First Response and Resolution Time: Set Your Own Targets

Definitions decide whether your numbers mean anything. Pin these down in writing:

• First response: the first human reply, not the automatic acknowledgment. Exclude auto-replies, or track them separately.
• Clock: business hours or calendar hours. Business hours match staffing, and calendar hours match what the customer feels. Choose one per metric and label it.
• Resolution time: from creation to the point the ticket is solved, and how you treat tickets reopened after that.
• Pauses: whether time waiting on the customer stops the clock.
• Channel: chat is measured in seconds or minutes, email in hours, so never blend them.

Publish the definitions next to the number. A first response time of 40 minutes means little until people know it excludes weekends and auto-replies.

A few very slow tickets can pull an average far from the typical experience. The median tells you what the typical customer sees, and a high percentile, such as the ninetieth, tells you how bad the slow tail is.

Track both. Say your median first response is 30 minutes but your ninetieth percentile is 9 hours: most customers are happy, but one in ten waits most of a day, and a target that covers only the median hides them. Set a target for each, for example a median goal and a ceiling for slow cases.

Report them by week, and add ticket volume alongside, because response times drift up when volume spikes or staffing drops. A time-series view lets you connect changes to causes like a product release or a holiday.

One target for everything satisfies no one. Split by the following:

1. Channel: live chat and phone need minutes, email can take hours, and social messages sit between them.
2. Priority: an outage or a blocked payment deserves a faster response than a how-to question.
3. Customer tier: if your plans promise faster support for higher tiers, measure each tier against its own promise.
4. Time of day: after-hours contacts follow a different rule, which should be explicit.

Write the resulting grid, such as priority by channel, as the support SLA policy and share it with customers where it's a commitment. Keep internal stretch goals separate from customer-facing promises, so you don't promise what only your best week delivers.

Use this method:

1. Pull twelve weeks of tickets and calculate the median and ninetieth percentile per channel and priority.
2. Look at the trend and the worst weeks, and identify causes such as understaffing or a release.
3. Set the target slightly better than your typical weeks, not your best.
4. Check staffing against volume: can the team hit the target at peak?
5. Review after a quarter, and tighten only when you're hitting it comfortably.

A worked example: say email tickets have a median first response of 4 hours and a ninetieth percentile of 20 hours. A reasonable next target might be a 3-hour median and a 12-hour ceiling, with a plan for the causes of the slow tail, such as weekend coverage. Don't set a target you'd need a new hire to reach without saying so.

Speed alone can mislead: a team can reply fast with a useless message. Pair it with quality measures such as customer satisfaction, reopen rate and the share resolved on first contact. If response gets faster but reopens rise, the replies are getting worse.

The main levers are staffing to volume, routing tickets to the right person, canned replies for repeat questions (see support macro templates), a good help center that prevents tickets, and automation for triage. See AI support agent resolution rates for how automation changes the picture. If you serve business customers in shared chat channels, B2B support in shared chat channels adds its own tracking needs. To choose a platform that reports these metrics, compare these help desk platforms and these B2B-focused options.