How do I analyse my review comments?

Every time your customers give a service review, they give a star rating. In addition to the star rating, they can leave a review comment that contains more information about the shopping experience.

You can use this information to…

… analyse the general sentiments of your customers.

… uncover weak points in your processes.

… discover opportunities for improvement.

However, a more in-depth analysis of review comments is very time-consuming. The Sentiment Analysis therefore helps you to use this valuable source of information with very little effort.

What is the Sentiment Analysis?

The Sentiment Analysis is a tool that examines your review comments using artificial intelligence. The Sentiment analysis is a two-step process:

Step 1: Review comment investigation

The Sentiment Analysis examines all the review comments you've collected. Two questions will be examined:

1. Which parts of the shopping experience are mentioned in the comment?
2. Are the parts mentioned rated positively, neutrally or negatively?

It is not always clear whether an area of the shopping experience is rated positively, neutrally, or negatively. From time to time, positive and negative experiences are mixed within a review comment.

In such cases, the Sentiment Analysis calculates an average value to decide whether a mention should be classified as positive, neutral, or negative.

Step 2: Calculation of the Sentiment Score

The Sentiment Analysis records how many review comments mention certain parts of the shopping experience as positive, neutral or negative.

The Sentiment Score is calculated based on the distribution of positive, neutral and negative mentions. The Sentiment Score is based on the following formula:

(Number of positive mentions - number of negative mentions) : Number of all mentions x 100 = Sentiment Score

This Sentiment Score ranges from -100 ("only negative mentions") to +100 ("only positive mentions").

The Sentiment Score allows you to see at a glance how satisfied your customers are with a particular part of the shopping experience. These insights will help you decide in which areas you can improve.

How do I use the Sentiment Analysis?

Follow these steps to use the Sentiment Analysis:

1. Booking the Sentiment Analysis
2. Using the Sentiment Analysis

Booking the Sentiment Analysis

The Sentiment Analysis is a fee-based add-on to your Trusted Shops membership. Follow these steps to book the Sentiment Analysis:

1. Log in to the eTrusted Control Centre.
2. Click on "Analytics" () in the main navigation. Click on "Sentiment Analysis" (). Select the desired channel using the "Selected channel" selection (). Then click on "Uncover deeper insights" ().
   
3. A contact form opens. Enter and submit your contact details.

You will then be contacted by Account Management and receive an individual offer for the booking. After booking, the Sentiment Analysis will be activated in your eTrusted Control Centre.

Using the Sentiment Analysis

1. Log in to the eTrusted Control Centre.
2. Click on "Analytics" () in the main navigation.
3. Click on "Sentiment Analysis" (). The "Sentiment Analysis" overview appears.
4. Select the desired channel using the "Selected Channel" selection ().
   
5. In the "Sentiment Analysis" overview, you will find a lot of information about your review comments:
   • The "Total reviews" display informs you about the number of reviews you have collected.
     
   This number also includes reviews that do not include a review comment.
   • Column "Categories" informs you about the parts of the shopping experience that the Sentiment Analysis searches for in the review comments.
     
   Currently, the Sentiment Analysis is looking for mentions of the following categories:

   • Returns/Refunds/Cancelation
   • Website & Shop
   • Product
   • Customer Service & Communication
   • Payment
   • Shipping & Delivery
   • Column "Mentions" informs you of how many review comments were used to calculate the individual category scores.
     
   In order for a review comment to be counted as a "Mention" of a category, the category must not only be mentioned but also evaluated.
   For example:

   • The review comment "I have ordered a shoe" is not counted as a "Mention" of the "Product" category, as no rating of the shoe was made.
   • The review comment "The shoe is great" is counted as a "Mention" in the "Product" category, as the shoe has been rated positively.
   The first number in the "Mention" column indicates the number of review comments in which the respective category was evaluated. The second number provides information on the number of reviews collected.
   For example:
   The display "651/2481" in the line "Returns/Refunds/Cancelation" informs that this category was evaluated 651 times within the 2481 collected reviews.
   • Column "Sentiment breakdown" provides information on how negative, neutral, and positive mentions are distributed within the individual categories.
     
   The negative mentions are shown in red, the neutral mentions in yellow, and the positive mentions in green.
   • Column "Sentiment Score" () informs you about the Sentiment Score of the individual categories. To calculate the Sentiment Score, the positive mentions are compared to the neutral and negative mentions.
     
   The sentiment score ranges from -100 ("only negative mentions") to +100 ("only positive mentions"). To make it easier for you to assess the individual results, they are coloured within the column. The legend () at the bottom of the overview explains the colour scheme.
6. Would you like to analyse your review comments for another period of time? Then, click on "Date Range" (). A selection field opens. Either select one of the preset periods () or define an individual period using the calendar (). Then click on "Apply" ().
   
7. Would you like to analyse the development of a category in more detail? Then, click on the arrow belonging to the desired category in column "Details". A detailed overview will open.
   
8. In the detailed overview, you will find further information on the selected category:
   • The "Sentiment Breakdown" () overview tells you how the Sentiment Score and the distribution of negative, neutral, and positive mentions have developed in the chosen period.
   • The "Mentions Breakdown" () overview tells you how the number of "Mentions" has developed during the chosen period.
   • The "Reviews in this category" () list shows you all reviews in which the selected category was reviewed. The mentions of the category are coloured green, yellow, or red within the review comment () – depending on whether the category was rated positively, neutrally, or negatively.
     
9. Do you want to filter the reviews within the "Reviews in this category" list? Then click on "Filter" (). In the drop-down menu that opens, you have various options:
   
   • You can display only positive, neutral, or negative mentions ().
   • You can display only reviews with a certain number of stars ().
   • You can decide whether the list should start with the latest or the oldest review ().
   Click "Apply" () to apply the selected filters.
   

10. Do you want to open a review in the inbox to reply to it, for example? Then click on "View in inbox".
    

11. Do you want to open the detailed overview of another category? Then use the "Categories" () drop-down menu to select a different category.
12. Do you want to close the detailed overview again? Then, click on "Close details" ().