Semantic analysis and verbatims overview

Verbatims are essential for understanding the needs and expectations of your customers and optimizing their experience with your brand.

The Skeepers Feedback Management platform offers you three levels of semantic analysis, with increasing complexity, allowing you to select the approach that best suits your needs to obtain insights and guide your decisions regarding the improvement of the customer journey.

Depending on the chosen option, you can select a quick analysis method for simple insights or more in-depth solutions with detailed information and classifications. This flexibility allows you to choose the analysis that best matches your strategic objectives and the complexity of your data.

Glossary of verbatim and semantic Analysis

Verbatim

A verbatim refers to a raw textual comment or feedback collected directly from survey participants. Verbatims are open-ended answers where respondents can express their opinions, feelings, and experiences in their own words. They can be displayed in word clouds, classification plans, or raw data visualizations.

Verbatim analysis

Verbatim analysis is a feature used to examine, understand, and extract insights from textual answers collected in surveys. This method helps to better interpret the opinions, feelings, and suggestions expressed directly by respondents.

Semantic analysis

Semantic analysis is a feature used to understand and interpret the meaning of words and texts in a given context. More advanced than verbatim analysis, it aims to identify meanings and relationships between terms to extract richer and more detailed insights from the content.

Verbatim desensitization

Verbatim desensitization is a feature accessible to everyone that allows masking personal data provided by a respondent in the text field of a survey.

Word cloud

A word cloud is a visual representation of the most frequently used words in answers to an open-ended question. Words are displayed in varying sizes, with the size of each word proportional to its frequency of appearance in the analyzed text : the more frequent a word, the larger it is represented in the cloud.

Classification plan

A classification plan is an organized structure that allows grouping, classifying, and analyzing verbatims according to predefined criteria, such as specific themes or concepts. It serves to structure data to identify trends and statistical insights from the collected information.

Choosing an appropriate level of semantic analysis

To decide and select the option that best suits your needs, you will find below a summary of the different levels of semantic analysis:

 

Semantic analysis by theme detection

(Level 1)

Semantic analysis by semi-automatic classification plan

(Level 2)

Semantic analysis by automatic classification plan

(Level 3)

Illustration
Components considered
  • Terms
  • Polarities (negative, neutral, positive)
  • Concepts
  • Polarities (negative, neutral, positive)
  • Concepts
  • Polarities (negative, neutral, positive)
Types of analysis
  • Automatic and non-retroactive feature
  • Real-time analysis
  • Semi-automatic feature with human intervention
  • Analysis available at D+2  
  • Fully automatic feature
  • Real-time analysis
Particularities
  • Quick and simple setup for basic needs
  • Longer setup for advanced needs
  • 500 to 600 verbatims required to co-construct the classification plan
  • Long setup for advanced needs
  • Approximately 100 verbatims required per context to co-construct the classification plan
  • Analysis and coding of verbatims by Synapse technology, with direct integration into the interface

Accessibility

  • Submit an activation request to your Customer Success Manager or Skeepers commercial contact : account permissions will be activated and a pack of verbatims will be provisioned
  • Submit a request to your Customer Success Manager or Skeepers commercial contact, who will engage our Professional Services experts to plan several exchanges (needs assessment, setup specifics, installation, etc.)
  • Submit a request to your Customer Success Manager or Skeepers commercial contact, who will engage our Professional Services experts and our partner Synapse to plan several exchanges for the creation of the customized classification plan
Setup time
  • Instantaneous upon activation of the feature
  • 2 weeks of setup
  • 6 to 7 weeks of setup

Semantic analysis by theme detection

Presentation of semantic analysis by theme detection

Semantic analysis by theme detection corresponds to the simplest level of verbatim analysis. It provides a view of the main topics addressed by your customers in their verbatims, as well as the sentiment associated with them. You will be able to :

  • Detect the most used terms in your comments to display them in detail in a raw data visualization or in a word cloud
  • Identify the polarity (positive, neutral or negative) of each sentence and analyze the terms in their context to spot the topics generating the most dissatisfaction and prioritize actions

Requirements

To activate this feature, you must first contact your Customer Success Manager or your commercial contact to grant your account the necessary permissions.

Verbatims are only analyzed if the permissions are activated and the feature has been checked in your text field question. If you deactivate verbatim analysis, the terms will not be analyzed, and your data will be lost, even if you reactivate the option later.

This option is not retroactive. When verbatim analysis is activated during the deployment of a campaign, it will only apply to feedbacks collected after its activation.

This option is available in only 9 languages.

Activate semantic analysis by theme detection

Integrating an open-ended question in a satisfaction survey is essential to allow your customers and/or collaborators to express themselves.

An open-ended question in the form of a 'Text field' element offers several advantages :

  • Better understand the reasons for satisfaction, and especially dissatisfaction
  • Enrich the analysis with more detailed perspectives
  • Capture the mindset of customers through their own words
  • Avoid "frustrating" respondents by allowing them to express themselves on aspects not covered by closed questions, but which are important to them

Once the permissions are activated in your organization's account following your request, you can enable verbatim analysis on an element by accessing your survey and following these steps :

  1. Add a 'Text field' type element.
  2. Open the 'Advanced settings' tab in your open-ended question.
  3. In the 'Verbatim analysis' field, check "Yes" and then validate the element and save your survey.

It is done ! No further action is required. As soon as you collect a few feedbacks to this open-ended question, you can already analyze the different elements.

Integrate verbatims into your analyses

After sending your survey, it is time to analyze the feedback and verbatims from your respondents. With the Skeepers Feedback Management solution, you can analyze verbatims left by your customers in real-time directly from your dashboard (learn more about dashboards).

Depending on your needs, you can configure new indicators (graphs or word clouds) or present your verbatims in raw data visualizations from your dashboard.

Want to integrate verbatims into your analyses ? Check out our dedicated article here : Integrate verbatims into your analyses.

Semantic analysis by semi-automatic classification plan

Presentation of semantic analysis by semi-automatic classification plan

Semantic analysis by semi-automatic classification plan corresponds to the intermediate level of semantic analysis offered by Skeepers. It relies on an internal algorithm supported by Skeepers, followed by manual review and adjustments performed by specialized teams.

This method of semantic analysis ensures remarkable quality by combining the power of automation with human expertise. Manual intervention allows for correcting, refining, and adjusting the results produced by the Skeepers engine, ensuring greater precision and relevance in the analysis. This helps better address the nuances and subtleties of customer feedback, providing more reliable and actionable insights.

Requirements and specifications

To co-construct a relevant classification plan, it is necessary to have a sample of 500 to 600 verbatims. This method is available only in French.

Since the method combines automation and human intervention, it is important to note that the analysis is not real-time : the results will be available within 48 hours after submission of the feedback (D+2).

Activate semantic analysis by semi-automatic classification plan

To activate this feature, please request it to your Customer Success Manager. Several exchanges with the Professional Services team will follow to define the classification plan and configure your semantic analysis project on your account. The complete setup process takes approximately 2 weeks, ensuring a functional and tailored configuration to your specific context.

Configure semantic analysis

Several exchange sessions with the Professional Services team will be organized to define and refine the classification plan together. These exchanges will help better understand your specific needs and precisely configure your semantic analysis project directly on your account. This collaborative approach ensures that the classification plan is perfectly adapted to your objectives, providing more relevant and actionable analyses.

Here are the different steps of the configuration process :

  1. Exchange sessions between you and the Skeepers Professional Services team to define the classification plan
  2. Technical and functional setup of semantic analysis on your Skeepers Feedback Management account by our Professional Services team
  3. Functional support provided by our Professional Services team to understand, take ownership of, and use this semantic analysis feature

Semantic analysis by automatic classification plan

Presentation of semantic analysis by automatic classification plan

Semantic analysis by automatic classification plan corresponds to the most advanced level of semantic analysis offered by Skeepers. In collaboration with our partner Synapse, we offer an automatic classification technology for verbatims.

Once the classification plan is defined with you, Synapse and the Skeepers Professional Services team deploy this technology to analyze and code the verbatims, then integrate the data directly into your Feedback Management interface. Additionally, the engine's performance improves proportionally to the volume of verbatims processed.

Requirements and specifications

To co-construct a relevant classification plan, it is necessary to have a sample of 100 verbatims per concept. This method is available for all languages, ensuring flexibility in an international context.

Since the method relies on automation, the analysis is in real-time.

Activate semantic analysis by automatic classification plan

To activate this feature, please request it to your Customer Success Manager or your Skeepers commercial contact. Several exchanges with the Professional Services team and our partner Synapse will follow to define the classification plan and configure your semantic analysis project on your account. The setup process for this analysis generally takes between 6 and 7 weeks.

Configure semantic analysis by automatic classification plan

To set up semantic analysis by automatic classification plan, the Skeepers Professional Services team and the partner Synapse will assist you in developing and configuring the classification plan on the Feedback Management platform.

Here are the different steps of the configuration process :

  1. Exchange sessions between you, the Skeepers Professional Services team, and Synapse to define the classification plan
  2. Technical and functional setup of semantic analysis on your Skeepers Feedback Management account by our Professional Services team
  3. Functional support provided by our Professional Services team to understand, take ownership of, and use this semantic analysis feature

Verbatim desensitization

There are several reasons why you might want to exclude certain words or expressions from your verbatim analysis. For example, it is common to remove personal information such as bank details or phone numbers, as well as religious or offensive statements. You might also want to suppress specific terms related to your business or industry.

To set up verbatim desensitization in an open-ended question ("Text field"), we invite you to follow the steps described in our dedicated article : Use verbatim desensitization to mask sensitive data.

Semantic analysis by automatic classification plan with a third-party provider

Want to perform semantic analysis in Skeepers Feedback Management with other tools ? Leverage classification plans designed with services like Erdil, Proxem, or Viavoo to optimize your semantic analysis.

ERDIL (@erdil_sa) / X     Proxem | Parisviavoo recrute des talents informatiques | ChooseYourBoss

This method allows you to :

  • Improve customer listening by making feedback richer and more relevant
  • Gain in-depth knowledge of your customers by analyzing their verbatims in a complementary way to your survey results
  • Identify the main factors of satisfaction and dissatisfaction, and spot opportunities for improvement

To set up a classification project, contractualization is necessary with one of our partners.

To learn more, feel free to consult our dedicated article : Connect Skeepers to semantic analysis tools

Best practices

To help you make the most of your semantic analysis, we have compiled articles available on our blog that detail best practices. With our expertise and experience with various partners and specific cases, we recommend consulting these resources to optimize your analysis methods and obtain more relevant and actionable results :

 

If you have any questions, feel free to contact our Customer Care team.

 

Was this article helpful?
1 out of 2 found this helpful