How to Measure Customer Support Quality
Customer support quality cannot be measured by gut feel or anecdote — it requires consistent tracking of specific metrics over time. The challenge is that most support teams track too many metrics superficially, or track the wrong ones entirely. This guide covers the 12 KPIs that genuinely reflect support quality, how to calculate each one, and what benchmark figures indicate good performance.
Why Most Teams Track the Wrong Metrics
Ticket count is the most commonly reported support metric. It is also one of the least useful. A team closing 500 tickets a day looks productive — but if 400 of those tickets reopened within 48 hours, or if CSAT scores are declining, high ticket count signals dysfunction rather than performance.
The metrics worth tracking fall into four categories: speed, quality, efficiency, and customer experience. Tracking at least two from each category gives a balanced picture of support performance.
Speed Metrics
1. First Response Time (FRT)
What it measures: How long a customer waits from sending their first message to receiving any reply.
How to calculate: Average time between ticket creation and first agent (or AI) response, across all tickets in the period.
Benchmarks by channel:
| Channel | Good | Acceptable | Poor |
|---|---|---|---|
| Live chat | Under 1 min | Under 5 min | Over 10 min |
| Under 4 hours | Under 24 hours | Over 24 hours | |
| Social media | Under 1 hour | Under 4 hours | Over 8 hours |
Why it matters: First response time is the single metric customers feel most directly. A fast first response — even if it is an acknowledgement rather than a resolution — reduces anxiety and sets expectations.
2. Average Resolution Time (ART)
What it measures: How long it takes to fully resolve a customer's issue from the moment they first contact support.
How to calculate: Average time between ticket creation and ticket closure (marked resolved), across all tickets.
Benchmark: Under 24 hours for tier-1 issues; under 72 hours for complex cases.
Why it matters: FRT tells you how quickly you acknowledge — ART tells you how quickly you actually solve. A team with a fast FRT and a slow ART is good at acknowledging problems but poor at fixing them.
3. Time to First Resolution (TTFR)
What it measures: Time to first resolution on initial contact — excludes reopened tickets.
Why it differs from ART: ART includes all reopen cycles. TTFR measures whether your team is solving issues cleanly the first time. A team with low ART but high TTFR is reopening tickets frequently, which inflates efficiency figures.
Quality Metrics
4. Customer Satisfaction Score (CSAT)
What it measures: Customer satisfaction with a specific support interaction, measured immediately after resolution.
How to calculate: (Number of positive ratings ÷ total ratings) × 100. Typically collected via a 1-click post-resolution survey ("Was this helpful?").
Benchmark: 85%+ is good; 90%+ is excellent; below 75% requires investigation.
Why it matters: CSAT is the most direct measure of whether customers felt their issue was handled well. Tracked over time, it reveals whether process changes improve or degrade the customer experience.
5. First Contact Resolution Rate (FCR)
What it measures: The percentage of tickets resolved in a single interaction, without the customer needing to follow up.
How to calculate: (Tickets resolved on first contact ÷ total tickets) × 100.
Benchmark: 70–75% is industry average; 80%+ is strong.
Why it matters: FCR is the most reliable indicator of support quality. A ticket resolved first time means the agent understood the problem, had the knowledge to solve it, and communicated clearly. A ticket requiring multiple contacts means at least one of those things failed.
6. Ticket Reopen Rate
What it measures: The percentage of resolved tickets that customers reopen because the issue was not actually fixed.
How to calculate: (Reopened tickets ÷ total resolved tickets) × 100.
Benchmark: Under 5% is good; over 10% signals systematic quality problems.
Why it matters: A high reopen rate means your resolution rate is artificially inflated. Tickets marked resolved that immediately reopen are not resolved — they are deferred.
Efficiency Metrics
7. Ticket Deflection Rate
What it measures: The percentage of potential support contacts resolved through self-serve channels (knowledge base, AI, FAQ) without reaching a human agent.
How to calculate: (Self-serve resolutions ÷ (self-serve resolutions + agent-handled tickets)) × 100.
Benchmark: 30–40% for teams with basic self-serve; 60–70% for teams with AI automation.
Why it matters: Deflection rate directly measures how effectively your self-serve and AI infrastructure is working. Every deflected ticket is a ticket your agents do not handle — with no degradation in customer experience when the deflection delivers accurate answers.
8. Cost Per Ticket
What it measures: The fully-loaded cost of handling one support ticket.
How to calculate: (Total support costs for period ÷ total tickets handled in period). Include agent salaries, benefits, software subscriptions, and management overhead.
Benchmark: £3–£8 for AI-assisted teams; £8–£18 for manual-only teams.
Why it matters: Cost per ticket is the metric that translates support quality improvements into financial terms. Reducing cost per ticket from £12 to £4 while maintaining CSAT is the quantified ROI of AI investment.
9. Agent Utilisation Rate
What it measures: The percentage of agent working hours spent actively handling tickets versus administrative tasks, meetings, and idle time.
How to calculate: (Time spent on tickets ÷ total working hours) × 100.
Benchmark: 65–75% is healthy. Above 85% indicates understaffing; below 50% indicates inefficiency.
Why it matters: Low utilisation may indicate agents are spending time on avoidable administrative work. High utilisation consistently over 80% is a burnout risk and signals a need for either automation or additional headcount.
Customer Experience Metrics
10. Net Promoter Score (NPS)
What it measures: How likely customers are to recommend your business, influenced significantly by their support experience.
How to calculate: Survey customers: "How likely are you to recommend us? (0–10)." Promoters (9–10) minus Detractors (0–6) = NPS.
Benchmark: NPS above 50 is excellent; 30–50 is good; below 0 indicates systemic problems.
Why it matters: NPS connects support quality to revenue. Customers who have a positive support experience are significantly more likely to repurchase and refer others. Customers who have a poor experience are the most common source of negative reviews.
11. Customer Effort Score (CES)
What it measures: How much effort a customer had to exert to resolve their issue.
How to calculate: Post-resolution survey: "How easy was it to resolve your issue today? (1–7)." Average of all responses.
Benchmark: Above 5.5 is good; below 4 indicates friction in your support process.
Why it matters: CES predicts churn better than CSAT in many industries. Customers who found support easy to use are more likely to stay. Those who found it effortful — even if their issue was eventually resolved — are at higher churn risk.
12. Support-Driven Churn Rate
What it measures: The percentage of churned customers who had an unresolved or negatively rated support interaction in the 30 days before churning.
How to calculate: (Churned customers with negative support history in prior 30 days ÷ total churned customers) × 100.
Benchmark: Varies widely; tracking the trend matters more than the absolute number.
Why it matters: This metric directly connects support quality to revenue retention. If a significant percentage of churned customers had a poor support experience beforehand, support quality is a retention lever — not just a cost centre.
How to Build a Support Metrics Dashboard
Track these 12 metrics in a simple weekly dashboard. Priorities:
Review weekly: FRT, CSAT, FCR, ticket volume Review monthly: ART, deflection rate, cost per ticket, reopen rate Review quarterly: NPS, CES, agent utilisation, support-driven churn
Use trends, not snapshots. A CSAT of 82% means little in isolation. CSAT declining from 91% to 82% over three months is a signal requiring action.
Key Takeaways
- Track metrics across four categories: speed, quality, efficiency, and customer experience
- First Contact Resolution Rate is the single most reliable indicator of support quality
- Ticket count and closure rate are lagging indicators — FCR, CSAT, and reopen rate are more diagnostic
- Cost per ticket translates support improvements into financial terms for business cases
- Use weekly reviews for speed and quality metrics; monthly for efficiency; quarterly for experience metrics
- AI automation directly improves deflection rate, FRT, cost per ticket, and — indirectly — CSAT and NPS
Frequently Asked Questions
What are the most important customer support KPIs? The most important KPIs depend on your business stage, but First Contact Resolution Rate, CSAT, First Response Time, and Cost Per Ticket cover the four dimensions of support quality most comprehensively. FCR tells you if issues are solved properly, CSAT tells you if customers are satisfied, FRT tells you if you are fast enough, and cost per ticket connects performance to business economics.
What is a good CSAT score for customer support? A CSAT score above 85% is considered good across most industries. Scores above 90% are excellent. Below 75% indicates systemic problems with resolution quality, agent training, or knowledge base accuracy. CSAT should be tracked as a trend rather than a point-in-time figure — a declining score is more important to act on than a stable low score.
What is a good First Contact Resolution rate? The industry average FCR is 70–75%. Teams with well-maintained knowledge bases and AI assistance typically achieve 80–85%. FCR above 85% is considered excellent. FCR below 60% suggests agents lack the knowledge, tools, or authority to resolve issues in a single interaction.
How do I reduce cost per ticket? The fastest ways to reduce cost per ticket are: increasing ticket deflection through AI and self-serve tools, reducing average handle time through better knowledge base access and AI-generated draft replies, and reducing reopen rates by improving first-contact resolution quality. AI customer support platforms typically reduce cost per ticket by 50–70% in mature deployments.
Should I track all 12 KPIs from the start? No. Start with four: CSAT, First Response Time, First Contact Resolution Rate, and ticket volume. Add cost per ticket and deflection rate once you have baseline data. Add the remaining metrics as your team and tooling mature. Tracking 12 metrics from day one without baseline data produces noise rather than insight.