We do a number of things to prevent it, but it ultimately comes down to providing enough value to the supply side of your market. If you build a product that drives real, ongoing value for your service pros, they will be much less incentivised to cut you out of the transaction. The same is true for the demand side.
Also, lawn care has a reduced risk profile compared to services that take place inside the home (maids/babysitting, etc). There is no key hand-off and the nature of the service is somewhat less intimate. Our customers don't need to be present for the work, which both reduces this risk and boosts the value of our platform as a means of managing your service.
For Homejoy, saying "we are bonded and insured" reduces reluctance for a first you or when on-demand service is the primary consideration. But for recurring service, having trust with an individual beats the crap out of being able to point to the contract or collect against a bonding agency if something goes wrong.
[aside: I found Adora Cheung's presentation for How to Start a Startup thought provoking: http://startupclass.samaltman.com/ ]
To incentivize that, you could offer discounts for neighbors who use the service together. If Homejoy had a bunch of apartments in one location, then cleaners would make more money per day.
Also, routing optimization is a fairly hard problem and requires significant critical mass in a geo. Pretty hard for individual service providers to build this on their own. The closest thing may be a small-town service pro who has been in business for 30 years and eventually reached sufficient mass / frequency to optimize his routes, but that's a rare case.