Analyzing Customer reviews
Stephane_SimpleDecisions
Last Update 3 jaar geleden
Monitoring customer satisfaction is must. But analyzing customer reviews can be really painstaking.
Analyzing customer reviews with Simple X is really easy and powerful. The above video will give you an overview of what can be done.
We assume here you know how to import your raw data from a spreadsheet into Simple X. If not please refer to our specific import tutorial here in our Knowledge Help Center.
With or without ratings
Most customer reviews come with a rating. Between 1 to 5 (eg Amazon or Google) or 0 to 10 (NPS). And this rating is complemented by a text review to be further analyzed.
If so, it is useful during import to select this rating field as a primary filter. It will be be displayed on the left sidebar, above the topics and will be generate specific charts.
Searching, filtering, counting
You can now browse very easily within the thousands of surveys by selecting the ratings, the keywords, plus the filters available in the right sidebar.
Sentiment analysis
You can add a sentiment analysis to every review by activating the sentiment analysis in the Advanced tags section.
Often ratings (if any) and sentiment are partially redundant information. Often the satisfied reviewer (eg 4 or 5 out of 5) will post a comment that will be detected as positive and conversely dissatisfied reviewers (1 or 2) will be detected as negative.
But not always:
- A very satisfied reviewer can use a single word, whose sense depends in fact of the rating ("Location!"). This comment is likely to be detected as neutral.
- A satisfied reviewer can mention something it did not like. Enough to be detected, but not to spoil his/her satisfaction
- Sometimes, sentiment analysis just get it wrong! Good news is that you can correct this by clicking on the corresponding smiley and adjust if required.
Building topics
Unlike many alternative tools, SimpleX allows you to build the exact coding plan that suits your particular business. You are in full control, with the help of AI.
You build the topics of your choice by using the "create new topic" button.
You can specify what you would like to put into each topic by providing a few samples.
At creation you can ask SimpleX to suggest reviews that fit into it because they mean something similar to your samples.
Once created, you can also click next to the topic and select suggest content, and it will suggest other reviews based on the reviews you have already put into it.
Once all you topics created, you can also use the "distribute the selection" to distribute across the different topics , again based on meaning.
Using pre-trained industry models
You can also get access to one of our pre-trained models, by using the Categorize function, in the Build topics button.
For customer reviews, 3 models are relevant:
- E-commerce
- Retail
- App
Sharing insights
Once categorized, you can browse, find, count, compare by topic.
In the chart module, you can visualize the relative weights of the different topics in the tree map chart.
Select Charts in the general left sidebar and click on the tree map. This chart will adjust whenever filters are activated in the right sidebar.
Note that our industry models are featured only in our Enterprise Plans. And of course, we can build for you a completely custom model to serve your specific needs, directly accessible from SimpleX.