Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / Why the Best Marketplaces Build Their Own Logistics
Transcript
- Lucas: So a few years back, Amazon started delivering its own packages. And a lot of people thought it was just about cutting costs. But the real story is more interesting. Luna: Because Amazon was already profitable on its core marketplace. Why add the headache of running a delivery fleet? Lucas: Exactly. And that's the question a lot of marketplace operators face eventually. Do you stay asset-light and let third parties handle logistics, or do you bring it in-house? The answer depends on a concept economists call the tipping point of vertical integration. Luna: I remember reading that Amazon's shipping costs were growing faster than revenue in the mid-2010s. That must have been the trigger. Lucas: Right. In 2014, Amazon was spending about 10 percent of its revenue on shipping. By 2018, it was over 12 percent — and that's billions of dollars. When you're that large, paying FedEx or UPS per package becomes more expensive than running your own network. But the key insight is that shipping volume needs to hit a certain density before the math flips. Luna: So what's the number? How many packages per day do you need before owning the trucks makes sense? Lucas: There's no universal rule, but analysts often cite a threshold of around one million packages per day in a metro area. Below that, third-party carriers are usually cheaper because they have density across many customers. Above that, you start leaving money on the table. Amazon crossed that threshold in several major cities by 2018. And then they launched the Delivery Service Partner program — basically franchising the last mile to local entrepreneurs. Luna: Which is a clever way to stay asset light ish. You don't own the vans or hire the drivers, but you control the operations and the data. Lucas: Exactly. And that control becomes a competitive advantage. When Amazon owned the last mile, they could offer same-day delivery on millions of items. That's hard to do when you're dependent on a carrier that's optimizing for everyone. Luna: So the question for a marketplace founder becomes: at what scale does logistics ownership start to generate a return on investment that beats just improving your software? Lucas: Right. And that's where a lot of founders get it wrong. They see Amazon or DoorDash and think logistics is the moat. But for most marketplaces, especially in early stages, logistics is a distraction. eBay is the classic counter-example. They never built their own shipping network. And they're still a hundred billion dollar marketplace. Luna: Because eBay's marketplace is fundamentally different. It's peer to peer, mostly non-new goods. The shipping expectations are lower. The buyer and seller can negotiate it. Lucas: Exactly. So the decision to build logistics should depend on three things: the homogeneity of the product, the delivery time promise, and the frequency of purchase. If the product is standardized — like a book or a phone case — and customers expect it in two days, you probably need to own the logistics eventually. If the product is unique or one-off, like a vintage watch on eBay, logistics can stay commoditized. Luna: Let's talk about DoorDash, because I think they're the most interesting example of a marketplace that started with logistics as the core value proposition. They didn't build a marketplace for restaurants, then add delivery. They built the delivery network first. Lucas: Right. DoorDash's insight was that the hardest part of food delivery is not matching supply and demand — it's the logistics itself. Getting a driver to the restaurant, waiting for the food, navigating traffic, and arriving hot. So they built a logistics engine first, then layered the marketplace on top. That's the opposite of most marketplaces. Luna: And it worked because the unit economics are driven by density. Every order in a given neighborhood makes the next order cheaper. So DoorDash had to own the delivery to capture that density benefit. Lucas: Exactly. And they took it a step further with DoorDash Drive — their white-label logistics service. They realized the logistics engine itself could be a product. So now other businesses, like Chipotle or your local bakery, can use DoorDash's network to deliver their own orders, even if they're not on the marketplace. Luna: That's a powerful moat. Because now you're not just a marketplace — you're the infrastructure for an entire industry. But it also means you're asset-heavy in a way that makes investors nervous. Lucas: Yeah, and that's the tension. The market rewards asset-light models with higher multiples. DoorDash trades at a lower price to earnings ratio than a pure software marketplace like Etsy, because investors see the delivery fleet as a fixed cost. But DoorDash's argument is that the logistics is their competitive advantage, not a liability. Luna: So how should a founder think about this? Let's say you're starting a marketplace for furniture delivery. Do you own the trucks? Lucas: I'd say no, initially. You need to validate that the marketplace works first. Rent a few trucks from a partner, negotiate a per-delivery fee, and focus on getting supply and demand to match. Only when you have consistent volume in a few zip codes do you consider buying a truck. Luna: And that's where the tipping point comes in. You calculate the break-even: at what volume does owning the truck cost less than renting it per delivery? If that volume is achievable in six months, go ahead. If it's two years away, stay asset-light. Lucas: Right. And one more thing: you need to think about strategic control. If your marketplace's core promise is speed or reliability, and the third-party carriers can't deliver that, then you might need to build logistics even if the unit economics aren't perfect. Because the alternative is losing customers to a competitor who does control the logistics. Luna: That's basically what happened with Amazon. They started losing Prime members because two-day delivery was slipping on some items. So they took control. Lucas: Yep. And now they're delivering over 4 billion packages a year through their own network. That's a moat that's incredibly hard to replicate. If you're a competitor, you can build a better website, but you can't build a nationwide logistics network overnight. Luna: If today's deep dive on marketplace logistics was useful for what you're building or running, that's exactly why we do this show. These conversations take time to research and produce, and the way we keep them ad-free is through listener support. If you'd like to help, you can find us at buy me a coffee dot com slash fexingo. Lucas: Yeah, and that support really does make a difference. It lets us spend more time on episodes like this one, where we can really get into the numbers and the strategy. So if this helps you, a coffee is a great way to say thanks. Luna: And I should add — the buy me a coffee dot com slash fexingo link is just that. No tiers, no commitments. It's a simple way to contribute if you find value here. Lucas: So back to the framework. Let's push this into a practical checklist. If you're thinking about building logistics for your marketplace, ask yourself: is your product homogeneous? Do customers expect delivery in under three days? Do you have at least 500 orders per day in a single metro area? If you answer yes to all three, it's worth a deeper analysis. Luna: And if you answer no to one or more, focus on reducing friction in the marketplace itself. Better search, better reviews, better payment flow. Those improvements will give you a higher return on engineering time than building a logistics network you don't need yet. Lucas: Exactly. And one final example. Instacart — they started as an asset-light marketplace connecting you to shoppers who would go to the store and deliver. But over time, they realized that controlling the delivery slot was key to customer retention. So they built their own logistics, in a sense, by hiring full-time shoppers and eventually building the Instacart platform as a white-label logistics for grocers. Luna: Right. They're now essentially a logistics company that also has a marketplace. And their valuation reflects that shift — higher than before, because the logistics gives them a stickier product. Lucas: So to wrap up: the best marketplaces don't build logistics for the sake of it. They build it when the strategic value — speed, reliability, control, margin — exceeds the cost. And they do it at the right moment, when density is high enough to make the math work. Luna: It's a classic build-versus-buy decision, but with the twist that the 'buy' option today might be the 'build' option tomorrow. The key is knowing when the tipping point arrives. Lucas: And that's the kind of decision that separates marketplaces that become platforms from marketplaces that remain features. We'll explore another operational lever next time.