Sprinklr AI+ Integration in Quality Scoring
Updated
What is AI Quality Scoring?
AI Quality Scoring in Quality Management allows Quality Managers to efficiently assess the performance of their customer service agents. With the rapid advancements in AI technology, the evaluation process has become faster and more accurate than ever before. Sprinklr AI+ enables the analysis of agent-customer conversations based on multiple quality parameters, including but not limited to the quality of the initial interaction, attitude, communication skills, and the quality of the closing interaction.
AI Insights in AI Quality Scoring
While Sprinklr AI+ assigns weighted scores to various quality parameters, it may lack in-depth insights. AI Insights in AI Quality Scoring addresses the need for more granular understanding of quality parameters in Quality Management.
AI Insights offers a solution by allowing Quality Managers to access AI-detected phrases/messages associated with the quality parameters. Quality Managers can drill down on each parameter in AI Score Breakdown widget and explore the highlighted phrases identified by the AI. Positive inferences are indicated by green highlighting, while negative inferences are highlighted in red.
Use Cases of AI Quality Scoring
Empathetic and Effective Communication: Sprinklr AI+ can analyze customer interactions to identify phrases and approaches that customers find empathetic and effective. This information serves as guidance for agents to enhance their communication skills and provide better customer experiences.
Deeper Insights into Performance: Sprinklr AI+ quality scoring provides Quality Managers with deeper insights into their agents' performance. By evaluating various quality parameters, Sprinklr AI+ offers a comprehensive assessment of agent capabilities, enabling managers to pinpoint specific areas for improvement.
Identification of Improvement Areas: Through Sprinklr AI+ quality scoring, Quality Managers can identify specific areas where agents may need additional training or support. This helps managers create targeted coaching programs and interventions, ensuring agents have the necessary skills and knowledge to deliver exceptional customer service.
Enhanced Customer Service and Satisfaction: By leveraging Sprinklr AI+ quality scoring, Quality Managers can drive improvements in customer service. The insights gained from AI analysis enable managers to implement strategies that lead to better customer experiences, resulting in increased customer satisfaction and loyalty.
How to View AI Quality Scoring?
Sprinklr AI+ analyzes 100% of the daily case conversation data for quality scoring. In the Case Analytics View for any case conversation, AI Score Breakdown provides the overall and parameter-wise AI quality scores. AI insights provide detailed explanations and recommendations for each quality parameter.
Sprinklr AI+ Quality Parameters
Following are few examples of Sprinklr AI+ quality parameters:
L1 Category | L2 Sub-category | Description | Examples |
Opening Quality | Greeting | Did the agent open the conversation with a Greeting? | - Hi, Hello, Hey - Good Morning/afternoon |
Opening With Brand Mention | Did the agent use Brand name while opening the conversation? | - Thank you for contacting Acme care - Thank you for reaching out to Sprinklr support | |
Introducing using Agent name | Did the agent introduce himself/herself by name? | I am John and I will be assisting you today | |
Closing Quality | Gratitude | Did the agent expresses gratitude while closing the case conversation? | - Pleasure to assist - You're Welcome - Thank you for your time. |
Closing With Brand Mention | Did the agent use Brand name while closing the conversation? | - Thank you for contacting sprinklr support | |
Feedback | Did the agent ask for feedback from customer on overall customer experience? | - If I was helpful, you could take a quick survey at the end of the chat - Please rate our conversation | |
Further Assistance | Did the agent confirm with the customer that any further assistance is required or not? | - Do let us know in case of further assistance. - Do reach out to us in case of further enquiries | |
Effectiveness | Efficiency | Customer messages which depicts that they have to repeat themselves during the conversation. | Example of customer verbatims: - I have told you that now like 7 times - I have repeated myself over and over - Have you read what I wrote? - Please read the chat above |
Attitude | Courtesy | Did the agent appreciate the customer for their time/loyalty and patience? | - We appreciate your cooperation - Thanks for bringing this to our attention |
Accountability | Did the agent response highlights ownership and can-do attitude? | - I will definitely help you with best possible options - I will do my best to alleviate your concern | |
Patience | Did the agent allow appropriate time to customers during case resolution: | - Please take your time, i will be right there - I will stay connected with you while you place an order | |
Promptness | Customer messages which depicts slow customer service/responses from the brand: | - I've been waiting for your response from so long - How long will it be for take for you to respond | |
Empathy | Did the agent response highlights that he/she can feel what customer is going through? | - I can imagine the trouble that your device is causing you. - I can totally understand your situation. - I apologize for the inconvenience caused | |
Communication Skills | Grammar | Did the agent commit any spelling mistakes/used incorrect verbs/nouns during conversation? | - I hvae filed the tciket. (Wrong Spelling) |
Profanity | Did the agent use any Profane/Inappropriate words during conversation? | - Bloody hell |
Note: Sprinklr AI+ also provides customizable AI quality parameters. You can refer this article on Customized use cases for AI quality scoring
Sprinklr AI+ Quality Scoring Criteria
Quality scores are determined using two scoring criteria:
Binary scoring - Evaluates agents on a simple yes or no basis, with a score of 100 for a positive result and 0 for a negative result.
Delta scoring - Assigns a penalty based on the number of mistakes made, with the penalty subtracted from the total score.
In either case, the highest possible score for any quality parameter is 100 and the lowest possible score is 0.