First contact resolution (FCR) is the percentage of customer issues closed in the first interaction, without a follow-up on the same problem. The formula is (issues resolved on first contact ÷ issues closed) × 100. Intercom does not label FCR as a native report; most Intercom teams proxy it with reopen rate, snooze patterns, and a documented window (often 24–72 hours).
FCR belongs next to CSAT, not instead of it. A closed conversation is an action; FCR tries to measure whether the problem actually stayed solved. Pair this page with customer service metrics for the rest of the weekly scorecard and what CSAT is for the satisfaction formula.
High FCR with low CSAT usually means customers gave up, not that the issue was solved. Supportman posts every CSAT rating to Slack as it arrives, so a low score next to a closed conversation is visible immediately instead of waiting for a monthly report.
What is first contact resolution?
FCR, sometimes called first call resolution from the phone-center era, asks: did the customer need to come back about the same thing? Email, chat, and Messenger all count. A new conversation about a different issue is not a miss.
You have to define three rules before the percentage means anything:
- What “resolved” means. Closed in Intercom is a teammate action, not proof the customer is done. See every Intercom metric explained for close vs reopen events.
- The follow-up window. 24 hours is tight for email; 72 hours is a common B2B default. Write it down and keep it stable.
- Same issue vs new issue. A billing follow-up after a technical close is a new contact, not an FCR miss, if you tag intents.
How to calculate FCR
FCR (%) = (conversations that did not reopen inside the window ÷ conversations marked resolved) × 100
| Week | Closed | Reopened on same intent | FCR |
|---|---|---|---|
| Example | 200 | 36 | (164 ÷ 200) × 100 = 82% |
That is a reopen-based proxy, which is what Intercom teams can actually operationalize. If you later add a “resolved on first reply” tag, keep both definitions labeled so week-over-week charts do not jump when the method changes.
There is no single industry FCR number worth copying onto an Intercom inbox. Channel mix, product complexity, and whether Fin handles tier-one questions all change the rate. Use your own baseline. Published call-center ranges (often quoted around 70–80%) describe a different channel and a different “contact.”
How to measure FCR in Intercom
- Export or report closed conversations for the week, grouped by close date.
- Count reopened conversations that return inside your window. Intercom tracks reopen events; it does not ship a built-in reopen rate.
- Exclude reopens tagged as a new intent when your tagging is reliable enough. If it is not, count every reopen and accept a conservative FCR.
- Split Fin-closed vs human-closed. A bot that closes quickly and reopens overnight will inflate blended FCR until you segment.
Put reopen count on the Friday review next to CSAT. Template: weekly customer support report.
When high FCR is a bad sign
- High FCR + low CSAT. Customers stopped writing. That is dropout, not resolution. Inspect remarks and unrated closes.
- High FCR + high contact rate. People can get an answer in one reply, which often means the help center should have answered it. Fix content, not agent speed.
- High FCR + rising snooze volume. Waiting-on-customer snoozes can hide unfinished work if you treat snooze as resolved.
Low FCR is not always an agent skill gap. Wrong routing, a macro that invites a follow-up question, a policy the agent cannot change, and a product bug all produce reopens. Tag the cause before coaching.
Six practices that raise FCR
1. Internal knowledge before more macros
Agents miss on first contact when the answer lives in Slack or someone's head. A searchable internal source for policy, edge cases, and “what we told customers last time” raises FCR more than another saved reply. Customer-facing help articles can lower FCR by deflecting the easy contacts; that is a win if contact rate also falls. Structure: how to structure a help center.
2. QA the reopen pile, not random tickets
Sample conversations that came back. Look for the second customer message: missing step, wrong assumption, or a question the first reply invited. Feed those patterns into the rubric. Start with customer service quality assurance and a QA rubric agents can use.
3. Answer the job, not only the literal question
A request to “hide my profile” is often a privacy job. A complaint about a gamified feature is often a stress job. First replies that treat those as feature tickets bounce. Train for the underlying job on the intents that reopen most.
4. Collect routing facts up front
If it takes two touches to learn plan, workspace, or which product the customer means, FCR is lost before the specialist replies. Ask for those fields in the Messenger, or use workflows that attach them before assignment.
5. Split Fin and human FCR
Fin can close a thread that the customer later sends to a human. Blended FCR then lies. Measure Fin reopen separately and QA Fin with the Fin QA checklist rather than treating every reopen as a person missing a step.
6. Close with the next step named
A first reply that says what happens next (and when) prevents “just checking in” reopens. If the work is blocked on the customer, say exactly what you need. If it is blocked on another team, say who owns it rather than closing as resolved.
Frequently asked questions
What is a good first contact resolution rate?
Use your own baseline on a stable definition. Call-center figures around 70–80% describe a different channel. Intercom teams should watch reopen rate and CSAT together rather than chasing a copied target.
Does Intercom have an FCR report?
No labeled FCR metric. Use closed conversations plus reopened conversations, then calculate reopen rate yourself. Keep Fin and human segments separate.
Is FCR the same as first call resolution?
Same idea, different channel history. First call resolution came from phone queues. First contact resolution covers chat, email, and in-app messaging as well.
Does a new question count against FCR?
Not if you can tell it is a new intent. If tagging is unreliable, count the reopen and accept a conservative number until tags are trustworthy.
Should agents be scored on FCR alone?
No. High FCR with low CSAT or high reopen after the window is a warning. Pair FCR with CSAT, QA, and cause tags so policy and product issues are not treated as agent misses.