You motivate the discussion by a comparison by two near-identical products.
In reality, even in your examples, what would matter is more the absolute judgement of quality (or comparison to entirely different alternatives).
I think thus even your slippery slope type Ice Cream ingredients dynamics mostly doesn’t happen that way.
Example Ice Cream: If you want to sell ice cram you’d like for most people to like it as much as possible (trading off with input costs, of course). And even the tiny drop in quality has its corresponding probability for one consumer here and there to eventually not buy your ice cram anymore. And as manufacturer you’ll obviously consider that (barring any principle agent where you as contracted manufacturer are trying to trick the actual owner, which may well happen).[1]
Similarly this can be said to apply in +- any realistic utility scenario, barely leaving any room for the supposed puzzle you bring
Note this does not mean we don’t have suboptimal quality. But that’s not primarily linked to any supposed ‘slow but near-continuous decline’ dynamics I’d expect.
You motivate the discussion by a comparison by two near-identical products.
In reality, even in your examples, what would matter is more the absolute judgement of quality (or comparison to entirely different alternatives).
I think thus even your slippery slope type Ice Cream ingredients dynamics mostly doesn’t happen that way.
Example Ice Cream: If you want to sell ice cram you’d like for most people to like it as much as possible (trading off with input costs, of course). And even the tiny drop in quality has its corresponding probability for one consumer here and there to eventually not buy your ice cram anymore. And as manufacturer you’ll obviously consider that (barring any principle agent where you as contracted manufacturer are trying to trick the actual owner, which may well happen).[1]
Similarly this can be said to apply in +- any realistic utility scenario, barely leaving any room for the supposed puzzle you bring
Note this does not mean we don’t have suboptimal quality. But that’s not primarily linked to any supposed ‘slow but near-continuous decline’ dynamics I’d expect.