First contact resolution (FCR) and its four formulas

The call centre name is first call resolution. Helpdesks, the tools support teams answer email from, sometimes call the ticket version one-touch resolution. At SQM Group’s current call centre average of 71%, 29% of customers have to get in touch again about the same problem.

Four ways to calculate FCR on the same 1,000 calls

The versions differ in what they divide by. ContactBabel’s guide to first-contact resolution, from an analyst firm that studies contact centres, runs one set of calls through four of them.

A centre takes 1,000 calls. Of those, 900 are about a new problem and the other 100 are call-backs about one not yet solved. In all, 950 could have been solved on the first call, and 800 were solved during it.

The first is gross FCR, and the second is net FCR. It leaves out the calls nobody could have solved on the first try, such as a repair that needs someone to visit. Those left-out calls are sometimes called carve-outs. MetricNet’s column on FCR, from a firm that benchmarks IT service desks, says most organisations do not track the gross figure at all.

What counts as the first contact

The two names mostly mean the same thing. SQM Group’s FCR guide says they are “commonly used interchangeably”, with call for phone and phone menus, and contact for email, chat and other channels.

Some draw the line at transfers instead. In ContactBabel’s guide, first call resolution allows the call to pass between agents, and first contact resolution means the agent who answered solved it. A transfer costs something either way, since SQM’s study of transfers and holds found 19% of callers were transferred and their FCR was 14% lower.

The first contact your team sees is often the customer’s second try. Atlassian’s guide to first call resolution says this happens when a customer tries self-service before calling.

Who decides the problem was solved first time

Asking the customer is the most direct way. SQM Group’s survey counts a call as solved first time only if the customer gives these three answers.

• Yes, to “Was your call resolved?”
• One call, to “How many calls did you make to resolve your call?”
• No, to “Did you use another contact channel after you used the call center?”

Measures that never ask the customer run higher. SQM’s comparison of the two methods puts them 10 to 20% above what customers report, partly because a customer who gives up never calls back.

Helpdesks count replies instead. In Zendesk’s one-touch report, a ticket counts if it was solved with one reply from your team, or none. The customer must send nothing after their first message in Freshdesk’s FCR report, and the ticket must stay closed. Freshdesk divides by resolved and closed tickets, so a ticket still open never lowers its figure.

For email, MetricNet’s FCR article says the standard now emerging is to count a ticket resolved within one business hour of arriving.

Call centres often watch for a repeat contact instead. If the same customer gets in touch about the same problem within a set window, the first contact failed. The window is seven days in a Genesys report, and anything from one to 30 days at SQM. SQM gives a claim that needs investigating 30 days and a status question five.

The least reliable way is to let the person who took the call mark it resolved. NiCE’s glossary says agents have a reason to mark calls resolved to hit their targets, even when the problem may still be there.

FCR rates measured in more than 500 call centres

SQM Group sells quality assurance software to call centres and runs the FCR benchmark many other pages quote. Good, by its bands, is 70 to 79%. World class is 80% or more, a level 5% of call centres reach. Its current average is 71%, from post-call surveys and its own AI scoring of call recordings. We read every figure below on 19 September 2026.

In SQM’s 2024 results, 73% of general inquiries were solved first time, against 60% of technical support calls and 48% of complaints. By industry, retail comes top at 77% in SQM’s 2026 chart, and tech support scores 64%.

Many vendor pages, Talkdesk’s guide to FCR among them, call 70 to 75% a good rate, usually with no source. The range is in SQM’s 2021 guide. In an ICMI survey of 298 contact centre professionals in 2008, just over one in three reported 90 to 100%, figures ICMI called “suspiciously high”. Comparing your FCR with your own past months works better, says ContactBabel, since other centres measure it differently.

When a high FCR means customers ask easy questions

Customers who keep calling about something simple, like the date their subscription renews, push FCR up. Those calls are easy to solve first time. An answer in a help centre or a chatbot would remove many of them. Zendesk’s FCR article says a high FCR rate “could indicate that your self-service resources are lacking”.

ContactBabel found the UK trend falling slightly as the easier questions moved to self-service and left call centres with harder ones. A falling rate can be a sign that your improvements are working, according to ICMI’s 2008 report.

Fixing the answers behind repeat contacts

The 100 call-backs in the example are where to look first. Each one tells you what a first answer missed. Sort them by question and fix the answers that come up most.

Check the targets you set beside it. Atlassian’s guide describes an 85% FCR target next to a rule that every issue is resolved or escalated within five minutes. Agents then have to choose between the two. Track the reopen rate too, because a ticket closed early and reopened is a failed first contact.

Paying agents on FCR pushes them to mark calls resolved early. Call Centre Helper’s article on calculating FCR says it links an agent’s pay to failures the agent cannot fix.

Deacon, our chatbot, asks for the missing detail when one would settle the answer. It asks up to three clarifying questions in a conversation, then hands the conversation to you with what the visitor has said.

FCR next to the resolution rate

The resolution rate counts every problem that was eventually solved, however many contacts it took. FCR is the part solved in the first. A centre can score high on FCR and low on resolution rate, according to NiCE’s glossary. It happens when first-contact fixes do not hold and the customer calls back.

First response time measures something else, how long the customer waited for a first reply.