[blarghyblarg]

Not to be argumentative, but... that's the second option. Algorithmic. The initial implementation would be simple. One point, two points, three points. Then, some categories turn out to have significant change requirements, some change infrequently, some changes turn out to be very important in some categories, some categories need immediate human review...

[bogwog]

I'm sure Amazon can commit more resources to solving this problem than some random individual on HN can commit to a ~3 sentence long comment.

If they wanted to solve this problem, they would have already. Clearly, they don't care. People buy products with good reviews, and Amazon makes money when people buy products.

I wonder if they've been sued over this? I'm not a lawyer, but this sounds like false advertising to me.

[rwalle]

Yeah, so many people are wasting their time on those useless arguments. What matters here is not how to solve the problem -- customers should not worry about that at all -- but whether Amazon is commited to address this problem, and the answer is a clear no.

[blarghyblarg]

It's armchair quarterbacking, for sure. Nobody at Amazon is going to read some random news article, even if they are here slacking at work, and say "That's the ticket! Lets do that!"

They either fix it, or they don't.

[gowld]

> People buy products with good reviews, and Amazon makes money when people buy products.

and lose money (or make bad sellers lose money) when people return products -- often times Amazon won't even ask you to ship back a product you complain about.

[bombcar]

The vast majority of the crap with this problem just gets thrown away. A very small percenrage bother returning it.

[lozenge]

This is like saying that stocking groceries is complicated because some items are perishable, some are frozen, some sell faster than others, they go to different parts of the store. Isn't it just a fundamental part of what the business is meant to do? To have reviews that actually relate to the product?

[blarghyblarg]

Yes, it is. And in a grocery store, how is a product handled when it comes in on a shipment and doesn't match the previous shipments and/or bills of lading? By a person, in every single grocery store everywhere on the planet.

Although, in a grocery store, you're less likely to see a product change from a bicycle to a scooter. You're more likely to see the cocoa content in your chocolate bar drop, and the oil change to a cheaper and less flavourful type.

I guess the comparison still works... should we have purchasers comparing products before setting them out? Under what conditions? What conditions trigger a review? In my experience, the stores just keep putting changed products out like nothing has happened, even if we notice over time that things just don't taste the same sometimes.

Turns out that... yes... it gets pretty complicated pretty fast, but many grocery stores also seem to just be ignoring this issue.

[fn-mote]

> You're more likely to see the cocoa content in your chocolate bar drop [...] > many grocery stores also seem to just be ignoring this issue.

I do not believe this is the kind of swap people are complaining about.

We're talking about a box that's labeled applesauce and is actually full of rocks. Grocery stores would kill a supplier who was doing this.

I can't imagine a "real-world" situation happening at the scale of reputation-stealing that happens on Amazon. I suppose it's just an extreme version of "Made In Your Country Tools" building up a good reputation and then quietly outsourcing the work to "Low-Quality Overseas Forge".

Is the problem is that "easy to acquire" reputation for low-cost products (get good rep by using good materials, which don't really cost that much more and eat the cost as a loss leader) is easily transferred to higher-cost products? It's not like you even need to be making bandsaws to get the good rep., then start using it to sell cheap bandsaws? The investment at the start is very low.