All free tools

First Contact Resolution (FCR) Calculator

Free first contact resolution calculator: gross and net FCR rate, survey margin of error, and blended FCR with an AI agent. Formula, examples, benchmarks.

How you measure resolution

Use one channel and one resolution rule at a time. Everything runs in your browser, and the inputs are saved in the page link so you can share the result.

Results

Updates live

First contact resolution rate

72.0%

720 of 1,000 contacts with no repeat contact in 7 days

Gross FCR
72.0%
All contacts in the denominator
Net FCR
80.0%
Over 900 resolvable contacts
Not resolved first time
280
28.0% of contacts

How to read it

280 customers needed another contact within 7 days. A longer window catches more of them and gives a lower FCR, so keep the window fixed from month to month.

Published benchmarks such as SQM Group’s come from customer surveys, so a repeat-contact FCR is not directly comparable with them. If an AI agent takes contacts first, switch to the AI vs human split.

How this is calculated

In "From totals" mode you choose how resolution is measured (no repeat contact within a window, or a customer survey) and enter contacts and those resolved on the first contact. It returns gross FCR, net FCR if you enter contacts that could not be resolved first time, and for survey data the 95% margin of error and the responses needed for ±5 points. In "AI vs human split" mode you enter contacts the AI took first, how many it resolved and handed off, how many handoffs a person resolved, and contacts sent straight to people. It returns a strict blended FCR that never counts a handoff, a customer-view FCR that counts resolved handoffs, the AI first-touch resolution rate, human FCR on handoffs and direct contacts, and compares them with your FCR before AI if you enter it. Everything runs in your browser.

What first contact resolution measures

First contact resolution (FCR) is the share of customer issues resolved on the first contact, with no transfer, callback or repeat contact about the same issue. 720 issues resolved first time out of 1,000 is an FCR rate of 72%. Enter your totals, or split contacts between an AI agent and people to see what a blended rate hides.

Teams measure it in one of two ways. A repeat-contact count treats an issue as resolved if the same customer does not come back about it within a set window. A customer survey asks whether the issue was resolved on that contact. The two can disagree, so state which one a number uses.

First contact resolution formula

  • Gross FCR = contacts resolved on the first contact ÷ all contacts
  • Net FCR = contacts resolved on the first contact ÷ (all contacts − contacts that could not be resolved on the first contact)
  • Blended FCR with an AI agent = (AI resolutions + first-time resolutions by people on direct contacts) ÷ all contacts

Net FCR removes contacts nobody could close in one go, such as a refund that needs approval or a visit by an engineer. It is useful for coaching, but the exclusions are a choice, so report gross FCR next to it. Call Centre Helper warns that counting a call passed to a second advisor as a first contact inflates the result, which is why the blended headline here does not count handoffs.

Worked example A: FCR from totals

A team handles 1,000 issues in a month. 720 have no repeat contact within 7 days. 100 needed a second step by design, such as a manager's approval.

StepCalculationResult
Gross FCR720 ÷ 1,00072.0%
Net FCR720 ÷ (1,000 − 100)80.0%
Not resolved first time1,000 − 720280 issues

If the same 72% came from 1,000 survey responses instead, the 95% margin of error is about ±2.8 points, so a month at 70% and a month at 74% may not be a real change. Switch the calculator to "Customer survey" to see the range for your own response count.

Worked example B: AI vs human split

Before AI, the team's FCR was 72%. After launch, an AI agent takes 600 of 1,000 contacts first and 400 go straight to people.

GroupContactsResolved first timeRate
AI agent600360 by the AI60.0%
Handed to a person180108 by the person60.0%
Left the AI unresolved600—
Straight to a person40026466.0%

Strict blended FCR is (360 + 264) ÷ 1,000 = 62.4%. Counting the 108 resolved handoffs gives 73.2%, a little above the 72% before AI. Both are true, but the second hides three things: FCR on contacts sent straight to people fell 6 points, a sign the AI kept the easier ones, 180 customers were passed to a second agent, and 60 left the AI with nothing resolved. Count those 60 as deflected and the AI's rate reads 70% instead of 60%.

Report the strict blend, the AI first-touch rate and human FCR side by side. For the AI's share of the whole queue, measured over the whole case rather than the first contact, see first-contact vs end-to-end resolution.

First contact resolution rate benchmarks

The most widely cited FCR benchmarks come from SQM Group, which surveys customers of inbound North American call centers within one business day of the call. As of November 2026, these are the figures its pages publish:

SegmentPublished FCRSource and basis
Industry average (2022)71%SQM Group; post-call survey
Range across centers (2022)40% to 91%SQM Group; post-call survey
Good70% to 79%SQM Group standard; 46% of centers
Top band (SQM: “world-class”)80% or higherSQM Group standard; 5% of centers
General inquiries (2022)74%SQM Group, by call type
Billing (2022)71%SQM Group, by call type
Technical (2022)63%SQM Group, by call type
Complaints (2022)47%SQM Group, by call type
Retail (2021)78%SQM Group, by industry
Insurance (2021)76%SQM Group, by industry
Financial (2021)71%SQM Group, by industry
Tech support (2021)65%SQM Group, by industry
Telco (2021)61%SQM Group, by industry

SQM's 2022 benchmark results state that "a good FCR rate standard is 70% to 79%", that "the call center Industry benchmark average for the FCR rate is 71%", and that "only 5% of call centers perform at the world-class FCR rate standard of 80%". The industry figures come from SQM's 2021 results, where "the retail industry has the highest performance for FCR at 78%" and "telco with the lowest FCR at 61%".

These are customer survey results from inbound call centers staffed by people. A repeat-contact FCR or an AI agent's resolution rate is measured differently, so use them as a reference point, and compare your own number with itself under the same rule each month.

Related RTCI tools and guides

Frequently asked

How do you calculate first contact resolution rate?

Divide the contacts resolved on the first contact by all contacts and multiply by 100. 720 resolved out of 1,000 is 72%. That is gross FCR. Net FCR leaves out contacts that could not be resolved first time by design: with 100 of those, it is 720 ÷ 900 = 80%.

What is a good first contact resolution rate?

SQM Group, which surveys customers within one business day of a call, rates 70% to 79% as “good” and 80% or higher as its top, “world-class” band, which only 5% of the call centers it benchmarks reach. Its 2022 industry average was 71%. Those are customer survey results from inbound call centers staffed by people, so compare a repeat-contact FCR or an AI agent's rate with your own trend rather than with them.

What is the difference between gross and net FCR?

Gross FCR uses every contact as the denominator. Net FCR removes contacts nobody could resolve in one go, such as a refund needing approval, so it is higher. Net FCR helps with coaching, but the exclusions are a judgement call, so report gross FCR next to it.

How do you calculate FCR when an AI agent handles contacts first?

Report three numbers: a strict blended FCR (AI resolutions plus first-time resolutions on contacts sent straight to people, over all contacts), the AI first-touch resolution rate, and human FCR. In the example on this page the strict blend is 62.4%, but counting resolved handoffs as first contact resolution lifts it to 73.2%, above the 72% before AI, while human FCR on direct contacts fell 6 points.

Should I measure FCR with a survey or with repeat contacts?

Either works if you keep it fixed. A repeat-contact count is objective but misses customers who come back on another channel, and a longer window gives a lower FCR. A survey reflects what customers say but only covers people who answer, so check the margin of error before reading a month-to-month change.