to be more precise, it's not that switching cost is zero, it's that the marginal cost of switching is small. the initial switch into ridehailing is a huge cost, particularly around opportunity costs, but adding another app is a relatively small difference.

it's not negative though. you still have the cost of signing up, taking your car to get evaluated, getting trained, figuring out how/when to get paid, etc. it adds cognitive and accounting overhead, and you run into contention issues between the apps, creating negative rider feedback that you must balance against the increased income potential.

i do agree generally that there is little defensibility in ridehailing, and that's why uber (and lyft) tried to employ its supplier power to create competitive advantage/barriers. for instance, uber eats wasn't an expansion so much as a flanking move to lock in drivers on uber by increasing utilization.