Amazon launches two new features, sellers are expected to pay less

Amazon has launched two new features, which are intended to allow sellers to improve products based on review trends, reduce returns, and improve customer experience. If the new features are practical, it is expected to reduce sellers' return processing fees. The platform's extreme customer experience has always hit sellers hard. Because of malicious returns, some sellers have to give up the categories they operate.

Use reviews to improve your product

Amazon recently announced that sellers can access new, data-driven product insights in the Customer Review insights tab of Opportunity Explorer to launch successful products.

This update includes the following aspects :

Review Trends (for brands) shows changes in customer sentiment over the past six months to identify trends in positive and negative reviews.

Parent ASIN insights provide an overview of customer sentiment and are especially useful for ASINs with multiple variations as they summarize customer review insights for the product. With Parent ASIN insights, brands can now gain an in-depth sentiment view of their entire listing in one place.

ASIN Benchmark compares your product’s star rating to the category average to strategically position your products and gain a competitive advantage.

TopicsDig deep into subtopics frequently mentioned in reviews to gain an accurate understanding of product attributes and customer experiences. Prioritize improvements that address core customer concerns to deliver better products.

Amazon's goal is to help sellers understand the specific problems of their products through review trends, thereby gaining a competitive advantage and further improving and launching better products. The platform's original intention is good and it hopes to help sellers.

However, sellers are more concerned about negative and malicious comments . One seller used 3 "ifs" to question whether the function has solved the seller's concerns.

It might help if Amazon only allowed real buyers to leave reviews ;

This might be useful if Amazon removed obviously abusive reviews ;

Maybe it would be useful if Amazon didn’t allow reviewers and fake accounts to leave bad reviews for a ton of products they never purchased.

Some sellers are looking for ways to delete malicious reviews. A colleague said that Amazon doesn’t care about fakes or abuse. They care more about customer experience. Another recent AI feature was launched, which uses artificial intelligence to find damaged products and reduce returns. This is another move by Amazon to increase customer experience. Is the platform going to take a strong approach to solve sellers’ pain points?

Using AI to detect damaged products before shipment

Amazon has launched an artificial intelligence (AI) model called "Project Pi" (Private Detective) that can scan items for defects and ensure that customers are satisfied with every order they receive. Currently, millions of products pass through the imaging tunnel every day in Amazon distribution centers across North America.

By combining generative AI and computer vision technology, before the product is shipped to the customer, it will pass through an imaging channel, where Project PI uses computer vision to scan the product and evaluate the image to detect defects. If a defect is found, Amazon will not ship it to the customer and further investigate whether there are more problems with similar products.

Amazon employees are responsible for reviewing Project PI-tagged items and then deciding whether the item is eligible to be resold on Amazon's website at a discount or find another use for it.

The model provides Amazon employees with an "extra pair of eyes" and has helped to strengthen manual inspections at multiple distribution centers in North America. During the system's operation, it has proven to be able to skillfully classify the millions of items that pass through the tunnel each month and accurately identify product issues. This AI technology is expected to expand to other locations in 2024.

Additionally, Project PI can help identify the root cause of an issue so preventive measures can be taken upstream to prevent the issue from recurring.

Project Pi is part of Amazon's customer-first culture, and preventing damaged or defective goods from reaching customers is critical to a good customer experience. Accidentally shipping defective goods can lead to unnecessary returns, resulting in losses for sellers.

Meanwhile, the Amazon team is leveraging a generative AI system that uses multimodal LLM ( MLLM ) to investigate the root causes of negative customer experiences. Amazon uses it to continuously improve the system when it learns from customers about defects that it did not identify. The system first reviews customer feedback and then analyzes Project Pi images obtained from fulfillment centers and other data sources to confirm what caused the problem.

Technology can help sellers more easily access data, and if a seller accidentally puts the wrong size sticker on a product, Amazon can communicate the issue to prevent the mistake from happening again.

More than 60% of sales on Amazon come from independent sellers, most of whom are small and medium-sized businesses, which can reduce the total number of returns to a certain extent by reducing the number of defective products to customers.

Since the beginning of this year, many sellers have been suffering from the pain of returns, especially the return processing fee that came into effect on June 1, which increased the costs of many sellers. The new AI function launched by Amazon based on the pain points of sellers may help sellers reduce some of the return costs.

Fraudulent returns force sellers to abandon categories they sell

In April this year , the Wall Street Journal published a report titled "Amazon Sellers Face Surge in Fraudulent Returns", which attracted widespread attention and seemed to put Amazon under some pressure.

Nicole Barton, a seller in the report , tried to seek help from Amazon after experiencing multiple fraudulent returns, but the complaint process was not satisfactory. In the end, she had to give up selling clothing and apparel and switch to selling consumables such as pet supplies.

This problem has caused losses to many sellers, such as sellers of household goods , who received returns of TV set-top boxes and old soaps ; sellers of outdoor coffee products such as travel French presses , who received returns of Christmas ornaments and toy airplanes, etc. They received completely unrelated returns.

There are also many sellers who receive returns that do not match the texture and color of the products they sell. Some of the returns appear to have been used and are obviously returned after being used up.

This fraudulent return behavior can take many forms, such as returning used items, stolen items, or counterfeit products.

Sellers are counting on Amazon to recover money lost to return fraud. But they say that because processing can take weeks, even if Amazon accepts a claim, sellers may not get a full refund for the goods. In many cases, sellers say they are left to absorb the losses themselves.

The National Retail Federation said that in 2023, about 13.7% of returns were fraudulent, costing retailers $101 billion. Sellers said that return fraud is becoming widespread and Amazon has not taken effective measures to prevent it.

There have also been horrific cases where Amazon accidentally repackaged a defective product upon return and shipped it as a new product to another customer, causing huge hassles and losses to the seller.

Amazon will also introduce some features such as those mentioned above to prevent such problems from happening, but many sellers said that these new changes mainly benefit buyers, and some sellers hope that Amazon will give them some control to verify returns before processing refunds.

The two new features launched by Amazon this time are intended to help sellers improve products and reduce returns, but their actual effectiveness remains to be verified.

Amazon

New Features