Product Reviews: Part 3

Broke the Amazon New Reviewer Rank of #2000 for the first time today, jumping from #2017 to #1961 in less than 24 hours, although this ranking fluctuates and is bound to dip again before going back up. Almost exactly two years ago, I posted the following comment in my first write-up about writing reviews for Amazon:

In my opinion, my "helpful vote" percentage, which has apparently decreased from 90% to 88%, is a direct result of my decision to review for Amazon Vine, although the Amazon reformulation of its "Classic Reviewer Rank" algorithm into its "New Reviewer Rank" to determine overall reviewer rank at the time boosted me from around 8500 to around 3000. Apparently, the original algorithm was highly based on total "helpful votes", whereas the new algorithm places importance on how recently reviews have been written as well as review quality, although it is uncertain how an algorithm can determine review quality. It is already apparent that review quality cannot be based on "helpful votes".

Amazon ranking algorithms still remain a mystery. About a year-and-a-half ago, I posted that I had received 500 helpful votes to date at the time for the various reviews I had written, but as of today I only have 888 helpful votes, while my ranking has jumped by over 1000.

While I enjoy writing reviews so that I can share with other readers, especially technology professionals, I still find it interesting to following the rankings because they often do not make sense, although by now it is apparent that my movement in the rankings depend as much on the helpful votes and recency of reviews that I receive as much as the helpful votes and recency of reviews of those who rank above me.

In the last couple years, I have made a couple of other observations. The first is that helpful votes being equal, reviews written for Amazon Vine always display on top. The second is that, unfortunately, unhelpful votes are often granted for reasons other than whether the associated reviews are helpful. The third is that it is fairly clear that some reviews are written by individuals who have not even read the book.

Amazon has attempted to rectify this third item by introducing the Amazon Verified Purchase designator that signals to readers that the book was purchased from Amazon, and although this at least shows that the book was actually purchased by the individual, some individuals seem to think that if this designator does not exist, the reviewer did not purchase the book at all, which is obviously an inaccurate conclusion.

Resolution of the second item, however, is probably not going to be possible. While granting helpful or unhelpful votes to reviews are intended to communicate whether they provide information that will influence the purchase decision, it is clear that unhelpful votes are often cast due to disagreement with what the reader presumes the book to be about.