KPIs and metrics

First Call Resolution: How to Measure and Improve It

What first call resolution means, how to measure it honestly, how it differs from handle time, and practical ways to raise it through knowledge and handoffs.

Call editorial team · Published · 5 min read

Nothing frustrates a caller like ringing back about the same problem. First call resolution, or FCR, is the measure of how often that does not happen. It is popular because it connects what customers care about, getting the matter settled, with what operations can influence. It is also one of the easiest measures to get wrong. This guide explains how to define it, count it fairly and raise it in ways that last.

A support agent calmly resolving a customer's problem on the phone
Illustrative image

FCR belongs with other basic measures such as those in a call-center KPI list, repeat call rate and post-call satisfaction. Read it as one signal among them.

What counts as resolved

There is no single definition, so write yours down. The simplest is: the customer's issue was dealt with on the first contact and they did not need to contact you again about it within a set window, such as a few days. Two ingredients matter. One is the window: too short and you miss repeat calls, too long and unrelated calls creep in. The other is what counts as the same issue.

Three ways teams define it

  • Agent says soThe agent marks the call resolved. Easy to collect, but optimistic.
  • No repeat contactNo further contact on the same issue within a window. More objective but needs matching of contacts.
  • Customer says soThe caller confirms in a survey. Closest to the truth but depends on response.

Many teams combine the second and third. If you only use the agent's judgement, expect the figure to look better than customers' experience.

Measuring it honestly

A practical way to measure

  1. Define the issueChoose how to link repeat contacts: same number, same customer record or same topic.
  2. Set the windowPick how many days count as a repeat and keep it fixed.
  3. Mark outcomesRecord whether each call was resolved, transferred, callback promised or abandoned.
  4. Match repeatsCount contacts from the same customer on the same issue inside the window.
  5. Cross-checkCompare with survey answers and with listened calls.

Be clear about what you exclude. Some calls cannot be resolved at first contact by nature: an order that is still being delivered, a complaint that needs an investigation. Count them separately rather than letting them drag down the figure or hiding them. Remember that callers may contact you through other channels, so a repeat by email or chat is still a repeat if you can link it; omnichannel support discusses this.

Why calls are not resolved first time

Before fixing anything, find out what is going wrong. Review a sample of repeat calls and ask what stopped the first call from finishing the job. The reasons usually fall into a few groups.

CauseWhat it looks likeTypical fix
Missing knowledgeAgent does not know the answer or gives a wrong oneBetter knowledge base, coaching
Missing authorityAgent knows what to do but may not do itRaise limits, clear rules for exceptions
Handoff failureTransferred call loses context or lands wronglyWarm transfers, better routing
Process gapsFix needs another team, who is slowService agreements, tracked tickets
Poor understandingAgent solved the wrong problemListening and confirming habits
Product problemsThe issue is a defectFeed back to product teams

How to raise it

Make the right answer easy to find

If agents hunt for answers, calls drag and mistakes creep in. Keep a short, accurate knowledge base, organised by the questions callers ask, and update it when policies change. Building a knowledge base and finding gaps from calls help.

Let agents decide

Many second calls exist because the first agent had to escalate a simple request. Look at what escalations are about, and give front-line staff the authority to handle the common ones within clear limits. Set limits that match the risk.

Make handoffs clean

When a transfer is necessary, share context and stay with the caller until the next person is on the line. If nobody can answer, take a callback request with enough detail that the second person does not start from zero. Hold and transfer etiquette and escalation paths cover the mechanics.

Close the loop before ending the call

Beware the trade-offs

FCR can be gamed or can distort behaviour. Agents under pressure to resolve everything may stay on long calls, avoid transferring when they should, or mark calls resolved too soon. Handle time and FCR often pull in opposite directions, so look at both. Reducing handle time explains how to shorten calls without damaging quality.

Healthy and unhealthy use of FCR

Healthy useUnhealthy use
PurposeFind and fix reasons for repeat callsRank agents against each other
ReadingWith satisfaction and repeat contactsAs a single number
Agent responseHonest marking and escalating when neededMarking resolved to look good
ResultFewer repeat callsBetter figures, same experience

An example: one repeat call, traced

Take a hypothetical caller who rings a clinic about a changed appointment. The first agent changes the time in the diary but does not tell the front desk, who still has the old slot on a printed list. The caller turns up, is told there is no booking and rings again, annoyed. On paper the first call was resolved. In the customer's experience it was not.

A repeat-contact measure would catch this, and a review of the second call would show the cause: a change made in one place but not shared. The fix is not a faster agent but a rule that every change updates one shared record, plus a closing question: "Is there anything else this change affects?" Small process fixes like this raise FCR more reliably than urging people to work harder.

A weekly review routine

Weekly FCR review

  • Listen to five repeat calls and note why the first call failed
  • Count repeat reasons by cause
  • Pick one cause and one fix
  • Check that last week's fix helped
  • Share one good example with the team
  • Update the knowledge base with anything new

Do this for a month and the pattern of causes usually becomes obvious. Often two or three fixes account for most of the improvement.

Doing it in Call

In Call, call history keeps outcomes, summaries and transcripts, so reviewing repeat calls does not depend on memory. Tickets collect callers who asked for a person or a callback, with owners and notes, which helps you close the loop. AI call summaries can speed up the reading, and quality review lets you score a call item by item. Reports can be exported to a spreadsheet for your own analysis of repeat contacts. Use these to find the causes, then use the guides on knowledge, handoffs and follow-up to fix them.

Frequently asked questions

What is a good FCR figure?

There is no universal target. Definitions and call types differ too much for outside figures to be reliable, so compare against your own history.

How is FCR different from handle time?

Handle time measures how long a call lasts. FCR measures whether the problem was solved. A short call that leads to a second call is worse than a longer one that solves the issue.

Should I count callbacks as resolved?

Be consistent. Many teams treat a promised callback as unresolved until completed, and count the completed callback separately.

How long a window should I use for repeats?

Pick a window that fits your type of problem, such as a few days for routine issues, and keep it fixed so trends are comparable.

Can FCR be measured without a survey?

Yes, by matching repeat contacts from the same customer on the same issue. A survey is still useful to check that the numbers match customers' experience.

What is the quickest way to improve FCR?

Review repeat calls to find the top causes. Usually knowledge gaps, missing authority and poor handoffs account for most of them.

Review unresolved calls and callbacks in Call

In the demo console you can try incoming calls in a simulation.

Related guides