A Food Business without the real estate

Over the last 12 months, almost a dozen investors/well heeled folks have approached me to talk about a food business without the real estate. Essentially they want to explore the possibility of setting up a delivery only food business with a kitchen in a non-prime low cost location. This would make the capital expenditure significantly lower, make the operations easier (due to no customer footfalls in the real estate space).

While on a excel sheet this model looks reasonably attractive, I have been a little skeptical about this business model for the following reasons:

1) Buying food from a restaurant is a impulse decision. It is also typically not a "Necessity" but a "Nice" purchase - though it sometimes feels like a necessity given our hunger pangs. Because of this, reinforcing your existence becomes very critical - i.e. a customer sees your place everyday when he comes home and simply by seeing you everyday, he decides to try it out one day. Without a retail location, you don't get this visibility. So you will have to invest in extensive branding - bill-boards, flyers, etc. repeatedly. The cost of doing this is also quite high and may offset the rental savings to a large extent, atleast in the initial few months/years while your brand is being established.
2) Several customers like to visit a restaurant, try out the food and then start ordering food from there. With the above business model this option is not available. So they will have to be coerced to place a delivery order and try it out once - again this will require some marketing spend. This can be done by placing attractive offers during the initial few months to get customers to try it out.
3) The Real Cost of delivery is higher than it seems. My estimate is that the variable cost of delivery ranges between 15-20% of the order value - this includes the following costs: Staff, Vehicle fuel and maintenance, order taking and tracking system. While a lot of these can be optimized on paper, in reality streamlining delivery is not easy. In a discussion with a Domino's executive, he told me that they undertook a study to optimize the delivery process and the conclusion to the study was to have more delivery vehicles per unit.
The rentals for a restaurant in a high visibility location is typically 15-20% of the sales. So the delivery costs are not lower than the rental costs, but the initial capex and capital at risk is significantly lower.

Despite the above skepticisms, I think it will be worthwhile for folks to experiment with this format - and like I have mentioned several times, if your initial plan on paper requires "X" amount of funding, keep 2-3X handy to give you enough cushion and enough of a runway to try and take-off.