How We Measure Email Response

All response times are measured in business hours only — time outside working hours is never counted against anyone.

If an email arrives at 4:30 PM on a Friday, only 30 minutes of that day count before the clock pauses. It resumes at 9:00 AM on Monday. Similarly, emails received on Saturday or Sunday start counting on Monday morning.

Response time is the number of business hours between when an email was received and when a team member sent their reply.

• The clock starts when the inbound email arrives in the member's inbox.
• The clock stops when the member sends a reply in the same email thread.
• If the sender sends a follow-up email before receiving a reply, the clock resets to the follow-up — since that is the message the sender is actually waiting on.
• Each reply in a thread is tracked as a separate response time record. Long threads with multiple back-and-forth exchanges produce multiple records.

Every email is classified into one of three categories based on the sender's email domain. These categories are not mutually exclusive — an owner client email is also an external email.

When a team member has an active Gmail vacation auto-reply, the system automatically detects this and applies a grace period after they return. Emails received during the OOO period get extra time before the response clock starts.

OOO periods are detected automatically by scanning each member's Gmail sent folder for vacation auto-reply messages. A minimum of two auto-reply messages must be found within a window to register as a valid OOO period — this prevents a single accidental auto-reply from creating a false absence. Detected periods are stored persistently and applied to all historical response time records, so corrected grace periods take effect retroactively.

An SLA breach is recorded when a response to an owner client email takes longer than the defined threshold — this applies to both first responses and follow-up replies.

Rules for SLA breaches:

• Only owner client emails count — vendor, prospect, and other external emails are tracked for response time but do not contribute to breach counts.
• Internal emails are excluded — team-to-team emails never count toward breaches.
• OOO grace periods are applied first — if the grace period extended the clock start, that adjusted time is used when determining if 8 hours were exceeded.
• The breach count and percentage shown on the dashboard refer to the selected date range only.
• The SLA Breaches table (below the charts) lists every breached email, sorted by longest response time first, so the worst cases are always at the top.

The system tracks response times by analyzing email threads — the full conversation chain — rather than individual messages. Here's how thread data is collected:

• The system looks at the Sent folder to find threads where a team member has replied. This avoids counting newsletters and automated emails that never received a response.
• Within each thread, the system walks through every message in order and records a response time for each incoming → outgoing pair. Long conversations produce multiple records.
• If a sender sends multiple messages before receiving a reply, the clock starts from their most recent message — that's the one they're waiting on.
• Proactive outbound emails (not replies to anything) are skipped.

Emails from automated systems are excluded from response time tracking. These do not represent a real person waiting for a reply.

An email is considered automated if the sender address:

• Starts with no-reply@, noreply@, donotreply@, do-not-reply@, mailer-daemon@, postmaster@, notifications@, bounce@, or similar prefixes.
• Contains keywords like automated, calendar-notification, or system@.

Automated emails are also excluded from the Top Senders list in the activity section, since they skew the data.

These metrics count raw email activity during the selected period:

Both the average and median are shown for each response time metric because they tell different stories: