Hey Adam Thanks for the article. I’m assuming that you are referring to a one-to-one delivery model where one food item (or product) is delivered from the production point (kitchen, restaurant, commissary) to one customer. What some On Demand companies (including Uber & Caviar) have been doing have been many-to-many, where they pick up many dishes of the **same item** and deliver them to many people within a geographical area (say SOMA in SF) based on an algorithm that optimizes the distance travelled from one customer to the next during lunch. (Btw, the food is kept heated (and safe) in thermal bags).
Since you’re picking up many items (of food) only once from the production facility, there are efficiencies you gain for # of Tasks Completed / Hour (all else being the same as one-to-one).
Did your analysis above also cover this model? If so, what are the metrics of one-to-one versus many-to-many? Thanks!