This seems like the simplist way. The company could take a sample of 100 Amazon parcels, analyze the marginal cost of delivering each (how full was each van, how much did it add to each route, etc), and use that as the basis for price negotiations with Amazon.
If that means Amazon pays more than other customers, fine.
Anecdotally, your claim doesn't seem correct, but I'm not sure what the exact rules are.
from the document
I know this because as I type this I’m sitting in a house in a rural area of upstate NY that doesn’t have USPS delivery. When clueless e-commerce merchants send me things via USPS anyways they never arrive and eventually get sent back.
This situation is not uncommon at all.
If what is being said is actually true, then amazon doesn't have any leverage at all. (again if what is being argued is true) These last mile providers might be taking a small hit to deliver amazon packages, so why deliver them at all?
If you answer is: "if they don't deliver these they won't get amazon's business." It would appear they are not getting any of amazon's profitable business (2/3 of delivers), so who cares about getting amazon's crap business?
It's a bit of a catch-22, they can reform their route and coverage to similarly profitable state with less coverage, potentially dooming them in the long run, or they can take the near 0 profit contract from Amazon and stay competitive while having almost no cards at the negotiation table. Amazon knows their business is huge and needed by these groups and is not afraid to point that out.