5 ms·
> Next they need to get rid of grouping reviews for different product versions. This is huge. It seems to be a common tactic for companies to get a product's r
by in_me_i_trust 10y ago
> Next they need to get rid of grouping reviews for different product versions.
This is huge. It seems to be a common tactic for companies to get a product's rating up to 4.5 stars, then replace it with a lower quality version. I notice only when I filter for the most recent reviews.
- wnevets 10y agoThere been many occasions where I didnt buy something because of this very reason.
- Alupis 10y ago> It seems to be a common tactic for companies to get a product's rating up to 4.5 stars, then replace it with a lower quality version What you are likely seeing is not a seller "pump-n-dump" scheme. Rather, it's more likely another seller came along and "attached" to the good listing - except they are selling a knock-off product of worse quality. Often these sellers are from overseas, but not always. Amazon rotates the "Buy Box" (the "Add to Cart" button) winner on a "percentage-share" basis - meaning if there are 5 sellers attached to a listing, and all stats on the seller accounts are equal (volume sold, number of positive feedbacks, account history, etc...), then each seller will get 1/5th of the buy box time. When certain sellers are "stronger" than others (meaning their stats), they get a larger percentage of the buy box time. You can see this sometimes if you refresh the page several times while on a product page. The "Sold by xxxxx" may change. So, it's likely some cheap knock-off seller has attached to a "hot" product listing, and simply based on Amazon's Buy Box system, they will get a certain percentage of sales. Unfortunately, this can ruin a perfectly good listing for the other sellers who sell the legitimate product, and is difficult to combat. It pays to pay attention to a seller's own feedback rating. Not just how many stars on average, but more importantly the recent 30 day trend for their feedback ratings, as well as the number of total feedbacks, since it will give an idea of seller volume (most sellers average about <10% of orders leaving order feedback, and usually <1% leaving product feedback - my company has sold on Amazon for the past 7 years, and we seem to stay right in line with what other sellers expect as well).
- abawany 10y agoExactly. I tend to filter by critical reviews (with the default sort of recent) and that often gets the noise down to a manageable level.
- chias 10y agoI think this is a slightly different issue, which first entered widespread public discussion in the SSD market. Basically Kingston (I think?) and a few others released a new SSD and used very high quality components in them, so that they exceeded advertised specification. After the early-adopter reviews came in with the benchmarks and whatnot, they altered the build to use lower quality / cheaper components. When discovered and confronted, their argument was that the lower quality builds still met spec, so it was appropriate to sell it under the old model number and listings, but the upshot was that they got to produce low quality products while benefiting from reviews and benchmarks that were run against a higher quality product. I don't think there's much Amazon can do about this particular issue... although yes it would be a great help if you didn't see reviews for the 900x version of something while looking at the 1x version page.
- ekianjo 10y agoSteam handles that problem by showing two scores for the item (games in their case): the total score, and the "past 30days" score. This way you can see if something is going wrong recently.
- chias 10y agoI love that mechanic. I think it's a lot more in the forefront with things like software that gets auto-updated, where the publisher pushes an unfortunate patch, or fails to follow through on a promise which was previously given the benefit of the doubt. I don't think physical products have yet caught up to the idea that this notion is relevant to them as well. The especially difficult part when it comes to reviews for physical products in that it just isn't sexy to benchmark an old piece of hardware. What makes the situation even worse is that most reviews aren't centralized (that is, you google "productname reviews") and so to a large extent the visibility goes to whoever publishes first. Your brand new benchmark announcement gets clobbered by the pagerank of the reviews that happened right when everyone was interested in the product.
- ropeladder 10y agoI had long assumed that Valve added the "past 30 days" aggregate to show when games had improved rather than gotten worse. (i.e. early access games) Kind of interesting that it helps clarify changes in both ways, and implies very different incentives for the game publishers/sellers.
- nommm-nommm 10y agoThis may also be a third party seller selling fakes.
- tesseract 10y agoThere are also Amazon listings where they have completely different products sharing the same page, using the UI that is normally intended for color/size/quantity selection. The reviews get mixed together and it can be difficult to tell which review is intended for which product. It would help if Amazon added fields to the review to indicate which variant the review was written for. Filtering would be even better.
- Pxtl 10y agoOr just boot terrible sellers off the Amazon store and stop being AliExpress.
- voltagex_ 10y agoThe problem is, the "terrible" sellers are back on Amazon the next day under a different name/brand.
- Pxtl 10y agoSo spin off a secondary site for the bazaar and protect the Amazon brand for selling quality stuff and only include the top-quality sellers on the main Amazon brand.
- makomk 10y agoThis is unfair to AliExpress. Unlike Amazon, they know better than to mix reviews for products from completely different sellers.