Latest / Marketplace Businesses with Fexingo: Two-Sided Networks, Liquidity, and Take Rates / Why Marketplaces Fail At The Last Mile
Transcript
- Lucas: If these conversations are useful for what you're building or running, consider supporting us directly. We deliberately don't run ads on these shows because we want the editorial focus to remain sharp and uncommercialized. If you want to support that choice, the link is buy me a coffee dot com slash fexingo. Luna: That keeps the signal clean, so thanks for making it possible. Let's dive into today's friction point: why delivery apps hit a wall when they try to go beyond the city center. Lucas: Right. So most marketplace listeners know the core loop. You need enough drivers to cover enough customers so that wait times stay under five minutes. That creates liquidity. But here is the thing that trips up even the best operators. Liquidity is not uniform across geography. It is highly concentrated in urban cores and thins out drastically in suburban or rural zones. And once you thin out, your unit economics break before you even think about marketing. Luna: So you are saying the problem isn't demand. People everywhere want food delivered. The problem is that supplying that demand becomes too expensive per order. Lucas: Exactly. Demand is actually pretty flat nationwide now. A pepperoni pizza tastes the same in Chicago as it does in Des Moines. But the cost to get that pizza from the kitchen to the porch changes wildly based on distance and density. In a dense neighborhood like Manhattan or downtown Austin, a driver can drop off three orders in a half-mile radius in twenty minutes. In a sprawling suburb with single-family homes spaced half a mile apart, that same driver spends forty minutes driving between drops that might only be two miles apart total. Luna: Which means the driver earns less per hour unless the platform subsidizes them heavily. And if the platform subsidizes them, who pays for that loss? Lucas: The customer pays through higher fees, or the platform eats it. But there is a third option that I see more often lately. They just stop trying. We are seeing major platforms essentially freeze their geographic expansion. Look at DoorDash. They have been incredibly aggressive in metro areas for years. But in the last eighteen months, their growth in new zip codes has slowed dramatically. They are not launching in new small towns anymore. They are optimizing existing ones. Luna: That feels like a defensive move. Are they worried about losing money in those new markets? Lucas: They are worried about burning cash for zero long-term value. Here is the concrete number that matters. In dense urban cores, the average delivery fee can cover the variable cost of the driver plus payment processing with room for margin. In low-density areas, the variable cost exceeds the average fee by fifteen to twenty percent. To make that work, you would have to charge the customer an extra four dollars per order. No one wants to pay that. So the marketplace chooses to let those customers go to slower, cheaper alternatives like traditional courier services or local restaurants that use their own drivers. Luna: So the liquidity pool dries up because the price point required to sustain it kills the volume needed to keep the pool liquid. Lucas: Precisely. It is a classic chicken and egg problem but reversed. Usually, you need volume to lower costs. Here, lowering the cost raises the price, which kills volume. And there is another layer to this that people forget. Driver churn. Drivers prefer dense areas because they can stack orders. When you move to a low-density area, drivers spend more time waiting in cars than earning money. So you lose your best drivers first. You are left with part-time gig workers who cannot sustain the consistency the marketplace needs. Luna: I remember reading data that showed driver satisfaction scores dropping significantly outside of top fifty metropolitan areas. It makes sense. Waiting around for thirty minutes between deliveries is boring and inefficient. Lucas: It is inefficient and it lowers the quality of service. And quality matters for retention. If your app takes forty-five minutes instead of twenty, customers cancel. Once they cancel, the merchant loses revenue. Then the merchant stops prioritizing your app. Then the cycle reverses. You lose the supply side, then you lose the demand side. It is a downward spiral that is very hard to reverse once it starts. Luna: Does this mean Instacart faces the same issue since grocery shopping involves heavier items and longer distances? Lucas: Grocery is actually worse because of the weight and complexity. You cannot stack grocery orders easily. A shopper for one family might take forty minutes to fill a cart. By the time they drive to the next house, which is two miles away in a suburb, the previous customer is already home. The driver utilization rate plummets. Instacart knows this. That is why they have invested so heavily in Shipt, their owned workforce model. They are trying to solve the liquidity problem by controlling the labor directly rather than relying on the open market. Luna: But owning the fleet is capital intensive. Doesn't that hurt their valuation multiples compared to asset-light models? Lucas: It does, but it also protects margins in the areas where they operate. You cannot scale an asset-light model into low density without breaking unit economics. So they accept the higher capex to secure profitability in their core regions. It is a strategic trade-off. They are choosing depth over breadth. They would rather dominate fifty zip codes profitably than spread thin across five hundred zip codes at a loss. Luna: That sounds like a much healthier business strategy than the old growth-at-all-costs playbook. Investors seem to be rewarding that discipline right now. Lucas: They certainly are. Look at the public markets this quarter. Companies that showed clear paths to positive free cash flow in their core segments are trading at premiums. Those still promising national ubiquity despite negative unit economics are getting punished. The narrative has shifted from reach to efficiency. And for marketplace builders, that is a crucial lesson. Your moat is not how many cities you are in. Your moat is how efficiently you serve the ones you have chosen. Luna: So density is the moat. If you can't build density, you can't build a defensible position. Lucas: Correct. And density is hard to fake. You cannot buy it with marketing spend. You have to earn it through operational excellence and consistent supply. Once you have it, it acts as a barrier to entry. Competitors cannot enter your dense zone because they cannot match your wait times without bleeding money. You win on logistics, not just on brand awareness. Luna: It reminds me of how ride-hailing works. Uber became dominant in cities not because they had the best app, but because they had the most reliable drivers. Lucas: Exactly. And reliability comes from liquidity. When you have enough drivers, you get shorter pickup times. Shorter pickup times mean happier riders. Happier riders mean more rides. More rides mean better earnings for drivers. It is a virtuous cycle. Break the cycle by expanding too fast into thin markets, and you get a vicious cycle of cancellations and driver exits. Luna: What about the role of technology in mitigating this? Can AI routing solve the distance problem? Lucas: AI helps, but it cannot change physics. Better routing can shave five minutes off a trip. It cannot turn a two-mile solo drive into a multi-drop efficient route if there are no other orders nearby. Technology optimizes within constraints. It does not create constraints. If the constraint is low population density, algorithms can only do so much. You still have to pay for the mileage. Luna: So the real bottleneck is demographic, not technological. We are stuck with the layout of American suburbs. Lucas: In America, yes. In Europe and Asia, where housing is denser, delivery marketplaces scale differently. They can achieve liquidity in neighborhoods that would be impossible here. That is why global comparisons are important. You cannot judge a marketplace model in isolation. You have to look at the underlying urban fabric. Luna: That explains why some European platforms feel more mature. They started in environments that naturally supported liquidity. Lucas: Right. And they evolved faster because the economics worked earlier. Here, we spent years fighting against geography. Now we are finally accepting the limits. And that acceptance is leading to smarter business decisions. Platforms are focusing on vertical integration. Instead of just connecting buyers and sellers, they are taking ownership of the middle mile. Dark stores. Micro-fulfillment centers. These allow them to bring inventory closer to dense populations, reducing the last-mile burden. Luna: So the solution is not to expand geographically, but to contract operationally. Bring the warehouse to the customer. Lucas: Yes. And that shifts the cost structure. You pay for rent and labor in the hub, which is fixed and predictable. Versus paying for miles and time on the road, which is variable and volatile. Fixed costs are easier to manage at scale. Variable costs kill you when volume dips. So the industry is moving toward a hybrid model. Asset-heavy hubs for efficiency, asset-light drivers for flexibility. Luna: It seems like the end game is a localized monopoly in every dense cluster, rather than a national utility. Lucas: I think that is exactly where we are heading. National brands will exist, but their power will come from local dominance. If you control the liquidity in your city, you control the pricing power. You don't need to be everywhere to be profitable. You just need to be indispensable where you are. Luna: Indispensable through speed and reliability, not just presence. Lucas: Precisely. Presence is cheap. Reliability is expensive. And that expense is the price of admission for a sustainable marketplace. If you can't justify that expense in your target zip code, you shouldn't be there. Simple as that. Luna: It forces a lot of humility on founders who thought they could copy-paste a successful model across the country. Lucas: It does. And it rewards patience. Building density takes time. You cannot rush it. You have to nurture the supply side, ensure the demand side is happy, and let the network effects compound organically. Try to force it, and the liquidity cracks. Luna: So the lesson for listeners building or investing in these spaces is to watch the zip code, not the nation. Lucas: Watch the zip code. Watch the driver hours per shift. Watch the cancellation rates. Those are your early warning systems. If they spike in a new area, you have a liquidity problem. Fix it or exit it. Do not hope it gets better with more marketing spend. Luna: Hope is not a strategy. Data is. Lucas: Data always wins. Especially when it tells you that some places are just not meant to be served by your model. Knowing when to stop expanding is just as valuable as knowing how to start. Luna: A powerful reminder that scale is not the same as success. Lucas: Exactly. Profitability at scale beats emptiness at any size. We will leave you with that thought as you evaluate your own growth metrics today.