Quick summary

• First contact resolution (FCR) is the strongest predictor of customer satisfaction. Industry average is 70-75%. Top quartile teams hit 85%+.
• Average handle time benchmarks differ by channel: phone (6 minutes), email (4-8 hours to first reply), live chat (2-4 minutes active handling).
• CSAT scores above 90% are achievable but require specific practices. Global average across industries sits around 77-82%.
• Ticket backlog is the most ignored metric. A backlog growing faster than your team’s throughput is an early warning sign most managers miss.

---

Help desk software vendors love to highlight resolution time. It looks great on dashboards, it trends in the right direction after you implement new tools, and it’s easy to show to leadership.

The problem: fast resolution with a wrong answer is worse than a slow resolution with the right one. Customers who get the wrong fix the first time contact you again. They also tend to leave negative reviews.

Here are the metrics worth tracking in 2026, what good looks like for each, and why the popular ones are often misleading.

First contact resolution (FCR)

FCR measures the percentage of tickets resolved without a follow-up contact. Most teams define “resolved” as the customer not reopening the ticket within 7 days. Some use 30 days for more complex products.

The ICMI (International Customer Management Institute) tracks this across industries. Their data consistently shows FCR at 70-75% as the industry average, with top-performing teams (usually those with dedicated knowledge bases and well-trained agents) reaching 85-90%.

Every 1% improvement in FCR typically corresponds to a 1% improvement in CSAT, according to the same ICMI research. That relationship is more reliable than almost any other metric correlation in customer service.

What hurts FCR most: agents who don’t have access to the right information during a call or chat. Freshdesk’s own internal data (published in their 2024 benchmark report) showed that teams using a connected knowledge base resolved tickets on first contact 23% more often than those without one.

Tools that help: Freshdesk ($0-$69/agent/month), Zendesk ($19-$115/agent/month), and Help Scout ($22/agent/month) all support knowledge base integration directly in the agent view. The key is keeping articles updated, which requires a process, not just the software.

Average handle time (AHT)

AHT is the average total time spent on a ticket, from first contact through resolution (including after-call work for phone support).

Benchmarks by channel:

• Phone: 6 minutes is typical for general support; technical support runs 8-10 minutes
• Live chat: 2-4 minutes of active chat time (agents handling multiple simultaneous chats)
• Email: measured differently (first reply time and resolution time rather than handle time)

The metric is easy to game. Agents learn quickly that transferring a ticket rather than resolving it lowers their personal AHT. Teams with AHT as a primary metric often develop handoff cultures where no one wants to own hard tickets.

Use AHT to identify outliers, not to set targets for individual agents. An agent handling consistently long contacts is either taking on harder tickets (good) or struggling with a knowledge gap (addressable). You can’t tell which without looking at FCR alongside AHT.

CSAT and NPS

CSAT (customer satisfaction) asks customers to rate their support experience, usually on a 5-point scale. The industry standard is to count 4s and 5s as satisfied.

Global average across B2B software companies: 77-82%, according to Zendesk’s 2025 Customer Experience Trends Report. Their benchmark for “good” is 85%. Teams above 90% are genuinely strong performers.

NPS (Net Promoter Score) measures loyalty more broadly: “How likely are you to recommend us?” Detractors (0-6) get subtracted from promoters (9-10). Most support organizations use CSAT more than NPS because NPS captures the full product experience, not just support.

One honest caveat: response rates matter enormously. Teams with 5% survey response rates often have higher CSAT than teams with 40% response rates, because the people who bother to respond after a good experience are more likely to fill out the survey. Benchmark against your own historical data more than against industry averages.

Ticket backlog

Backlog is the total number of open tickets at any point in time, or more usefully, the trend over time.

An absolute backlog number tells you little. A team of 5 agents with 300 open tickets might be performing well if they’re resolving 150 per day. The same 300 tickets for a 20-person team is a problem.

What matters is backlog growth rate. If you’re adding more tickets than you’re resolving, the backlog grows. If ticket volume spikes (new product launch, outage) without a matching increase in capacity, the backlog can compound quickly.

Track backlog by age buckets:

• Under 24 hours (fresh tickets)
• 1-3 days (needs attention)
• 4-7 days (getting stale)
• Over 7 days (customers are likely frustrated)

Most help desk platforms can generate this breakdown automatically. Zendesk calls it “views by age,” Freshdesk calls it “ticket aging.” Help Scout shows it in the reports dashboard.

A practical target: less than 10% of your backlog should be older than 7 days. If that number is growing, the problem is usually either ticket volume exceeding capacity, or tickets getting stuck waiting on other teams (engineering escalations, billing, etc.).

The metric most teams track but shouldn’t prioritize alone

Time to first reply matters, but not as much as what happens after that first reply.

A 5-minute first reply that says “thanks for contacting us, we’ll look into this” followed by 3 days of silence is worse than a 2-hour first reply that actually addresses the issue. Customers can tell the difference between a canned acknowledgment and a substantive response.

Track first reply time, but treat it as a constraint rather than a goal. Set a threshold (under 4 hours for email, under 2 minutes for chat) and then optimize for FCR and CSAT within that constraint.

Frequently asked questions

What FCR rate should I target for a new help desk team? Start by measuring your current rate before setting a target. Most teams discover their baseline is 60-65%. A realistic first-year goal is to reach 75%. Getting above 80% typically requires investing in knowledge base quality and agent training, not just software.

How do I improve CSAT without hiring more agents? The highest-leverage changes are usually knowledge base updates (so agents can answer accurately faster), better ticket routing (right issue to the right agent), and post-resolution follow-up for complex tickets. Freshdesk, Zendesk, and Help Scout all have routing rules that can match ticket type to agent expertise.

Should I track AHT by individual agent? Use it for coaching, not compensation. Agents with high AHT who also have high FCR are handling harder tickets effectively. Agents with high AHT and low FCR need support. Penalizing AHT without context pushes agents to close tickets fast rather than right.

What’s a healthy ticket backlog growth rate? Zero or negative. You want to resolve tickets as fast or faster than they come in. If your backlog is growing consistently, you either need more capacity or need to reduce ticket volume through better self-service options.

How often should I review these metrics? Weekly reviews catch emerging problems before they compound. Monthly reviews are fine for CSAT trends. Daily backlog checks are useful during high-volume periods like product launches or incidents.