The End of the Driveway Beats the Delivery App: The Economics

Some of what follows is roadmap, not product. I’ll tell you which is which — the fee math is real today; the order-ahead future it points to is labeled as vision below.

Why are food delivery fees so high?

Because point-to-point delivery is genuinely expensive. When you order through a delivery app, a driver makes a dedicated trip: to the restaurant, then across town to you, then back. One trip, one order. That driver’s time, gas, and the platform’s cut all ride on your single meal — which is how a $12 order becomes $25 after service fees, delivery fees, small-order fees, surge pricing, menu markups, and the tip. The fees aren’t (mostly) gouging. They’re the honest cost of a fundamentally inefficient shape: one vehicle, one order, one house.

I say this as a paying, grumbling customer of these apps: the product is miraculous and the math is miserable. But my gripe isn’t that delivery apps exist — it’s that we’ve forgotten there’s an older shape with radically better math, and it’s been circling our neighborhoods playing music since before any of us were born.

The two shapes of getting food to a house

Point-to-point (the delivery app): every order is its own expedition. Cost scales with orders — each one needs its own driver-minutes and miles. Batching barely helps; even “stacked” orders detour one customer’s dinner to serve another’s.

Batched-by-geography (the truck): the vehicle goes where many customers are and they come the last thirty feet themselves. One pass down a street can serve every interested household on it. Cost scales with trips, and a trip’s cost is split across everyone who buys. The more neighbors participate, the cheaper each sale gets to fulfill — the opposite of the delivery-app curve, where demand spikes make your fees worse.

That’s the whole argument, honestly. Everything else is arithmetic. Say a truck’s evening pass through a neighborhood costs some fixed amount in time and fuel — split it across two customers and it’s steep; across twenty and it’s pocket change per order (illustrative math, but the shape is what matters, and the shape is undefeated). Meanwhile the customer covers the famous last mile personally, at the end of their driveway, in flip-flops. The last mile is the most expensive leg in all of logistics, and the truck model gets it for free.

So why did the inefficient shape win?

Information. The delivery app knows where you are and what you want; the truck never did. Point-to-point dispatch plus perfect information beat batched efficiency plus total blindness — not because the shape was better, but because visibility was. The truck’s model was better economics wearing a blindfold.

Which tells you exactly what happens when the blindfold comes off.

What exists today, and what’s vision

Today, shipped: Snack Signal gives the batched model its missing information layer — residents raise the snack signal, trucks see confirmed demand and drive to it. The truck that comes down your street because six households summoned it is already beating the delivery app’s economics: no dispatch cost, no platform fee stack, no last mile. You pay the price at the window, mostly in cash for now.

Vision, clearly labeled — none of this is in the app yet: the endgame is order-ahead: your raised hand carries your actual order, the truck arrives with it bagged, and you collect it at the end of the driveway, paying by tap-to-approve from your account. At that point the comparison gets almost unfair — delivery-app convenience (order from the couch, food arrives) at truck-model costs (batched trip, no last mile, no fee stack). Not because anyone subsidized anything, but because the shape of the trip is simply cheaper — and the vendor keeps more, too.

The neighborhood dividend

One more thing the fee math misses. Delivery-app dollars leave: to the platform, to a driver criss-crossing the whole metro. Truck dollars mostly stay with a small business whose livelihood is your ZIP code — a vendor whose kids might go to your kids’ school, whose success means your street gets better service next summer. The efficient shape isn’t just cheaper. It’s local in a way an app-dispatched expedition can never be.

The end of the driveway was always the best delivery address in America. It just needed the truck to know you were standing at it. 🍦


The full picture this points to: Beyond the Jingle · No truck required: Your Neighbor’s Cupcakes