I'd be interested to see how well this would work out.
Furthermore, it's clear that it's not just "pickup" and possibly "delivery"... the author is making a clear buy/sell/bargain decision on the spot... what happens when the pickup results in a fraudulent item?
Clearly this means that whoever is driving around needs to be trained, and once trained, can be your competitor (I see few barriers between a trained pickup-person and the author scanning craigslist)
Kudos to the entrepreneurial spirit here, but I highly doubt it's scaleable.
Eventually, you scale to having multiple drivers in multiple cities who spend their work time on the road, with the "back office" alerting them to new potentially valuable postings. Maybe even tracking their general location, and alerting them to lower-value postings that happen to be near their current location?
Given the entrepreneurial nature of the buy/sell/bargain decisions that need to be made on the spot, maybe it should be structured something like a franchise? You sign drivers up, they get a territory, and get sent alerts for their their territory. Several people mentioned paying the drivers hourly wages; instead, you could charge them (per month or per alert?) for the alert service, or you could do some sort of percentage profit sharing (though you'd have to trust the drivers to accurately report all transactions).
It is possible to pull it off, it's just going to take a lot of work finding the right detail oriented/trustworthy person in another city that's in the right situation to pull it off. Hopefully in a year I'm writing about how I was able to scale it to multiple cities.
Thanks for the note!
of course, it can't analyze the data as efficiently, but it would definitely be faster to have an automatic filter that saves time in looking through obviously overpriced or useless postings