First Contact Resolution: The Benchmark Everyone Quotes Measures Four Different Things

Every page on this query gives the same formula and the same benchmark. The formula is right. The benchmark is only meaningful if you know how the number was produced, and almost nobody states it.

Internal measurement overstates FCR by 10 to 20% against post-call survey. The window for what counts as a repeat runs anywhere from one day to thirty, with no industry standard. Those two facts make most FCR comparisons meaningless.

The formula, briefly

Contacts resolved on first contact, divided by total contacts, times one hundred. That part is not in dispute and it is not where the difficulty is.

Everything hard about FCR is in the word resolved and the word first.

The four ways it gets measured

Published research puts internal methods 10 to 20% above external ones for the same operation. A team at 78% measured by agent disposition and a team at 65% measured by survey may be running identical service — and the first will win a board comparison it has no claim to.

The reopened-ticket method deserves a specific warning. Most customers do not reopen; they start a new conversation, often through a different channel. Unless your matching is genuinely good, that method counts your worst outcomes as successes twice over.

The window nobody agrees on

Internal FCR is judged on no repeat contact within a timeframe, and published guidance puts that timeframe anywhere from 1 to 30 days with no industry standard for what is appropriate.

This is a bigger lever than it sounds. A 24-hour window counts almost everything as resolved, because most people do not come back the same day even when the fix failed. A 30-day window catches genuine failures and also catches unrelated contacts, which drags the number down for reasons that are not your fault.

Seven days is the defensible middle, and the important thing is not which you choose but that you state it. FCR reported without a window and a method is a number without units.

The benchmarks, with the caveats attached

If your figure is internal, subtract before comparing. A 75% internal FCR against a 70% survey-based benchmark is not a pass — adjusted, it is roughly 60%.

What each point is worth

The relationships published in this area are unusually clean, which is why FCR survives as a metric despite the measurement mess.

• Customer satisfaction falls around 15% (top-box) with each additional callback. One repeat contact does most of the damage.
• Every 1% of FCR improvement corresponds to roughly 1% of customer satisfaction improvement.
• Every 1% of FCR improvement raises transactional NPS by around 1.4 points.
• 60% of companies measuring FCR for a year or more report improvement of between 1% and 30%.

The first line is the one to act on. The gap between resolving something on contact one and resolving it on contact two is worth more than the gap between contact two and contact four. If you are triaging where to spend effort, the first repeat is the expensive one.

How it gets gamed

FCR on a wall reliably produces three behaviours, all rational and all bad.

• Closing conversations before they are resolved, then handling the return as a new contact with its own fresh FCR.
• Marking resolved on agent disposition when the customer never confirmed anything.
• Discouraging the customer from coming back — which improves FCR by exactly the mechanism that produces churn.

The counterweight is re-open rate, and the two should never be reported apart. FCR rising while re-opens rise is not improvement, it is a queue being cleared into next week.

What FCR does to your documentation backlog

Under-used framing: FCR is a documentation metric wearing an agent's badge. A contact that cannot be resolved first time is usually one where the agent had to ask someone, and the thing they had to ask is a gap in what is written down.

Tracking why each non-FCR contact failed, as free text, produces a ranked list of internal documentation gaps within about a month. That list is worth more than the percentage it came from.

FCR with AI in the queue

Two things change and both need handling before the number is readable.

First, split by handler. The AI takes the easy contacts first, so blended FCR rises for reasons that have nothing to do with performance while the human-handled figure falls — the remaining tickets are the hard ones. A team that does not split this reads a genuine improvement as an agent problem.

Second, the escalated contact. A conversation the AI attempted and handed to a human is not a first contact resolution by any honest reading, and most platforms will not count it as a repeat because it is one ticket. If you are measuring FCR while running automation, decide explicitly how escalations count, and write it down before the first quarterly review rather than during it.

Benchmarks and the 10-20% measurement gap are from published FCR research and cited below. The window guidance and the reading of FCR as a documentation metric are ours.

WRITTEN BY AR · UPDATED 2026-08-04

I read the fine print. Vendor pricing pages, billing definitions, terms, funding filings and acquisition notices — then I do the arithmetic nobody publishes: what a platform actually costs at your volume, what its headline metric is really counting, and who owns it now. I do not run benchmarks, and no page here pretends otherwise.

How we work · About the author