CSAT Report

Generate a structured CSAT report with score trends, satisfaction driver analysis, segment breakdowns, improvement recommendations, and industry benchmarks.

Workflow

Step 1: Collect CSAT Data

Gather satisfaction scores and associated metadata:

• Score data: Individual CSAT responses with timestamps and ratings.
• Response metadata: Ticket category, product area, agent, channel, resolution time.
• Customer metadata: Account tier, region, tenure, ARR.
• Response rate: Total tickets vs. surveys sent vs. responses received.
• Comments: Verbatim feedback attached to scores.

Step 2: Calculate Core Metrics

Compute the standard CSAT metrics:

• Overall CSAT score: Percentage of positive responses (typically 4-5 on a 5-point scale).
• Period-over-period change: Compare to the previous period of the same length.
• Response rate: Percentage of customers who completed the survey.
• Distribution: Breakdown across all score values (1 through 5).
• Median and mean: Both central tendency measures for nuanced interpretation.

Step 3: Segment Analysis

Break down scores across meaningful dimensions:

• By team or agent: Which teams are performing above or below average?
• By category: Bug reports vs. how-to questions vs. billing issues.
• By channel: Email vs. chat vs. phone.
• By customer tier: Enterprise vs. mid-market vs. self-serve.
• By resolution time: Correlation between speed and satisfaction.
• By first-contact resolution: FCR vs. multi-touch impact on scores.

Step 4: Driver Analysis

Identify what drives high and low scores:

• Top positive drivers: Factors most correlated with 4-5 ratings.
• Top negative drivers: Factors most correlated with 1-2 ratings.
• Comment themes: Common praise and criticism from verbatim responses.
• Controllable vs. uncontrollable: Separate issues the support team can fix from product or process issues.

Step 5: Benchmark and Recommend

Compare performance and propose improvements.

Output Format

## CSAT Report: <Period>

**Overall CSAT**: <score>% (<change> vs. prior period)
**Response Rate**: <pct>% (<surveys returned> / <surveys sent>)
**Total Responses**: <count>

### Score Distribution
| Score | Count | Percentage | Change |
|-------|-------|------------|--------|
| 5 (Excellent) | <n> | <pct>% | <+/- pct> |
| 4 (Good) | <n> | <pct>% | <+/- pct> |
| 3 (Neutral) | <n> | <pct>% | <+/- pct> |
| 2 (Poor) | <n> | <pct>% | <+/- pct> |
| 1 (Bad) | <n> | <pct>% | <+/- pct> |

### Segment Breakdown
| Segment | CSAT | Volume | Change | Notable |
|---------|------|--------|--------|---------|
| <segment> | <pct>% | <n> | <+/- pct> | <key observation> |

### Satisfaction Drivers

**Top Positive Drivers**
1. <driver> -- seen in <pct>% of positive responses
2. <driver> -- seen in <pct>% of positive responses

**Top Negative Drivers**
1. <driver> -- seen in <pct>% of negative responses
2. <driver> -- seen in <pct>% of negative responses

### Industry Benchmark
| Metric | Our Score | Industry Median | Percentile |
|--------|-----------|-----------------|------------|
| Overall CSAT | <pct>% | <pct>% | <nth> |
| Response rate | <pct>% | <pct>% | <nth> |

### Recommendations
1. **<recommendation>**: <expected impact and effort>
2. **<recommendation>**: <expected impact and effort>
3. **<recommendation>**: <expected impact and effort>

### Tracking
| Initiative | Status | Impact to Date |
|------------|--------|----------------|
| <prior recommendation> | <status> | <measured effect> |

Examples

User: "Generate the monthly CSAT report for March" Action: Pull March CSAT data, calculate overall and segmented scores, compare to February, identify drivers, produce the report.

User: "Our CSAT dropped 5 points -- help me figure out why" Action: Compare current and prior period by segment, identify which categories or teams drove the decline, analyze negative comments, recommend corrective actions.

User: "CSAT report by customer tier for the Q1 board deck" Action: Segment Q1 data by enterprise, mid-market, and self-serve tiers, highlight tier-specific drivers, format for executive consumption with benchmarks.

Best Practices

• Always report response rate alongside CSAT score. A 95% CSAT with a 5% response rate is less meaningful than an 85% CSAT with a 40% response rate.
• Use benchmarks carefully. Industry benchmarks vary by sector, support model, and survey methodology.
• Distinguish between statistical noise and real trends. A 1-2 point swing on low volume is not actionable.
• Track driver changes, not just score changes. The same score can mask shifting underlying causes.
• Include prior recommendations and their impact to show whether improvement efforts are working.
• Present actionable recommendations with effort and expected impact estimates, not just observations.
• Segment by controllable factors (agent, process) and uncontrollable factors (product bugs) to direct effort appropriately.
• Set realistic improvement targets. Moving CSAT from 80% to 85% is harder than 60% to 70%.
• Report on detractor follow-up rates if your team has a closed-loop process.