I know there are tens of millions of products on Amazon, but I wonder how many reviews a dedicated team of in-house of reviewers could do per day. Amazon could even charge for the privilege of having their (ostensibly) unbiased team review the product, or prioritize them based on sales trends etc. It feels like shoppers are already indirectly doing this by buying the in-house Amazon brands.
IMO, Amazon should consider hiring a team of professional reviewers to generate curated reviews for select products. I understand they couldn't/wouldn't review everything, but given the stakes of public perception, it appears it might be a good investment.
Would it be too costly for Amazon to just employ people to review most or all of their products, and maybe even rebrand the best products under its own brand where the seller agrees?
It would dramatically increase the quality of their store and brand, and it benefits from economies of scale.
Assuming they pay $100 for an high quality review, then a $10 item would increase by only 1 cent if at least 10000 are sold, which seem reasonable conditions.
For categories with many very similar items, reviews should be cheaper since a standard review procedure could be designed and applied quickly, offsetting the lower sales per items.
I think it only becomes a problem when the system gets big enough. If there's 20 different online storefronts, it's a lot more time and effort to game the reviews on all of them, than if there's only Amazon.
Fake reviews aside, it's core to Amazon.com's business model. Given Amazon's success I infer that a lot of consumers in general research products by going through reviews.
This used to be the space that paper magazines handled, and to a limited extent orgs like Consumer Reports still do.
"User" reviews are bullshit, and far too easy to game. And Amazon's system bodges together distinct products, so you can't even tell if you're reading a relevant review of the thing you want to buy or a different model/size/anything.
I sometimes wonder if an independent review site would be a business idea. Crap merchants would avoid it anyway, but anyone with a solid product would get a tear down, lab report, subjective review, and objective performance details, with a couple of follow-up purchases just to check the product hadn't been substituted.
Result would be high-value products standing out from the noise, with some objective reassurance for customers - which is the opposite of the Amazon model.
Well, if Amazon were to crack down on this sort of thing, it would incentivize unscrupulous sellers to solicit fake 5-star reviews (through a 3rd party) for their competitors so that they would be punished. The only reason we have to doubt this is happening now is that there's no incentive to do it, since Amazon seems to be turning a blind eye.
It's a hard problem. The solutions most likely to work are also expensive. Amazon could create internal product review teams staffed by employees instead of letting users do it all for free, but it's hard to make that scale. It's also less transparent; one of the big appeals of Amazon early on was the novelty that you could go to a retailer and if you don't like one of their products you can say so for the world to see. Now most other retailers do the same thing and we sort of take it for granted now.
I can appreciate that policing reviews at scale is difficult but why does amazon need to have 10,000 options for every single thing? If you want to create a system where the only differentiator is the number of positive reviews then you are basically creating the same monster that you are also attempting to sleigh.
The company makes 378 real purchases then mails 378 real products to random addresses scattered throughout the US. This allows them to leave verified reviews.
There's a whole industry around this - The Reply All podcast interviewed people who make these reviews. They talk about the economics of reviews and Amazon's reluctance to admit there is a problem.
In some categories, Amazon is putting more weight onto editorial reviews. It'd be a good thing for Amazon and a good thing for product oriented media companies for those reviewers and reviews to regain some of the clout and influence that they have lost to easily manipulated user reviews.
On the flip side I think media companies have to do more to deserve the trust of shoppers. More Consumer Reports style product evaluations and less clickbait trolling, more investment into respectable writers/video reviewers who have integrity and earn a real salary to make them resistant to PR bribery.
I don't think most shoppers fully understand how extreme the review manipulation is on the brand and merchant side of things. One issue is that the Federal government does not enforce the laws on the books against many of the worst types of manipulation. I'm not a particularly pro-regulation type of guy, but this is one of the consequences of zero law enforcement in a particular area of business: you become a sucker if you abide by the law when none of your competitors are.
The Yelps and Amazons of the world are kind of split on this because they both want to restore/retain the trust that users have in these reviews without drawing attention to how absurd the level of manipulation is.
Another issue is the entirely legitimate manipulation of reviews that nonetheless confers a competitive advantage. One local car dealership that did some service on my car has earned thousands of positive five star reviews by being extremely aggressive with e-mail followup to people who show up for even minor service interactions. I gave them five stars out of bemusement and because we legitimately received good service on a free interaction paid for by the car manufacturer with a rebate. But I would have been much less generous with my five star review had I had to pay for the service, as everyone is. That's why Amazon as of last year bans that practice.
I have myself counseled clients in ecommerce and written email sequence scripts to do just this kind of manipulation. But in the case of this dealership and of most people, is the dealership that is the most efficient at running an email marketing program really the best car dealership in the area by an order of magnitude? While it may be an indicator of the overall professionalism of the company and a signal of business health, it's pretty unrelated to the business of selling cars and servicing them. As a rather clueless car owner, I'm not even really qualified to give an intelligent opinion about the quality of a dealership apart from how I feel about it and whether or not I (rather clueless) think that I was ripped off or not.
Why is my unqualified opinion based on that minor interaction worth more than that of someone who really knows a lot about cars, bought a car from that dealership, and has had dozens of service appointments with the mechanic? Why is my dumb 5 star rating worth the same amount as that guy's? It makes no sense.
I guess they'll never go for this but they could do something like select 1% of buyers at random and pay them (or refund product cost) to write a meaningful review with pictures with updates at fixed intervals so the product is evaluated throughout its intended lifespan. Perhaps with some pre-filtering to improve the likelihood that reviews will be useful and accurate (eg "spent more than $10k on amazon", "billing address is shipping address", "previous reviews were voted as useful" etc)
Amazon may not be able to stop it, but they could surely mitigate it significantly. They choose not to, because in the end, having lots of reviews for lots of products is better for them than having a few reviews for a few products, even if a significant number of the reviews are fake.
Sure, but it would surely reduce the scale of these fake reviews. Right now, companies can buy reviews in batches of thousands; the only actual cost is in somebody/something creating that review and some overhead.
If Amazon also required actual purchase of that item, now there's the extra cost of actually buying that item. Even more so when you're talking about things that are expensive. This would drastically lower the amount of spam that could be generated unless the company wants to throw some huge money at it.
Returned items would still potentially be an issue, but I'm guessing Amazon already has things in place to kick off users that return too many items.
There will always be a desire to pay for reviews, since 70%-80% of shoppers check reviews before buying[0][1] (and growing[2]).
Amazon could help fix this problem by creating "sponsored" reviews. This way a company can pay for reviews in a non-scummy way, a reviewer (and Amazon) gets paid, and a customer can tell whether the review was paid for or not.
I like the idea, but if this is crowdsourced how can we trust that the reviews will be unbiased? Amazon reviews are supposedly crowdsourced, and we know what a shit show those are.
reply