Matt Levine's latest Money Stuff column does a good job explaining why a lot of smart people are trying to figure out a market-making model for homes (see companies such as Opendoor):
People want to apply the market-making model to homes. This makes sense. Buying or selling a home is a long slow uncertain annoying process. The value of immediacy is high, especially for a seller. If you decide to sell your house and go to a website and spend 10 minutes filling out a form and then someone wires you cash for the value of your house, that is much much much better than hiring a broker and listing the house and holding open houses and so forth. You’d be willing to pay a market maker a lot for that immediacy. (By selling your house to the market maker at a discount.) And if the market maker is good at acquiring houses, then it will have a lot of inventory, which will make it a good seller of houses. If you want to buy a house, you will naturally go to the market maker’s website, because it’s where the houses are.
Levine also explains why a market-making model is that much more difficult for homes compared to things like stocks. In a slowing/slumping housing market, it's pretty easy to lose money as a market maker. (That is, unless you can somehow accurately predict that a slump is coming.)

