Customer Service Metrics Every Seller Should Track
First Response Time
First response time (FRT) measures the elapsed time between when a customer submits a support request and when they receive a meaningful first reply from your team. This is the single most impactful metric in customer service because the longer a customer waits, the more frustrated they become, and frustration compounds with every passing hour. Research from SuperOffice found that the average ecommerce first response time is 12 hours, but top-performing stores average under 2 hours, and customer satisfaction drops measurably for every hour beyond the 4-hour mark.
Benchmarks by channel: Email should target under 4 hours during business hours and under 12 hours including off-hours. Live chat should target under 60 seconds. Social media messages should target under 2 hours. Phone should target under 3 rings or 20 seconds.
How to improve FRT: Set up help desk alerts that notify agents when tickets approach their response deadline. Use auto-replies on email to acknowledge receipt and set expectations. Staff live chat during peak traffic hours rather than trying to cover the entire day. Implement a chatbot that provides instant responses to routine questions, which handles 40% to 60% of inquiries without human involvement and dramatically reduces the FRT for the remaining tickets that need human attention.
What to watch for: Track FRT separately for each channel and each day of the week. If your Monday FRT is twice your Wednesday FRT, you have a weekend ticket backlog that needs either off-hours staffing or a larger Monday support allocation. If your live chat FRT exceeds 2 minutes during certain hours, you are understaffed for those peak periods.
First Contact Resolution Rate
First contact resolution (FCR) measures the percentage of support tickets that are fully resolved in a single interaction without requiring follow-up messages, callbacks, or escalation. This metric matters because every additional touch on a ticket doubles the resolution time, doubles the agent effort, and doubles the customer frustration. A store with 80% FCR resolves 4 out of 5 tickets in one reply. A store with 50% FCR requires multiple interactions for half its tickets, effectively doubling its support workload.
Benchmarks: Top-performing ecommerce stores achieve 70% to 85% FCR. The industry average sits around 60% to 70%. If your FCR is below 55%, your team is systematically failing to resolve issues on the first try, which means either they lack the information they need, the training to handle common scenarios, or the authority to take action without escalation.
How to improve FCR: Build comprehensive response templates that anticipate follow-up questions and address them proactively. Give agents direct access to order data, shipping status, and refund tools within their help desk so they do not need to switch between systems. Empower agents to issue refunds, send replacements, and apply discounts up to a defined dollar amount without manager approval. Train agents to identify the underlying need behind the stated question, because a customer asking "has my order shipped?" really needs the tracking link, the expected delivery date, and confirmation that everything is on track.
What to watch for: Track FCR by ticket category. If your FCR for shipping questions is 90% but your FCR for return requests is 40%, the return process needs simplification or your agents need better tools for processing returns within the help desk. Also track FCR by individual agent to identify training opportunities and top performers whose techniques can be shared with the team.
Customer Satisfaction Score (CSAT)
CSAT measures how customers rate their support experience after a ticket is resolved. The most common format is a simple survey sent after ticket closure asking "How would you rate your experience?" on a 1-to-5 scale. CSAT is calculated as the percentage of respondents who give a 4 or 5 rating. It is the most direct measure of whether your customers are happy with the help they received.
Benchmarks: Ecommerce CSAT targets are typically 85% to 95% positive. Scores above 90% indicate excellent support quality. Scores between 80% and 90% are acceptable but have room for improvement. Scores below 80% signal systemic problems that are actively damaging customer relationships and retention.
How to measure: Most help desk platforms include built-in CSAT surveys that send automatically after ticket resolution. Keep the survey to a single question with an optional comment field. Multi-question surveys have much lower response rates without providing proportionally more insight. Expect 10% to 25% of customers to respond. Low response rates do not invalidate the data as long as you have a reasonable sample size (at least 30 to 50 responses per month).
How to improve CSAT: Read every piece of negative CSAT feedback to understand what went wrong. Common themes include slow response times (which shows up in your FRT metric), needing to explain the issue multiple times (which shows up in your FCR metric), unhelpful or robotic responses, and policy rigidity where agents followed the rules rather than solving the customer's problem. CSAT improvements usually come from improving the metrics that feed into it: faster responses, better first-contact resolution, warmer communication tone, and greater agent empowerment to resolve issues generously.
Ticket Volume and Distribution
Ticket volume tracks the total number of support requests your team handles over a given period, and the distribution reveals what customers are contacting you about. This metric is less about performance and more about operational intelligence, because ticket volume data tells you where your business is creating unnecessary friction that generates avoidable support requests.
Ticket-to-order ratio is the key derivative metric. Calculate it by dividing total support tickets by total orders in the same period. A healthy ecommerce store generates 1 support ticket per 8 to 15 orders. If your ratio is worse (1 ticket per 5 orders or fewer), something in your product, website, or fulfillment process is systematically creating problems. If your ratio is better (1 ticket per 20+ orders), either your self-service resources are excellent or customers cannot find your contact information.
Category distribution reveals your biggest friction points. Tag every ticket with a category (shipping, returns, product question, billing, account issue, complaint) and review the distribution monthly. If 30% of tickets are "where is my order" questions, invest in proactive shipping notifications and a self-service tracking page. If 20% are sizing questions, improve your product page size charts. If 15% are return requests, examine whether your product descriptions or photos are setting inaccurate expectations. Each category represents a specific operational problem with a specific operational fix that permanently reduces future ticket volume.
Volume trends matter as much as the absolute number. If ticket volume is growing faster than order volume, your support burden is outpacing your business growth, which means something is getting worse. If ticket volume grows slower than order volume, your self-service improvements and process fixes are working. Track this ratio monthly to ensure your support operation scales sustainably.
Cost Per Resolution
Cost per resolution (CPR) measures the average cost to resolve a single support ticket. Calculate it by dividing your total monthly support costs (agent salaries or hourly wages, help desk software, phone system, any outsourced support costs) by the total number of tickets resolved. This metric connects your support operation to your business financials and helps you evaluate whether investments in tools, training, or automation are paying off.
Benchmarks: The average cost per resolution for ecommerce email support is $5 to $10. Live chat costs $3 to $8 per resolution because agents handle multiple conversations simultaneously. Phone support costs $8 to $15 per resolution because the agent is fully occupied for the duration of the call. AI chatbot resolutions cost under $1 per conversation. Self-service resolutions through your FAQ page or help center cost effectively $0 per resolution.
How to reduce CPR: Shift volume from expensive channels (phone) to lower-cost channels (chat, email, self-service) without degrading quality. Invest in self-service resources that resolve routine questions before they become tickets. Improve FCR so tickets are resolved in one interaction instead of three. Use automation to handle routine tasks like sending tracking information, processing simple returns, and answering common questions.
Additional Metrics Worth Tracking
Average resolution time measures the total time from ticket creation to final resolution. While FRT measures speed of the first response, average resolution time captures the entire lifecycle including back-and-forth communication, waiting for customer replies, internal investigation, and final closure. Ecommerce benchmarks are 4 to 12 hours for simple tickets and 24 to 72 hours for complex issues.
Agent utilization measures the percentage of available working time each agent spends actively handling tickets versus idle time. Target 70% to 80% utilization. Below 60% means you are overstaffed for your current volume. Above 85% means agents are overwhelmed, which leads to burnout, slower responses, and lower quality.
Ticket backlog tracks the number of open, unresolved tickets at any point in time. A growing backlog means your team is receiving more tickets than they can resolve, which will eventually cause response times to spiral. Monitor backlog daily and take action (reassigning tickets, pausing non-urgent projects, or bringing in temporary help) before the backlog reaches a point where response time targets are missed.
Building a Support Dashboard
Combine your core metrics into a single dashboard that you review weekly. Most help desks provide native dashboards showing FRT, FCR, CSAT, ticket volume, and agent performance. If your help desk dashboard is insufficient, export the data to a spreadsheet or a tool like Google Looker Studio to build a custom view. The dashboard should show current-week metrics alongside the 4-week trend so you can spot improving or declining performance immediately.
Share the dashboard with your support team. Agents who can see their own performance metrics alongside team averages are motivated to improve and can self-correct without constant management oversight. Celebrate improvements publicly and investigate declines collaboratively rather than punitively. The goal of metrics is continuous improvement, not performance policing.