GrubHub charges a 15-30% fee, of which only 10% goes to the delivery person. On average the app owners are making twice as much as the person who does all of the actual work.
Seems like it. 10-15 years ago I had a bunch of paper menus in a folder taped to my fridge that I would have to look at to order food. Perhaps I didn't have access to every restaurant in the area, but the selection was broad enough. The only extra fee I paid was whatever tip I chose to give the driver which I always assumed got added on top of a reasonable wage. That system seemed to be sustainable--at least it operated fine for many decades.
It's too bad the old system was cannabilized by the apps. I'm sure some smaller operations that couldn't afford to run their own delivery before have seen an uptick in sales, but as a consumer the choice paralysis, screwed-up orders from over-worked kitchens, and insane fees/price bloat has turned getting food delivery into an unpleasant experience I now try to avoid.
I think the unprofitability is purely due to inefficiency. A lot of "gig economy" companies have taken the startup model of "don't worry about profit yet, just grow as fast as possible to get investor
s attention" and pushed it beyond reason. GrubHub has ~2700 employees [0], do you really need that many people to run a food delivery app that doesn't even employ delivery people? And then you get totally avoidable stupidity like pizza arbitrage [1]. There is no reason that most people on this site couldn't build an equivalent app for their city, and in fact people are doing just that [2].
