Week 6: Platform economics

The subsidy-war years, and why venture-funded growth makes labour cost the one line item everyone treats as adjustable

What a subsidy war actually buys#

Uber Eats’ subsidy-heavy expansion against DoorDash and Grubhub spent enormous sums making orders artificially cheap for customers and artificially rewarding for restaurants to list — not because any of the three expected loyalty in return, but because market share, once captured, is expensive for a rival to win back. The spending eased off once the market had consolidated enough that another dollar of subsidy returned less than finally trying to make money, not when any of them became profitable. Uber is a company listed on the New York Stock Exchange (NYSE: UBER), which means its annual 10-K and quarterly filings are public disclosures, not estimates — its Delivery segment’s results are the closest thing this week has to primary-source financial evidence, rather than reporting about reporting. China’s Meituan–Ele.me subsidy war ran the same play a few years earlier, on a larger scale — worth knowing the pattern repeats, not just once.

One order, laid out in full#

Delivery fee, restaurant commission, customer discount, rider pay — put side by side for a single order, it becomes visible which party actually absorbs the cost once a price war’s subsidies are withdrawn. That’s the week-6-checkpoint exercise: build that breakdown yourself, from a published rate card, rather than taking anyone’s summary of “who pays” on faith. Do the same breakdown for a second, differently-priced order if you can — comparing the two is usually more revealing than either number alone, since it shows you which line items scale with order value and which stay fixed regardless.

Why rider pay is the flexible line item#

A restaurant’s commission rate is set by contract and hard to move without losing the restaurant. A customer-facing price is constrained by what a competitor charges. Rider pay per delivery is neither — nothing structurally stops a platform adjusting it order by order, region by region, in response to supply. That’s not malice; it’s just the one variable in the whole equation without an external party contractually defending it, which is exactly why it absorbs the pressure the other three line items are protected from. It’s also why every regulatory mechanism studied in weeks 7–10 — however different their legal machinery — ends up converging on the same target: giving rider pay the kind of external defender the other three line items already have.

Before the session#

Build the cost breakdown for the week 6 checkpoint: one order’s fee, commission, discount, and rider pay, each sourced from a published rate card.