Thread by @ProtagorasTO

1/ $DASH about to IPO for around $30B. On top of the profitability question, there's another one about the durability of that profitability (moats). Having launched 2 and 3 sided mktplces for Eats, I can tell you the latter was exponentially harder to build *and* manage.

2/ In food delivery, a 2 sided mktplace is a traditional aggregator. A sales team adds restos to the platform and marketing acquires users. Restos complete their own delivery, so the aggregator gets to skip all the messy parts and friction in the physical world.

3/ 2 sided food delivery different from ride hailing where it's viewed as an incremental earnings. If restos don't get any orders, that's OK because they're running their dine-in biz. This takes pressure off the aggregator, especially as they launch new markets and demand lags.

4/ 3 sided food delivery is an entirely different beast. It's the platform that is acquiring and managing driver supply, not the resto. Now the platform needs to manage all the messiness of the real world.

5/ First issue is what we call "mktplce balance". Demand and supply need to be at an equilibrium. If you have more drivers than orders -> earnings / hr go down -> drivers are mad and churn. Too few drivers -> ETAs go up -> orders don't get picked up -> users and restos pissed

6/ This balance is extremely hard, drivers are independent contractors and come and go from the platform as they choose. Snow, sports games, construction, etc. all impact supply. Building the systems, processes, on the ground teams, for drivers etc. is hard and highly defensible

7/ Finally, the actual execution of logistics is extremely tough. First, you need to accurately determine how long the resto will take to finish the order. Second, you then need to find a driver and tell them to get to the resto just as the food is finishing.

8/ Send drivers too quick -> earnings go down and restos are pissed because drivers ramming their store. Send drivers too late -> restos and users pissed because order is late / getting cold. The fact that $DASH can deliver McD's end to end in under <15 mins is a moat

9/ Pulling all of this off requires a deep focus on tech and Ops to solve complex problems. Building this new competency is hard and requires a seismic shift in culture (cant just hire these people in a vacuum) for 2 sided mktplces. Not impossible but definitely hard to copy

PS. 3 sided food delivery services have the dual advantage of removing inefficiencies out of the system and driving down costs through integration / centralization *AND* improving customer experience. Ie in 2019, some 2 sided mkplces had 45 min delivery times vs. 30 for 3 sided.