It’s a bad day for businesses who buy fake reviews to juice their product ratings on Amazon. Already inundated with customer reviews, Amazon is turning to the power of artificial intelligence, or machine learning, to make sense of all the reviews.
The new system will focus on a combination of freshness, written by people who actually bought the product and how frequently other buyers cite the review as ‘helpful.’
These reviews will also be weighted, changing how the five-star system operates under the hood. The ‘helpful’ reviews will add more to the overall rating while old or ‘unhelpful’ reviews won’t impact the product’s rating.
Take a book with ten reviews. Under the old system, the star rating would be an average of those ten. With the new review system, the weighted reviews come into effect, either positively or negatively. It’s good news for consumers and businesses.
The consumer can rely more heavily on the reviews, and businesses don’t have to worry as much about their bottom line being impacted by spam reviews. It also allows for new product iterations to recover from initial bad reviews.
Say a business releases a blue widget. It turns out to be less than advertised, and consumer feedback speaks to it being a terrible product. The business goes back to the drawing board and releases an improved version. The reviews on the new product gain more prominence over the reviews for the old product.
Amazon Review System
Of course saying something cannot be gamed on the Internet has not met the Internet marketing community. Amazon is effectively saying it’s building the Titanic for reviews. Guess what? The company can expect a lot of icebergs.
How will the fake services change their offering to keep enticing business? Here’s how it could work as the new system rolls out.
The two obvious methods are to spam the ‘helpful’ section and keep reviews fresh. That’s a short-term game, and will thankfully be squashed. Next is for those who like the long game. Have an aged and non-bs Amazon account? Services will use it as a Tier 1 account. Purchase the product, and use the spam accounts to thank the reviewer.
Should companies engage in this? Hell no. Inevitably, both methods will fail. Unlike spam you eat, Internet spam always has a shelf life. And when it goes bad, it is apocalyptically bad.
Build great products, write amazing books and build the right way. Patience as a business can be agonizing, but it beats gaming the system for months to wake up one day to see your business shattered.
With the gradual rollout, consumers won’t notice much of a difference unless they are making a journal record of the average star ratings of the products they buy. Hey, maybe that’s their hobby. Not judging.
One area of Amazon that will use the new system immediately is the Amazon Choice program. Say you are searching for a general product versus a brand. Amazon would automatically suggest a brand for you.
The new reviews will be incorporated to offer the best search results. It’s a great way for Amazon to avoid any perception it is pushing certain brands towards consumers.
Overall, the system is win-win. Consumers can be more trusting of the reviews and businesses that have quality products get rewarded.