When measuring customer support (CS) quality, "satisfaction (CSAT) alone is insufficient." You also need "did the issue get resolved" and "how much effort did it take" — only then do improvement priorities become visible.
This article covers the design framework for support surveys and how to convert results into action.
Three metrics for support quality
1. CSAT (Customer Satisfaction Score)
"How satisfied were you with the support?" on a 5-point scale. The simplest metric.
2. CES (Customer Effort Score)
Measures how much effort the customer had to expend to resolve the issue. Has stronger predictive power for churn than CSAT.
3. Resolution rate
"Was your issue resolved? (yes/no)" — a simple binary.
Combining the three lets you separate improvement areas:
- Resolved with low effort (high CSAT, high CES, resolved ○) → ideal
- Resolved but high effort (mid CSAT, low CES, resolved ○) → improve self-service paths
- Not resolved (low CSAT, resolved ✗) → structural problem in support setup
Standard 5-question version
Q1. How satisfied are you with the support you received? (5-point)
※ CSAT
Q2. "Thanks to support, my issue was easy to resolve" — how much do you agree? (5-point)
※ CES
Q3. Was your issue resolved?
○ Yes, fully
○ Partially
○ Not resolved
※ Resolution rate
Q4. Anything the agent did well or that could be improved? (open text, optional)
Q5. Any other feedback about support overall? (open text, optional)
5 questions in under 2 minutes. Good for "right after support contact" distribution.
Extended 10-question version
For deeper probing into issue characteristics:
[Q1–Q3 as standard]
Q4. How long did resolution take?
○ Immediate (<5 min)
○ Within 1 hour
○ Same day
○ Next day or later
Q5. Did you try to self-serve before contacting support?
○ Read FAQ
○ Searched help center
○ Looked at past tickets
○ Didn't try
Q6. Why couldn't you self-serve? (Multiple)
□ Couldn't find relevant info
□ Documentation unclear
□ Tried but couldn't solve it
□ Didn't think to try
Q7. How was the contact channel? (5-point)
Q8. Was the agent's expertise sufficient? (5-point)
Q9. For similar issues in future, what would help most?
○ Improve FAQ / help
○ Stronger chatbot / AI
○ Improve the product itself
○ Expand support team
Q10. Other feedback (open text, optional)
This expansion surfaces self-service improvement opportunities too.
Distribution timing
Immediate (recommended)
Right after support resolution — in the chat widget, in the email reply, on the ticket-closure screen.
- Fresh memory → high accuracy
- No distribution delay → higher response rate
Delayed
Email 1–3 days after closure.
- Pro: captures the "after they've cooled off" assessment
- Con: lower open and response rates
Combination approach: capture 1 question (CSAT) on the closure screen, send remaining questions by email later.
Turning results into action
1. Auto-alert on low scores
CSAT 1–2 or "not resolved" responses → auto-notify the support manager → immediate follow-up.
2. Per-agent trend analysis
Aggregate CSAT and CES per agent. If a specific agent is consistently low, training is the answer.
3. Per-category trend
Aggregate "why I couldn't self-serve" to prioritize FAQ expansion.
4. Continuous self-service improvement
Quickly-resolved tickets often could have been self-served. They're prime FAQ expansion candidates.
Agent feedback design
CS work has well-known motivation challenges. Tying survey scores too tightly to staff evaluations creates distortions:
- Agents avoid difficult tickets
- Agents start begging customers for surveys
Recommended operation
- Team-level scores as the primary metric
- Individual scores used for coaching, not bonuses
- "What went well" open text → fuel for internal recognition
What not to do
- 10+ questions: a long survey right after a support interaction adds insult to injury
- Track only satisfaction: without resolution + CES, you can't see what to improve
- Pressure individual agents: keep it at team level
- Ignore the data: no improvement actions → field staff feel "this is just extra work"
Summary
Three pillars of support quality measurement:
- CSAT — satisfaction
- CES — effort
- Resolution rate — outcome
Combine these, then run the improvement cycle on three axes: self-service paths, FAQ, and agent skill.
Repoan's customer support post-inquiry survey template ships with this 10-question structure.
Embed the survey on the support closure screen (see embedding surveys in your site) for immediate measurement. AI reports (see AI-driven response analysis) auto-classify "why I couldn't self-serve" from open text, turning FAQ expansion priorities into a data-driven decision. CES agreement-format questions (see CES — Customer Effort Score) are supported out of the box.