This feature is currently in limited availability.

Understanding VoC (Voice of Customer): Sentiment Analysis

What is VoC (Voice of Customer)?VoC represents consumer feedback regarding a product, service, or overall shopping experience. Negative VoC is an important signal for assessing the health of user experience, as it may indicate potential issues such as product quality problems, poor logistics experience, or discrepancies in product descriptions.Understanding and proactively monitoring VoC not only helps improve customer satisfaction and repurchase intention. The module aggregates data from product reviews and return/refund requests, using NLP techniques to detect negative sentiment signals in these channels, providing a diagnostic view of customer sentiment for each product.
Where to find VoC Diagnosis:The VoC Diagnosis module is located under Analytics > Product Analytics.With the VoC module, sellers can:

• View the VoC status of each product
• Understand the issues causing negative feedback and their frequency
• Identify the core problems affecting consumer experience

By leveraging these insights, sellers can detect potential risks earlier and take effective actions before sales are impacted.

Why is VoC Important for Sellers?

Customer feedback, especially negative feedback, is the most direct signal for assessing the health of the shopping experience. A poor experience not only reduces a user’s willingness to purchase but also weakens their trust in the seller, ultimately affecting overall sales performance.
For both the platform and sellers, timely identification and handling of negative VoC is crucial to maintaining healthy and sustainable business growth.
In the past, however, sellers often only noticed issues after VoC problems had already escalated. By the time they received platform alerts, VoC metrics were usually already high. Therefore, sellers need earlier and more proactive insights to detect potential risks at an early stage and take corrective actions before GMV is impacted.

How is the VoC level calculated?

In the VoC column, sellers will see four status states. We compare the negative VoC of each product against the category benchmark and classify the product's VoC into four levels:

• At Risk – Significantly above the benchmark
• Needs Attention – Slightly above the benchmark
• Market Average – In line with the benchmark
• Below Average – Below the benchmark

We recommend that sellers prioritize products marked as At Risk and Needs Attention, as these statuses indicate overall negative VoC from users. Timely attention and adjustments for these products can help address potential issues and improve performance.

How to check your VoC performance?

Seller Center >> Analytics >> Product analytics >> Product list

Diagnosis Summary

• In the Product list, we added a Diagnosis summary on top of the page, which reveals the VoC performance of all products.
• We made a special call out for those high GMV products (GMV ≥ 10,000 or items sold ≥ 300 in the past 30 days)

Filter products by status

• You can click on any VOC status category of products in the Diagnosis Summary to view all products with that status
• You can also directly select the status you want to view from the Status dropdown menu
• You can also select multiple statuses at the same time to view products with those statuses

Filter products by GMV

• You can filter the product list you want to focus on based on their GMV performance and the number of items sold
• It is recommended to prioritize products with high GMV and high sales volume

Check VoC on product list

This content is only supported in a Feishu Docs

• Clicking on a product's VoC diagnosis status in the product list will open a drawer window on the right
• We only display the top 5 most frequently mentioned issues along with their respective proportions
• We highlight in red any VoC with a proportion greater than 50%. If all VoCs have proportions below 50%, we highlight the top-ranked VoC

Check the reason and take actions

This content is only supported in a Feishu Docs

• Click on a product's VoC status to open the right-hand drawer, where you can view the current status and compare it with the previous period.
• The system uses AI to aggregate negative VoC feedback for the product and displays issues in descending order of frequency (Note: a single customer feedback may involve multiple issues, so total percentages may exceed 100%).
• Below the AI summary, we display a set of important keywords extracted from customer feedback. We recommend that sellers review these keywords as a supplement, ensuring they consider the full set of issues highlighted by both the AI summary and the keywords.
• Sellers can click the buttons below to quickly access the Refund Management or Product Rating pages to monitor and address the corresponding issues.

Previous Success Cases

Sellers who have leveraged user feedback from VoC to make targeted optimizations were able to turn their products' VoC from negative to positive, resulting in improved overall sales and performance eventually.
We recommend that all sellers, like those in the examples below, regularly monitor their products' VoC performance and make corresponding adjustments based on user feedback. This proactive approach can help enhance overall customer sentiment toward their products.
Fashion & Sports3C & ElectricalBeauty & Personal care

FAQ

What does "At Risk" and "Needs Attention" stand for?These tags compare your product's VoC with the overall VoC level in your industry. Our goal is to provide sellers with a benchmark across the industry, offering a target they can aim to reach.
Why can't I directly see the VoC-related issues after entering the Manage Refund / Product Rating pages?At the moment, individual VoC cases are not surfaced on these pages. The goal of this feature is to help sellers stay focused on the key issues highlighted through aggregated user feedback, so they can better understand what consumers are saying and take clear actions to improve performance.
Why is a hyphen "-" displayed under the VoC diagnosis for some of my products?There are two possible reasons for this:
First, all VoC data has been collected starting from October 16th. If the selected date is earlier than October 16th, the VoC status will be displayed as a hyphen "-".
Second, it will also show as "-" when the data volume related to that product is insufficient to be statistically reliable.