Average Touches per Ticket
Average Touches per Ticket (also called Average Reply Count or Average Interactions) measures how many agent replies a support ticket requires before it is resolved. It is a proxy for resolution efficiency — fewer touches generally mean clearer, more accurate responses. High average touches can signal unclear initial responses, knowledge gaps, or complex issues requiring back-and-forth. It also helps identify ticket types where knowledge base articles or macros would have the highest impact.
Average Touches = Total agent replies sent across all tickets ÷ Number of tickets resolved in period
Count only agent-initiated replies, not customer replies. A ticket requiring one agent reply to resolve = 1 touch. Most helpdesks report this natively. Exclude ticket types where extended back-and-forth is expected (technical investigations, onboarding flows) when comparing against standard targets.
B2B SaaS, async support
Calculate average touches per ticket
- 1Conflating touches with resolution quality — a 1-touch ticket is only good if the issue was actually resolved. Always pair with FCR and reopen rate.
- 2Not separating simple ticket types from complex ones — password resets at 3.5 average touches indicates a training problem; enterprise API integrations at 3.5 is excellent.
- 3Using average touches as an individual agent metric without controlling for ticket type — some agents handle the complex escalations by design.
- 4Ignoring the trend, not just the number — rising average touches over a release cycle often signals a product documentation gap.