Speech and text analytics

1. Click Admin.
2. Under Quality, click Speech and Text Analytics.
3. Select the Settings tab.
4. Under Voice Transcription, select Enabled based on Queue configuration or Flow action.
5. Under Low Latency Transcription, select Enabled if you want to minimize the latency of all transcripts sent through the Notification API.
6. Under Transcript Confidence Filter, set the minimum confidence threshold required for a word to be displayed in a transcript. The default value for Word Confidence is set to 40.
7. Turn on Transcript Content Search if you want all voice transcripts and digital interaction content to be searchable for a limited number of days.
8. Click Save.
9. Complete the remaining steps in .

1. Click Admin.
2. Under Quality, click Speech and Text Analytics.
3. Select the Settings tab.
4. Under Default Program, select the fallback program to use for topic detection if no program is mapped to a Queue or Flow.
5. Complete the remaining steps in .

1. Click Admin.
2. Under Quality, click Speech and Text Analytics.
3. Under Customer Sentiment Analysis, select Enabled if you want to automatically detect instances in an interaction when the customer expresses a positive or negative attitude.
4. Click Save.

1. Click Admin.
2. Under Quality, click Speech and text analytics.
3. Select the Actions tab.
4. From the Select dialect… list, select the dialect that best represents your interactions.
5. From the Add new topics only list, select one of the following options:

   • Add new topics only – enables you to add one or more topics to the topic list.

   • Merge topics and phrases – enables you to merge new topics with existing topics that have the same name and selected dialect. This may reinstate Topics you previously deleted.

   • Replace and Merge – replaces unmodified topics to get the latest version, merges updated topics with the latest phrase updates and deleted deprecated topics and phrases.

6. Click Generate, to generate the new out-of-the-box topics.

This section is designed to provide you with a walk-through of important information in the required order to successfully get your Speech and Text Analytics solution configured and up and running. The information in this section, assumes that you have already set up a Genesys Cloud organization and have either Genesys Cloud CX 3 level licenses or WEM add-on in place.

Step 1: Define initial KPIs and Goals

Speech analytics implementations should start with a narrow focus and target a specific key performance indicator (KPI). The KPI must be clearly defined and measurable, have financial impact and/or be recognized as important to the business.

Common KPIs to focus your attention and tailor your speech and text analytics solution are:

• Sales Improvement
• First Call Resolution
• Call Volume Reduction
• Handle Time Optimization
• Compliance Management
• Customer Retention

The information outlined in the  article will guide you through this step.

Step 2: Turn on voice transcription

Voice transcription transcribes contact center voice interactions (that is, audio) into text that is stored as speaker separated conversational language. Voice transcripts generated by the system are shown in the Transcript tab as part of the interaction detail. The information outlined in the  article will guide you through this step.

Step 3: Set or create your default program

Programs are packages of topics that instruct speech and text analytics which business level intents to look for in recorded conversations between interaction participants. Programs are mapped to specific queues or flows and can contain topics of varying languages and dialects. 

The information outlined in the article will guide you through creating and editing a program.

Step 4: Define initial KPIs and Goals

Speech analytics implementations should start with a narrow focus and target a specific key performance indicator (KPI). The KPI must be clearly defined and measurable, have financial impact and/or be recognized as important to the business. The information outlined in the article will guide you through this step.

Step 5: Setting up Out-of-the-box topics

A topic is made up of phrases that represent a specific intent (for example, cancellation). Each program is associated with one or more topics. When a topic is included in a program, the system searches for all the phrases included in the topic’s definition, in all the interactions associated with the program. When one of the phrases is found, it is identified as an event and the topic is registered as found at a specific time during the interaction. The  article will allow you to review out-of-the-box topics and decide which ones you want to deploy.

Step 6: Create custom topics

Topics help to boost the recognition of specific words and phrases in voice transcription. That is, they adapt the underlying language model to look for organization-specific language in conversations. For this reason, it is key to create a set of topics that not only supports the goal of the initial speech analytics implementation, but also includes topics and phrases that are specific to your industry and business. The article will guide you through creating and editing a topic.

Step 7: Search and retrieve actionable data using the content search view

Use this view to search for interactions that contain specific words that are included or are not included in a transcript. You can also use this view to filter interactions with transcripts by additional interaction details.  The  article provides you with the required information to create your own search queries and retrieve actionable data

Step 8: Using sentiment analysis

Sentiment analysis is the interpretation and classification of phrases within an interaction based on the attitude expressed by the customer (positive, negative, and neutral). By capturing the sentiment of the customer’s phrases, users can gain valuable insight into the customer’s experience and can use this information to improve service delivery. The  article provides you with the required information to understand and utilize sentiment analysis data.