Amazon launches new feature, the number of reviews is no longer important

Amazon is testing a new feature that allows buyers to quickly get answers about products without browsing the review page. The feature appears to be an optimization of AI-summarized reviews, weakening the importance of the number of reviews.

As a seller, you cannot control how Amazon responds to buyers' questions or displays your products using this new feature. This may reshape the sales landscape and affect the sales of some products.

Amazon launches new feature, buyers don't have to read reviews

An Amazon spokesperson has confirmed that the company is rolling out a new feature that will use artificial intelligence tools to answer consumer questions about a product.

The new feature is available in the mobile app. When users ask questions about a specific item, Amazon will give an answer within a few seconds. This answer is mainly obtained by summarizing product information and reviews based on Amazon's big data.

The feature is currently in testing and is only available to some consumers, not everyone.

It is worth noting that this feature allows shoppers to find product information without browsing through review pages or browsing product lists. The answers given by artificial intelligence directly summarize the review content and product selling points, which seems to make the number of product reviews no longer important. As the new feature matures, it may gradually weaken the importance of the number of reviews on the listing.

For example, under the new feature, when a consumer sees a sweatshirt, they can ask questions like , "Is this sweatshirt suitable for running?" or "Is this shirt made of thick material?" or even "Will it fit someone who is 6 feet 7 inches tall?" Amazon will respond to these specific questions.

The new feature aims to replace reading individual reviews by summarizing specific information that shoppers care about, which is similar to Amazon's AI-powered review summary, but different.

Like other AI assistants, Amazon AI Assistant will happily answer questions that Amazon wasn’t designed to answer. On a listing for a women’s vest, it can write a description of the product and describe it in the style of Yoda from Star Wars.

An Amazon spokesperson said : “We are constantly inventing products to help customers live a better, simpler life. We are currently testing new features to improve shopping on Amazon by helping customers get answers to common product questions.”

The new feature still has significant limitations

This new feature of Amazon seems smart, but it still has great limitations, mainly reflected in the following four points:

First, they limit themselves to using only one product and refuse to compare other similar products or look for substitutes;

Second, no other actions, such as adding products to the shopping cart, are performed;

Third, it cannot provide useful information such as historical prices;

Fourth, they tend to provide false information about products or refuse to answer basic questions, such as stating the price of a product.

Unlike OpenAI's ChatGPT, Amazon's new feature can't hold a conversation, it can only respond to creative questions. And the tool doesn't stray off topic, it's limited to answering questions about Amazon products . If you ask a question like "Who is Jeff Bezos?" it will return an error message.

Importantly, sellers cannot control how Amazon uses AI to answer buyers’ questions. As a seller, you definitely want AI to display good product information and reviews, because wrong information or bad reviews can directly affect sales or increase returns.

In the case of a bike on Amazon, an AI assistant answering the question “Is this bike good for a 5-mile commute up a hill?” pulls useful information from the product reviews and description. It understands the distance of 5 miles, what commuting means, and how an e-bike can go up a hill.

And when consumers ask , "Is this bike suitable for a romantic date?" it says, "While an e-bike can be fun for a date, it may not provide an intimate or romantic setting. Consider a more traditional cruiser bike or tandem bike for a more romantic riding experience."

The next step for Amazon is to have an AI ranking of the best bikes based on how well they match the question if you type “good bike for a 5 mile mountain commute” into an Amazon search .

In this process, outputting review summaries is the first step, and the shopping assistant is the second step, whose output is dynamic output based on the query. Next, artificial intelligence will help consumers purchase directly or save items to their wish list.

Is Amazon’s continued investment in artificial intelligence effective?

Over the years, Amazon has seemed to be trying to innovate using machine learning and artificial intelligence to help customers have a better shopping experience. The company is also very persistent in artificial intelligence and continues to invest in innovation.

ChatGPT became very popular in early 2023, and the concept of artificial intelligence has been continuously magnified. Both world giants and cross-border e-commerce sellers have connected to ChatGPT, trying to use artificial intelligence to improve output value and efficiency.

In the cross-border e-commerce industry, some big sellers have achieved gratifying results by using ChatGPT. For example, after Jihong Co., Ltd. connected to ChatGPT, its digital operation capabilities were effectively improved, which boosted its revenue and net profit. In addition to Jihong Co., Ltd., Anker Innovations, Lechuang Co., Ltd., Huakai Yibai, Giant Star Technology, Hanwang Technology and other cross-border big sellers have also entered the market to test the waters.

As an e-commerce giant, Amazon also wants to get on the express train of artificial intelligence.

In June 2023, Amazon began showing AI-generated product review summaries to US customers. When users purchase products on their mobile devices , AI will provide a brief review summary, highlighting product features and customer sentiment frequently mentioned in the reviews to help buyers determine at a glance whether the product is suitable for them.

At that time, there was a view in the industry that the number of Amazon reviews was no longer important, because with AI summarization, a large number of reviews were just a number to buyers. However, some people believed that if there were not enough reviews, AI would not be able to summarize them, and reviews represented weight, so Amazon would not kick out old sellers. Today, Amazon’s new feature of answering product questions through artificial intelligence is equivalent to further optimizing the AI ​​review summary function.

Since then, Amazon has added a new AI listing function to allow sellers to list new products faster. Sellers only need to use a few words or sentences, and Amazon will generate high-quality content. Sellers can optimize the content or use it directly. However, as time goes by, some sellers find that the AI ​​is too outrageous, and listings in multiple categories have been removed from the shelves.

Amazon's artificial intelligence has been moving forward. Whether it is answering customer questions this time or the previous review summary and listing functions, there seem to be quite a few flaws.

About a year ago , ChatGPT was very popular, but now it has cooled down rapidly, and the number of times it is mentioned has dropped sharply. Can Amazon's persistent artificial intelligence really achieve the expected effect? ​​Perhaps it remains to be verified by time.

Amazon

Comment

AI