<span id="hs_cos_wrapper_name" class="hs_cos_wrapper hs_cos_wrapper_meta_field hs_cos_wrapper_type_text" style="" data-hs-cos-general-type="meta_field" data-hs-cos-type="text" >First Call Resolution Rate: What's Good and How to Improve It</span>

First Call Resolution Rate: What's Good and How to Improve It

Q: What is a good first call resolution rate?

According to SQM Group's benchmark , 70% to 79% counts as good and 80% or higher is world-class, a level only 5% of call centers reach.

Q: Why is FCR lower for technical and complaint calls?

Those calls are inherently more complex. In SQM's data , technical calls resolve first time 63% of the time and complaints only 47%, versus 74% for general inquiries.

05 de November, 2026

Last updated: November 04, 2026

Every customer who calls back about the same problem is a small failure that shows up three ways: a frustrated buyer, a second paid conversation, and a slot in your queue that someone else could have used. First call resolution rate, or FCR, is the metric that captures all three. This guide covers what a good FCR rate looks like, how to measure it honestly, why it stalls, and how AI can lift it without adding headcount.

What first call resolution actually measures

FCR describes the share of customer issues solved on the first contact, so the customer never needs a second call. TechTarget's definition frames it as agents properly addressing a customer's need the first time, and gives the simple formula: cases resolved on the first try divided by total cases, times 100.

The arithmetic is easy. The definition is where teams disagree. Does a follow-up chat message count as a repeat? What about a transfer, an escalation, or a customer who abandons the call? The same TechTarget article flags exactly these open questions, and your answers will move the number by double digits. Before you set a target, write the definition down and keep it stable, because a metric that quietly changes meaning cannot show a trend.

It also helps to separate FCR from its cousins. Containment measures whether automation handled a conversation without a human; our guide to containment rate covers that. Customer satisfaction measures how people felt, which you can explore in our CSAT guide. FCR sits between them: did the problem actually go away?

What is a good first call resolution rate?

The most-cited reference is SQM Group's benchmark of North American inbound call centers. Its 2022 document puts the industry average at 71%, defines 70% to 79% as good, and reserves world-class for 80% or higher. Only 5% of centers get there, while 49% fall below 70%.

Key takeaway: If your FCR sits near 70%, you are average, not broken. The real question is which call types drag you down, because the average hides a wide spread.

FCR varies sharply by call type

A single blended number is misleading. The same SQM data breaks FCR out by call type:

Call type	Average FCR (SQM)
General inquiries	74%
Account maintenance	73%
Orders	72%
Billing	71%
Technical	63%
Claims	59%
Complaints	47%

The lesson: benchmark each queue against its own type, not against a headline average. A support team handling mostly technical issues that scores 65% is outperforming the table, while a billing desk at 65% is lagging.

Longer calls resolve less often

SQM also found that FCR falls as calls get longer: 76% for calls of one to three minutes, 71% for five to ten, and 62% at fifteen. Cutting handle time for its own sake is risky, but our piece on reducing AHT without hurting CSAT shows how to shorten calls by removing friction rather than rushing customers.

Why FCR moves cost and satisfaction together

Most service metrics trade off against each other. FCR is unusual because improving it tends to help the customer and the budget at once. SQM's benchmark document states that each 1% improvement in FCR corresponds to a 1% improvement in customer satisfaction and a 1% reduction in operating costs. It also reports that when a customer's issue is resolved on the first call, 95% of customers continue doing business with the organization.

Think about what a repeat contact costs. You pay for the second conversation, the customer's patience shrinks, and the agent now starts from a warmer, angrier place. Multiply that by a queue and the economics are clear. If you track cost per resolution, FCR is the lever that bends it, since a ticket solved once is always cheaper than one solved twice.

How to measure FCR without fooling yourself

There are two broad approaches, and they disagree more than you might expect.

Survey-based measurement

Ask the customer right after the interaction whether the issue was resolved. SQM's standard method relies on an external post-call survey completed within one business day. This captures the customer's own view, which is the one that drives loyalty.

Internal repeat-contact tracking

Count a call as resolved if the same customer does not contact you about the same issue within a fixed window. It is cheap and automatic, but it can flatter you: SQM notes that internal measurement tends to overstate FCR by 10 to 20% compared with post-call surveys, because customers who give up or switch channels vanish from your repeat data.

The practical answer is to use both. Track repeat contacts continuously for trend and speed, then calibrate against a short survey every month. If the two numbers drift apart, trust the survey and investigate.

Why first calls fail to resolve

SQM's analysis of non-resolved contacts assigns the source of failure to three places: the organization 49% of the time, the agent 38%, and the customer 13%. That ordering matters. Nearly half of the failures come from policies, systems and processes, not from people.

In practice the root causes cluster into a few patterns:

Missing authority. The agent understands the problem but cannot approve the refund, change, or exception, so the customer must call again or wait for a callback.
Scattered information. Order status lives in one tool, contract terms in another, and past conversations in a third. The agent spends the call hunting rather than solving.
Bad routing. The call lands with someone who lacks the skill, then gets transferred. Every handoff is another chance for the context to be lost.
No closing step. The agent answers the question but never confirms the customer is actually done.

How to improve first call resolution

1. Fix the organizational blockers first

Since the organization is the biggest source of failure, start by reviewing your top repeat-contact reasons and asking what policy or system forced the second call. Give front-line agents authority for the five or six most common exceptions and watch the repeat rate for that call type.

2. Put the answer in front of the agent

A searchable, current knowledge base shortens the hunt. Our guide to an AI-powered knowledge base explains how to keep articles accurate, and real-time agent assist surfaces the right article while the customer is still talking.

3. Route by skill and context, not by queue order

Send technical issues to people who solve technical issues. When a handoff is unavoidable, pass the full context along so the customer never repeats themselves. Our walkthrough of AI-to-human handoff covers what a good transfer packet contains.

4. Let AI resolve the simple calls completely

Simple, high-volume requests such as order status, appointment changes and account questions are ideal for automation, because they follow clear rules and need no judgment. This is where Darwin AI's Eva fits: she handles customer conversations end to end across voice and messaging, pulls the customer's context from your systems, and hands off to a human with the full history only when a case truly needs one. Resolving the easy contacts at the first touch frees your people for the complex ones, which is where first-call resolution is hardest to win.

Example: A billing queue receives many questions about invoice copies. Automating the lookup and resend means those customers are resolved in the first interaction, and the remaining billing calls are the genuinely complicated ones your agents have time to handle properly.

5. Close every contact with a confirmation

Train agents and automations to ask one closing question: "Is there anything still unresolved?" It costs seconds, and it turns hidden repeat calls into visible ones. Pair it with the customer effort score to see where customers still work too hard.

6. Review quality with data, not anecdotes

Sample calls where the customer called back and read them for the pattern. Modern AI quality assurance can scan every conversation instead of a small sample, so repeat-call causes surface in days rather than quarters.

A simple 30-day plan

In week one, fix your definition and baseline FCR by call type. In week two, list the top ten repeat-contact reasons. In week three, remove the two biggest organizational blockers and automate the simplest request. In week four, re-measure and compare against your baseline. Small, well-chosen changes beat large programs, and each point of FCR you recover compounds across every future call.

FAQ

What is a good first call resolution rate?

According to SQM Group's benchmark, 70% to 79% counts as good and 80% or higher is world-class, a level only 5% of call centers reach.

How do you calculate first call resolution?

Divide the cases resolved on the first contact by the total number of cases and multiply by 100, as TechTarget explains. The harder part is deciding who confirms the resolution: the customer, the agent, or a no-callback window.

Does improving FCR really lower costs?

SQM reports that each 1% improvement in FCR brings roughly a 1% reduction in operating costs and a 1% lift in customer satisfaction, per its benchmark document.

Why is FCR lower for technical and complaint calls?

Those calls are inherently more complex. In SQM's data, technical calls resolve first time 63% of the time and complaints only 47%, versus 74% for general inquiries.