Repeat contact rate is the percentage of customers who contact support again about the same underlying issue within a defined window. It's a closer proxy for "did the problem actually go away" than first contact resolution, because it counts a second contact only when it's the same problem — not just another message from the same person.
Pair this with first contact resolution for the fuller picture: FCR tells you how many contacts an issue took, repeat contact rate tells you whether the ones you closed actually stayed closed.
Repeat contact rate is only as good as your ability to match the underlying issue across conversations. We're exploring a layer in Supportman that connects related Intercom conversations automatically — tell us how your team spots repeat problems today.
Free trial · no card · about 2 minutes to set up
What is repeat contact rate?
Definitions converge on the same shape across support vendors: a customer contacts support again about the same issue within some defined window. What varies is the window and what counts as "the same issue" — and both need a written answer before the number means anything.
How to calculate it
Repeat contact rate (%) = (conversations that returned on the same issue within the window ÷ total conversations closed) × 100
| Week | Closed | Same-issue repeats | Repeat contact rate |
|---|---|---|---|
| Example | 200 | 18 | (18 ÷ 200) × 100 = 9% |
Report it next to FCR, not as a replacement for it. A team can have decent FCR and a rising repeat contact rate if closes are happening a little too eagerly.
The window problem: there is no standard
Support platforms vary widely on the repeat-contact window — some default to 24 hours, others use ranges as wide as 7–30 days. There's no universal methodology to inherit. Pick a window that matches your product: 24–72 hours suits a quick chat issue, 7–14 days suits anything that involves engineering, a refund, or a shipped fix. Write it down, keep it stable across quarters, and don't let it drift when a number looks bad.
Repeat contact vs. repeat issue
This is the distinction most repeat-contact reporting misses: a repeat contact from the same customer is not automatically a repeat issue. A customer who messages Monday about a billing question and again Friday about SSO has raised two unrelated problems — that's not a support failure, it's two normal contacts. A customer who messages Monday because CSV exports fail, and again Friday because CSV exports still fail, has a repeat issue, and Monday's close should be reopened as evidence, not treated as a separate ticket.
Matching on customer alone inflates the metric with false positives. Matching on customer and a consistent intent tag (or semantic match, once you have one) is what turns "repeat contact" into a real signal. See how to QA Intercom Fin for a related case where the same distinction — same customer vs. same problem — matters for scoring bot-handled conversations.
How to measure it in Intercom
- Pull closed conversations for the window, grouped by customer.
- Flag customers with more than one conversation inside the window.
- Filter to matching intent tags, or read the flagged pairs manually if tagging isn't reliable yet.
- Divide matched repeats by total closes for the period.
- Segment by issue, agent/team, and Fin vs. human — a bot that closes fast and reopens overnight will inflate the blended number until it's split out.
Intercom's reopened-conversation tracking covers the narrower case — the same thread coming back. It won't catch a repeat issue that arrives as a brand-new conversation, which is the more common shape once a customer has given up on the original thread.
What to do when it rises
- Rising on one intent. Usually a content or product gap, not an agent skill gap — check the help article and the macro before coaching anyone.
- Rising on Fin-handled threads specifically. Fin may be closing on an assumed resolution rather than a confirmed one. Segment Fin and human repeat rates separately.
- Rising after a policy or pricing change. Expected short-term noise — track whether it decays over 2–3 weeks as agents adjust their answers.
Low repeat contact rate isn't automatically good news either — it can mean customers with unresolved problems simply stopped contacting support at all. Read it next to CSAT and unrated closes, not on its own.
Frequently asked questions
What is repeat contact rate in customer support?
The percentage of customers who contact support again about the same underlying issue within a defined window. It differs from a raw repeat-contact count by requiring the second contact to be the same problem, not just another message.
What window should I use for repeat contact rate?
There's no industry standard — vendors range from 24 hours to 30 days. Match the window to how long a real fix for your product takes to confirm, and keep it stable once chosen.
What counts as a repeat ticket in customer support?
A new conversation from a customer who already contacted support about the same underlying issue inside your defined window. A second conversation about an unrelated topic isn't a repeat ticket.
How is repeat contact rate different from first contact resolution?
FCR counts whether an issue needed more than one contact. Repeat contact rate specifically checks whether a later contact from the same customer is the same problem resurfacing, which catches cases FCR's contact count alone can miss.
Does Intercom have a built-in repeat contact rate report?
No labeled metric. Intercom tracks reopened conversations natively; a full repeat contact rate — including repeats that arrive as a new conversation rather than a reopen — has to be built from closed-conversation exports and intent tags.