Depending on your requirements, you can choose to calculate time-based metrics (e.g. first response time, subsequent response, time to close) in one of two ways:

In Intercom's new report templates and custom reports, you can control whether office hours are included in metric calculations at the individual chart level.

A "Within office hours" option is available for the following metrics:

Conversation metrics

Teammate metrics

Team metrics

By default, any pre-built template will calculate metrics based on All hours. You can choose to calculate metrics Within office hours when adding a metric from the chart library using the dropdown.

If you have access to custom reports, you can create or edit an existing chart and select the checkbox below the metric to calculate it Within office hours, then click Update chart.

This means you can have some metrics in your report which are calculated within office hours, and others that include all hours if you want to compare them.

You can assign individual schedules to teammates so that time-based metrics reflect each person's actual working hours, rather than your workspace default.

When "Within office hours" is selected for a teammate metric, the first matching schedule is applied in this order:

The “Effective schedule” column in Settings > Office hours > Teammate schedules shows which schedule will apply to each teammate based on their current assignments.

You can assign schedules to individual teammates or in bulk from Settings > Office hours > Teammate schedules in the "Reporting schedule" column. This is an assignment surface so you reuse the office-hours schedules you've already set up.

Assignment is optional - any teammate without an explicit schedule automatically falls back to their team's schedule, then the workspace default, so you can start with just the teammates whose reporting needs correcting.

In Intercom's legacy pre-built reports, the inclusion of office hours in metrics is controlled by your default office hours which are configured in Settings > Workspace > Office hours.

This toggle allows you to "Track team response metrics within office hours only".

Here are some practical examples to illustrate how office hours affect various metrics.

Response-time metrics use the schedule that was in effect at the time each reply was sent. Changing your office hours only affects data going forward — past conversations keep their original calculated values. After switching to a narrower schedule (for example, removing weekends), conversations that previously had data outside those hours will show no value for the within-office-hours metric, so a period that used to show data may appear emptier.

Within-office-hours metrics only count time inside your configured hours, so they're almost always smaller than the raw elapsed time. Two settings commonly narrow the hours more than expected: a 'Weekdays' schedule overrides any per-day configuration, and your default (workspace) schedule may cover fewer hours than you assume. Team inboxes use their own custom hours only if configured — otherwise the workspace default applies. To check the exact hours in use, go to Settings > Office hours.

Neither indicates lost data. A response sent entirely outside office hours has no countable in-hours duration, so it shows as blank in CSV exports and is excluded from averages. If the in-hours duration works out to zero (for example, the reply was sent exactly at the boundary of office hours), it can show as 0s in a chart. Both are expected behavior.

First response time measures from the customer's first message to the first human reply — not from when a bot or Fin escalated. Any time the conversation waited before a human replied (within office hours) is included. The 'bot inbox excluded' variant removes only the time the conversation spent in the Bot Inbox (where conversations wait while an automated bot or Fin AI handles them before a human teammate takes over). A duration of '1 day' in reports means 24 hours of in-hours time, not a calendar day.

The two metrics are calculated independently and can occasionally produce the same value — for example, a conversation that was answered and immediately closed in a single action within office hours. An equal value is a coincidence, not a sign the metrics are linked or mis-configured.