Statistical Mode in CFM Copilot helps you move beyond surface-level metrics and understand the underlying drivers of your CX outcomes. Instead of only showing scores, it identifies why those scores exist by applying appropriate statistical methods to your survey data.

Business Use Cases

• Identify What Drives NPS: Statistical models assist in pinpointing the most significant drivers that contribute to NPS. Copilot utilizes linear regression to measure the impact of each variable. This enables you to differentiate between factors that are associated with NPS and those that genuinely affect it.

  You can prioritize improvements based on statistical significance and impact.

• Detect Regional or Segment-Based Patterns: To assess if customer experience problems are confined to specific areas or are more widespread, Copilot employs a Chi-Square test alongside Cramér’s V to analyze the connections between categorical variables. This approach aids in determining if differences among regions or segments hold statistical significance.

  This allows you to concentrate on specific interventions rather than implementing sweeping changes.

• Validate Assumptions About CX Drivers: When evaluating hypotheses like the effect of wait time on satisfaction, Copilot performs a Pearson correlation analysis for numerical variables. This analysis assesses both the strength and significance of the relationship.

  This approach guarantees that decisions are grounded in data instead of mere assumptions.

• Understand Relationships Between Ratings: To relate various customer experience attributes, Copilot employs Spearman rank correlation for ordered rating scales. It determines if metrics change in tandem and the degree of their association. This aids in revealing dependencies among experience factors.

• Compare Performance Across Groups: To evaluate performance variations among segments, Copilot utilizes an ANOVA test to compare metrics across various groups. It takes into consideration differences in sample size and confirms statistical significance. This helps distinguish real variation from noise.

• Identify What Drives Promoters: To comprehend what enhances the chances of customers turning into Promoters, Copilot employs logistic regression to assess the influence of drivers on binary results. It emphasizes elements that notably raise or lower the probability of becoming a Promoter. This enables you to replicate experiences with a substantial impact.

• Analyze Drivers Across NPS Segments: To obtain insights that extend beyond Promoters and to comprehend all NPS segments, Copilot employs multinomial regression for outcomes with multiple categories. It pinpoints factors that cause customers to move among the Promoter, Passive, and Detractor segments.

  This assists in balancing efforts to enhance positive experiences while minimizing negative ones.

Recognize the genuine factors influencing CX metrics. Make informed decisions based on data with statistical confirmation. Concentrate on significant enhancements. Obtain more profound insights across various segments and customer journeys.

Prerequisites

You would need CFM Copilot permissions at the App Level.

Navigation

Follow the below steps to navigate through the platform:

1. Navigate to Spinklr Insights and then access Customer Feedback Management Persona App.

2. You can access Co-pilot in these two ways:

   1. Analytics Tab: First, access a Survey. From there, navigate to the Analytics tab to open Sprinklr Copilot.

   2. Custom Dashboards: In the left pane, proceed to Custom Dashboards. Then, choose a custom dashboard of your preference, and finally, open Sprinklr Copilot.

3. Click the Sprinklr Copilot button towards the top right of the Dashboard. This will open Sprinklr Copilot in the sidepane.

4. Click Try Different Copilot Modes (+ option) and select Statistical Analysis.

5. In the Ask Sprinklr Copilot anything space, enter your query in plain language (for example, “What is driving NPS this quarter?”).

6. Review the generated output, including:

   1. Key findings

   2. Statistical method used

   3. Significance levels

7. Use follow-up questions to refine your analysis (for example, apply filters such as region or time period).