I think that's, you you're doing yourself a disservice if you're not thinking immediately of the data and all of the use cases on day zero. as the industry matures, again, and this is mostly in supply chain, as the industry matures,
I think people are starting to realize the power of this data and the things that can be created from it. So if you're not thinking about the use case of your data and what you could do with this exhaust, like that is an asset. Like if you're building a TMS and you're not tracking usage behaviors, like you're absolutely looking at the wrong stuff. And just for those who aren't in supply chain, a TMS is kind of like a system of record for folks in transportation. think of it as kind of like a database.
And so it's one thing to actually have the store of that information. So is the shipment going from here to here, but that's just like part of it. There's all of this other stuff that you could be looking at. Did someone ask for quotes from here to here? That could be a signal that there's going to be demand in this area. Are you seeing more bookings now than you have been? There's just all these secondary and tertiary things that...
that I highly recommend folks to just think about in the back of their heads and make sure they're not losing ⁓ that data as they're building products.
Aaron Alpeter (24:09)
That's so fascinating. I guess other than being able to better predict why people are going to your tool for answers, what would be a benefit of tracking usage data in terms of, this person is a VP or a manager or they're in this department who logged in and checked things?
Marc (24:25)
I mean, of course there's the internal aspect of like, how do I build a product that serves my customers better? Can I monitor, they're doing this, they're not doing that. Maybe we should invest more in this product area or that one. think quantifying that always nice. But I think just having this usage data could also be a proxy if you get enough of it for like other broader things. So if all of a sudden we can see a spike of people searching for
freight between Shenzhen and Long Beach, that is a leading indicator before someone actually goes to book something. So if you can see a few weeks ahead of when that event were to happen, you can sell that data to someone.
Aaron Alpeter (25:09)
Yeah, that's fascinating. we've been talking a little bit in terms of like acquirers who are interested in that data, but you're kind talking about even as an existing business being able to sell that signal, obviously you've to be careful about what you're able to do and things like that. But I feel like so many people just have noise in their data and they have a Google Drive or a database that has lots of stuff in it.
Marc (25:29)
Yes.
Aaron Alpeter (25:34)
that stuff doesn't really mean anything. It's not telling a story. how do you actually extract signal from all of that noise that's there?
Marc (25:41)
I think that's a little bit of an art. I don't know if there's like a framework for folks to figure out how to do that, but I think it just takes, I don't know, maybe take some mushrooms and read data. don't know. Yeah, I think look at the shape of the data that you have, talk with some smart data folks. I don't think people even need to know what specific like the records look like.
but just the shape of it. you is this a interaction record and you're looking at what part of the system they're in versus other stuff. If you were to have a bunch of smart people look at that shape, think folks will eventually recognize that there are additional things that you could do with that data. ⁓ I just highly, highly caution folks from, making sure that they're tracking as much data as they can.
Because it's better to have it and not need it versus need it and not have it.
Aaron Alpeter (26:44)
I think a lot of people, especially if you're a product founder, your data is gonna live in Google Drive, right? Or if you're a tech product, it's gonna live in a database somewhere. How do you actually operationally manage and control that data? So one, that it's being identified correctly and stored correctly, but then two, you're able to go back and actually know what you have and ask questions of it later without saying, oh, you know what?
Marc (26:54)
Mm-hmm.
Aaron Alpeter (27:08)
we were so concerned about focusing how many wheels were on the road that we forgot to ask is it a car or a motorcycle?
Marc (27:14)
Yeah, I think the structured versus unstructured data problem is a hard one. I think when you're talking about this Google Drive scenario, when you've got literally gigabytes, terabytes of stuff in Google Drive and it's thrown all over the place and you have some document ontology, like this is my finance folder, this is my quotes, this is my sales, like all of that stuff, that becomes really, really tough to figure out what to do with.
I think this is one of the nice things that we have with LLMs is that they will absolutely not take breaks. So you can, yeah, yeah, you just throw an LLM at it and be like, yo, what can I do with this? In fact, I, the other day, someone gave me a bunch of files and I literally threw it into Claude and I said, Hey, write me an investment memo based on all of this information. And it did an absolute fantastic job. And traditionally, like you have teams of
Aaron Alpeter (27:43)
Go learn.
Marc (28:04)
folks that are doing this, you've got analysts and interns that are trying to scrape through the data and try to figure this stuff out. I think the barrier of being able to understand what to do with your data has just gotten much lower than it has been in the past.
Aaron Alpeter (28:19)
I agree with you. think sometimes you have a lot of people who are building things for the sake of building and you don't know, is it actually good data? Is it good thoughts or assumptions that are based onto it? And so I feel like,
it's never been easier to go out and use LLM to soak up vast stores of data. But I think the bar is therefore that much higher to actually have something that's useful. And so I'm curious to know, like, from your perspective, when someone is going through that noise and they're starting to pull out signal, it's nice to know what someone ate for lunch a couple of weeks ago. But when does a signal start to actually influence the pricing power, the risk, or the financial outcome in a way that could
Marc (28:40)
Sure. Yeah.
Aaron Alpeter (29:02)
really spell a big benefit from evaluation point of view.
Marc (29:06)
So there's a nerd answer to this and then there's like a human answer to this. The nerd answer to this is, you have a hypothesis and you've got stuff that you're going to use to prove or disprove that hypothesis. And when you're analyzing data, like you can see the statistical significance of that one piece of information with respect to a model from an output. So like,
Aaron Alpeter (29:09)
All right, near first.
Marc (29:26)
it
to see whether or not there is signal within noise if you're a nerd, like.
Life is easy. ⁓ If you're not ⁓ a nerd, ⁓ it becomes a little bit more complicated. people have been building analysis on Excel spreadsheets forever. I think the process is still similar where you, have a hypothesis and you want to either prove or disprove it. And I think you can build models within Excel or,
one of the Power BI or one of these tools to see if you can get a curve to fit some sort of shape. It's just a little bit more challenging versus throwing a k-means clustering algorithm and doing ⁓ all sorts of funky analysis.
It helps if you have the tools. Frankly, again, this is something that Claude can absolutely help with or OpenAI, any of these LLM-y things. I use the best 200 bucks a month I spend is literally on Claude Max or whatever the $200 one is.
Aaron Alpeter (30:36)
Yeah. Anthropic, if you're listening, we do take sponsors on. So I think I saw it OpenAI about a podcast for 200 million dollars. We'll give you a great deal. That'll be great.
Marc (30:41)
Hahaha!
There
you go. There's the strategic pitch. You're the answer to that. ⁓
Aaron Alpeter (30:54)
Yeah, no, I love that. mean, I feel like so many times people are, they're building something and they're like, look at this cool graph. But then it's also like, what is this graph actually telling or teaching or encouraging me to do? let's just kind of dive a little bit deeper into like the tech side or the design side. It is easy to have lots of data. It's easier.
to show visualizations and just throw a lot of stuff on the screen. I find that it's really hard to take stuff away and to be very clear and say, hey user, this is what you should do. Here's the conclusion you should draw and like helping them guide through all that.
Marc (31:25)
Yeah.
Yeah, this is the art in data science where like anyone can analyze the data, but to actually extract a story from it, I think takes either some expertise or some perspective. Like I almost feel bad for kind of the next generation of engineers and data scientists where like a lot of folks are going to be throwing lots of dollars at Anthropic
will be doing the same thing that you as a data scientist will do. It just happens to be that if I am someone who tells these stories for a living, it's much easier for me to go to this bot to have it do its stuff versus train up a whole entire team of people to do it. So yeah, I think it's gonna be really tough over the next few years for folks to kind of build that muscle of
I've got all this data, I've got all this analysis, how do I tell a story from it? What's my perspective? How am I thinking about the utility of this data for folks? ⁓
Aaron Alpeter (32:20)
Yeah.
Well, you know, I think there's gonna be a big shakeout, right? my wife is in marketing and we were having a conversation about how with all the generative AI, it's really easy to create.
videos and ads and copy and things that you can spin up an entire campaign in in an hour and something will have taken you thousands of dollars and weeks to create is just done instantly. And so as a result, everybody's doing this and it's all starting to look the same. And so the parts that are starting to separate at least from a marketing point of view is going back to basics. And can you tell a story? Can you be authentic? Can you do these things that were always probably the most important part of marketing in this case? Or are you just trying to do whatever
Marc (32:32)
I didn't.
Aaron Alpeter (33:00)
else is doing and you're just trying to put stuff out because you know you're told to put stuff out.
Marc (33:05)
Yeah, I always tell my team that, you know, these LLMs are just tools. At the end of the day, it's still your job to do the thing that you're supposed to do. You got to tell the story, you got to build the product, you got to make this stuff happen. It's just if you're at war, why would you be using rocks and sticks when everyone else is using a bazooka? And that's what these LLMs are. it's a pure arms race for, the war that is capitalism. So
If you're not using it, you're being left behind. But again, it still requires someone to go point and shoot and say, hey, this is my enemy. This is what I got to do. So yeah, totally agree with you. the bar to creating stuff of value has gotten super, super low. You just got to figure out what it is that that value is. There's only a finite amount of time, finite amount of resources to do this analysis, to do
these exercises and if you aren't able to steer the the LLM to the outcome that you're looking for, then you're screwed.
Aaron Alpeter (34:04)
It's also is like these things are not infallible, right? mean, ChatGPT and Claude say at the bottom of every session, Claude can make mistakes, right? And we kind of have a channel in our Slack group where we point out the mistakes that Claude make. And sometimes, you know, it's just like, yeah, I just made up that number. And you're like, don't do that. Like, and so I think that there's, there's this piece here where
Marc (34:14)
Thank you.
Aaron Alpeter (34:25)
A couple of years ago, it was magic, right? I remember talking to a friend of ours and she was like a technical writer. I was like, oh, well, look at this cool thing called Chat GPT. she's like, ask it about this really interesting, complex medical thing. And it came up with a blurb about it. She's like, I don't have a job anymore. I'm done. Like I said, initially it was like scary because this was going to replace everybody. But the reality is that I don't think AI is gonna make people lazier because those lazy people are gonna be found out, exploited.
and like shut down. Like I think that that's just what's gonna happen. And so it's really gonna help you do what you've been doing better. If you were already creative, if you were already analytical, if you're already technical, you're gonna be able to do that more. And it's not gonna take somebody who wasn't those things and suddenly make them the best worker on the planet.
Marc (34:55)
brain equalizer.
And I think this also goes back to a lot of the hypothesis around ⁓ &A and startups. the best product doesn't always win. I think it's a byproduct of a lot of marketing. It's byproduct of lot of stuff. Like these are all just tools that exist in order for you to solve a problem. So you've got these LLMs that can help you generate these ⁓ marketing campaigns faster. Like you absolutely should be using them.
it's all about the user of these tools to help figure out what do I need to be doing with this technology in order to drive to the outcome that I want. So if it happens to be that the value that I can provide to a strategic acquire is purely from creating an LLM-based wrapper and it just requires me to...
go sell it to a bunch of folks and then it gets scooped up, then like you absolutely should do that. Otherwise, yeah, use the tools to your advantage to figure out whatever it is that the acquirers use. mean, hell, even use the chat bots to look through the investor relations documents. Like I would be willing to bet, like you could probably say, hey, who are the acquirers in this ecosystem? Like what does these folks care about?
It's like use the tools. They're basically interns. yeah, interns that don't fight back.
Aaron Alpeter (36:31)
Yeah, had a fight back, I love that. I wanna kind of shift over to something else, because now you're doing venture investing, and you can go lots of other stuff you're doing. And when we met last month, you made a comment that really kind of struck with me, resognated with me. And that was that venture often optimizes for the next round, not the exit. And so I'd love for you to go into more detail about what you mean by that and how you see that showing up in companies today.
Marc (36:40)
Yes.
Sure, so for those who don't live in venture, usually when you have a startup, you can go and raise some money, say, hey, give me a piece of your company, I'll give you some cash. And the economics of these things are such that the investors that are investing in your current round really only care about your next round. And that's really just a byproduct of the fact that they, from an economics perspective, need to have a write up.
and a write up is just a term for when your investment has increased in value. So let's say you've got a company and you've valued it at $10 million and you raised a million dollars, you've given up 10 % of your company. And if a year later, let's say the company valuation is $20 million because you've raised some additional money and they valued your company at that, the original investor gets to go to their investors and say, look, like,
I made a bet, I was right, this company is worth more, give me more money. And so it's this constant, you know, back and forth where these cycles happen in a much shorter time period versus an exit, which usually venture capitalists own, I think they assume it's like a seven year outcome. So the economics of these venture funds are such that they want you to raise your next round as quickly as possible at the highest valuation. And like, it would be nice to have liquidity, but like, that's not the primary
objective.
Aaron Alpeter (38:23)
you would think that as a founder, you and the venture investor, your incentives are aligned, right? You're looking to grow the value of this company. But, you know, I've had conversations with founders where they took too much money too early and they're like, I've got a deal. can sell my company for 50 million bucks, but I just raised it a $75 million valuation. And like, they're not gonna let me do this. I have to keep going. And it's gonna be a $500 million outcome. like, I don't know how I'm gonna do that. So.
Marc (38:36)
yeah.
Aaron Alpeter (38:49)
It's such an interesting dynamic here where it's not a one for one just because you're both in the cap table your incentives may not be fully aligned
Marc (38:58)
Yeah, this is one of the weird little dichotomies or funky bits of the ecosystem where I think another byproduct of how venture funds are structured is that it's very much a power law distribution. So the top funds get most of the returns and one huge unicorn deal will return the fund for everybody else. incentives are not necessarily aligned where like
you're to sell your company for 50 million bucks and you have most of the business, that's like a life changing amount of money for a person. But for most venture capitalists, that's like a loss. They do not want to see that they want to see 500 million or a billion like, again, it's a function of the size of the fund and the strategy and blah, blah, blah, blah. But like most of the time, the economics are not necessarily aligned across the people building the companies and the people investing in them. And that's a tough
pill to swallow, I think, for a lot of folks. I have a lot of friends that are in venture, I have lot of friends in startups, and anytime I have the ability to tell someone, don't raise money if you don't need to, I'm absolutely going to tell them that. Because once you get on the venture treadmill, it's hard to get off.
Aaron Alpeter (40:12)
Where do see venture companies making decisions that actively reduce the likelihood of having a successful exit?
Marc (40:20)
I think there's sometimes some short sightedness that happens. And this is the same thing with public companies and public markets. You know, I am here trying to get a markup versus an outcome. Like if I can go not raise this round, but sell at three to four X, whatever my initial investment was like, that's a fine outcome. It's not a return the fund outcome. they're optimized for getting to that next round so they can go out and raise more money. It's just this vicious cycle.
I'm going to invest in you so that you can make my money worth more, but I don't really care about the outcome yet. I might care at some point.
Aaron Alpeter (40:54)
well, you know, five or six years ago, it was very common for founders to make most of their money on secondary too. So now you have another vehicle where, the VCs may not be aligned, but the founder's like, hey, yeah, I don't care if this thing sells either because I just, trying to take a couple million bucks off the table. And so then you're an employee and you're like, what's going on here? I thought we were trying to keep things going.
Marc (41:01)
Yes.
Yeah.
Brad Feld has a really good book called Venture Deals. If you haven't read it and you're in startups, like please read it. It will make your life a lot better.
Aaron Alpeter (41:24)
Yeah, if you can have, learn lessons, not scars, that's better. That's great. If you were to start a company today, I know you've always got ideas and things cooking, and you were gonna start with the explicit goal of selling it in the next five to seven years, how would your approach to capital be different today than maybe a traditional venture back path?
Marc (41:28)
Mm-hmm. Mm-hmm.
So I think that for most people, angel investors are the right investors, period. You won't necessarily get as much money as going through traditional venture, but angel investors, they can provide resources, they can provide capital, they can provide unfair advantage, especially if you're within a relatively complicated industry. Let's say the next thing I wanted to do was in customs brokerage.
Right? I think the universe of venture capitalists that understand customers brokerage is not very large. I would go out of my way to find some angel investors, some families that are in the space that can help me close deals, that can help me get in front of customers, it can help me, you know, do the thing. And frankly, they're not going to be shooting for that unicorn outcome. They're going to, you 3X outcome is great. It's fantastic.
You know, can I 3X my money in five years? Great, yes. I'm going to do that all day, every day. Because what is it? The standard is you double your money every seven years at 7 % or something. I can triple it in five years and help. great. That's definitely someone's strategy. in fact, like I think more people should be taking angel checks and I wish that more angel investors existed. So one of the things that I personally am trying to do is
talk to all of my friends in supply chain to write small checks, but aggregate a lot of them, because I think there's a lot of benefit that supply chain people that aren't necessarily seen as angel investors can actually do to support their businesses. Can this guy who works at this company open the door by connecting him, the startup to someone else?
It's worth more than the $5,000 check. The deal gets done.
Aaron Alpeter (43:34)
Well, you know, it's really interesting because I feel like angels are the right fit in some cases, but they can also be the ones that get crushed if that startup then goes the venture route and they just, you know, they have basically no upside there. And so, you know, do you find it more common that that companies are committing early on and say we're going the venture route or going the angel route and they're basically sticking to that for the most part?
Marc (44:01)
I have seen startups maybe do a small friends and family angel round and then get on the venture treadmill.
I think the universe of companies that I've seen that have done just angel is not very large. ⁓ I think that's a shame. I think more people should do just angels and not raise a lot of money and then get an outcome. ⁓ But it's definitely something I don't see a lot. And granted, this is just in my personal experience with the startups that I've seen and been involved
Aaron Alpeter (44:35)
I have conversations with folks all the time. And I used to do quite a bit of angel investing. And granted, I wasn't very sophisticated. I'm better now than I was, but I'm probably nowhere near as close as you are. looking back, those were lot of donations. were, yeah, was a vote of confidence. I like you. I like this idea. I feel like this should exist. And that was there. But the ones that did do really well,
Marc (44:49)
Yeah, yeah, you should never expect to get your money back.
Aaron Alpeter (45:00)
end up going the venture route and end up with like lots of point, 0.000 something. You're like, well, that's great. I might get my money back even though you had this great outcome. And so the person I haven't been doing much angel investing at all, I just kind of do more venture state because it's like, all right, these guys are already working toward an exit. I know I'm to get my money back in two to three years. And that's been a pretty good fit for me.
I kind of want to just last kind of questions we wrap up here. When we think about companies that look really promising on the outside, maybe they've got lots of press, they've raised a bunch of money, they've got great customers, the revenue's up there. What sorts of things tend to fall apart in due diligence or at late stage conversations with buyers where it's like, you know what, this is, there's a lot of lipstick on this pig and this really isn't that great of a business.
Marc (45:45)
you would be shocked at the amount of deals that get done without proper due diligence. So there have been some sizable outcomes that have happened where diligence happened after the deal closed. And then you get buyer's remorse. Like it happens a lot. And I think that is a byproduct of the acquirer, acquiry, the person who has the company that is being acquired.
⁓ doing a really good job of creating a FOMO, Fear of Missing Out vehicle.
whether or not that's ethical, different conversation, but from a strategic perspective, like I think that's the way to do it. I think so long as you're not like miscategorizing revenue and doing things that
the acquirer you're probably in a good state. ⁓ is it? Rule number one for me is stay out of jail. Like everything else is secondary. So if there is a chance that,
⁓ I am saying something that is a little bit questionable. I'm not gonna say it. It's just not worth it.
Aaron Alpeter (46:50)
Well, Marc, this has been absolutely fantastic. Every time I talk to you, I learn so much more and I'm like, I want to talk again. This is awesome. ⁓ We've got one question we always like to ask our guests on Build a Business Worth Buying, and that is what is the best moat that you've seen another business build?
Marc (47:07)
So there's one that I'm obsessed about right now. This is a company called Genlogs. I don't know if you're familiar with them. They are a company that basically is putting cameras in all of these random industrial locations and they're monetizing the data that's coming out of those cameras. And like that to me is one of these things that like in retrospect makes so much sense. I don't understand why someone hasn't done it. And they've been using it to like locate chassis that
like traditionally, you wouldn't be able to find or track individual trucks because they can look at the license plates. They can see a truck going from point A to point B and then they can aggregate that data. All of a sudden, you can start to see this lane getting a bunch of volume that you... It's just one of these insane pieces of data that I don't understand why someone hasn't done this sooner. Huge amount of respect to the GenLogs people. They're really smart folks.
there's a lot of these companies in supply chain that I think are sneaky. Selfishly, I think there's another company called Pascal Tags. ⁓ They basically are making a chipless RFID company. That is an interesting moat in that you're effectively building RFID tags at barcode prices. So if you're familiar with the universe, traditionally you've got an RFID tag that's got a chip and an antenna. They're just an antenna. They figured out how to print it so you can put it on anything.
which for all intents and purposes means that you get a lot more data than you traditionally are able to do with barcodes or with just RFID. ⁓ That to me is one of these really scary, ⁓ holy shit data modes that no one else has. That's simply a byproduct of this technology. ⁓ GenLogs is a byproduct of the time. Pascal is a byproduct of the technology.
Aaron Alpeter (48:53)
That's amazing.
Yeah, that's awesome. Well Marc, if people want to talk to you, if they've got an investment idea, if they want advice, what's the right way for them to reach out to you?
Marc (49:08)
ping me on LinkedIn or you can send me an email, marc@cartel.vc my personal email address. So give me a holler.
Aaron Alpeter (49:16)
Love it. Well, thank you so much for being on this episode and thank you all for tuning in to this episode of Build a Business Worth Buying. I'm your host, Aaron Alpeter, and good luck building.
Marc (49:25)
Thanks for having me.