The latest algorithms of reviewing stars and praising are revealed
Some time ago, the crisis of trust in the review has severely damaged Amazon's stock, and many media reported the false problems of Amazon review, which is also the root cause of Amazon's took actions, and this piece of review is also the focus of Amazon's attack in the future. It is also an indisputable fact that the risk of brushing one's comments is also increasing day by day.
From the promotion point of view, the absolute quantity of a review is one aspect, but the most important thing is that it is an overall star of the review. As you can recall carefully, your star dropped from five to four, which has great influence on your sales. However, if the absolute quantity of the listing evaluation is not enough, the influence is not as big as the decline of the star.
Today, let's have a good chat. If the review star of listing falls, apart from adding five-star comments to the list, is there any other methods to control the review star?
First of all, let's take a look at an accounting standard and algorithm of the review star. The review star is not accounted for according to simple arithmetic average. Let us take a look at the following very typical example:
It can be known that the overall proportion of five-star and four-star reviews is less than 70 %, but the overall listing of review stars is 4.3. It is a clear explanation of this problem. How many stars does a listing review have, not according to how many five-star reviews and how many four-star reviews have, and the arithmetic average can be obtained. If so, then this listing will never have an overall evaluation of 4.3 again.
Therefore, a listing review star must be calculated by weighted average. Besides considering only how many five-star reviews and how many four-star reviews there are, it is also necessary to consider more important weighting factors.
Let me explain to you some weighting factors about review rating considerations.
That is to say, the longer the review remains, the higher the weight value has, and the more new review will have less influence on the overall weight value.
At that time, we could call it Amazon's golden years. A scripting system that we developed at that time cost almost nothing, but it was profitable. But the good times will always be over, and the platform needs to go through a process of gradual improvement.
After analysis and investigation, we found that Amazon's ascending and descending system is similar to Amazon's review star mode now, and we have taken weighted average consideration, instead of taking a quantitative action of a simple point, which is as the only criterion for consideration.
I especially feel that Amazon has been ravaged by our Chinese sellers, so the following operational space will become narrower and narrower.