Consumers ratings -- the little stars in the listing -- play a major role in what many people choose to buy from Amazon.com (NASDAQ:AMZN).
The score on the five-star scale tells potential buyers a lot about the experience existing owners have had with a product. How those reviews are calculated has, up until recently, always been simple math, an unweighted average. Each review is counted equally and when it was written did not matter.
In theory, that led to some problems because a well-ranked product might have been supplanted quickly by new technology or improved models. The rating would remain four or five stars only because new people weren't writing reviews. The reverse could happen if a company made a change to a product but kept selling it under the same name. New ratings might be higher than old ones, but that would only reflect in how they change the overall weighted average.
This left an important part of Amazon's system somewhat outdated and not as useful as it should be. The online retailer is about to change that and modernize how it handles reviews and product ratings.
What is Amazon doing?
Using a new machine-learning algorithm, the online retail giant will give greater weight to "newer, more helpful and verified customer reviews (written comments) and ratings (the 5-star system) when determining top reviews to display, and when calculating a product's overall rating," GeekWire reported.
The changes, which have already begun to be implemented, are designed to make the ratings more reliable and current. Changing the formula could also make them very volatile. For example, if a well-liked product develops a problem, it could fall in the ratings.
"Amazon is enhancing the customer reviews system, adding a few changes we hope will help make product feedback even more useful to customers," an Amazon spokesperson told GeekWire. "The enhanced system will use a machine-learned model to give more weight to newer, more helpful reviews from Amazon customers. The system will continue to learn which reviews are most helpful to customers and improve the experience over time."
Why does this matter?
In many cases when I'm buying something on Amazon.com (which I do nearly every day) the ratings and, to a lesser extent, the written reviews heavily influence my opinion. For example, I recently purchased a charging case for my iPhone. After ruling out a number of the better-known, more expensive models, I chose one of the cheaper off-brand models based on it having a 4 1/2 star rating.
Had another charging case at a similar price been better rated then I would likely have bought that one. That type of decision happens millions of times a week on Amazon and a new ratings system may greatly impact what gets sold by the retailer.
This is good news for consumers
Improved, more accurate ratings offer Amazon customers an improved tool in making informed buying choices. That should lead to higher satisfaction with those purchases (I have mixed feelings about my charging case) and perhaps drive even more sales to the online retailer.
An improved ranking system that factors in who's writing the review and when they wrote it should weed out some obvious fakes which can skew the average. The algorithm will also get better the more it's used.
"The system will learn what reviews are most helpful to customers ... and it improves over time," Amazon spokeswoman Julie Law told CNET. "It's all meant to make customer reviews more useful."
These are positive changes for Amazon and its customers that should make the site even more useful. The old system was surprisingly non-technical for such an innovative company, and though it may take time, the new one should be vastly superior.