MTTA: Time From Page to Acknowledged

MTTA is response readiness. Shrinking it shrinks MTTR.

Definition

MTTA is time from page-sent to on-call acknowledgement, not from incident-start (which is MTTD plus MTTA). Acknowledge means “I am responding”, not “fixed”: the pager stops re-paging and the team knows someone is on it. Typically measured per page and aggregated per shift, per engineer, per service.

• Page-sent to ack. Not from incident-start; that’s MTTD plus MTTA; the two metrics are distinct.
• Ack means engaged. Not fixed; the pager stops re-paging; the team knows someone is on it.
• Aggregated views. Per page, per shift, per engineer, per service; different views surface different problems.
• Per-team baseline. The team’s own MTTA history is the comparison; absolute numbers vary by domain.

Reasonable targets

Targets scale with severity. Sev 1: under 5 minutes; Sev 2: under 15; Sev 3: under 1 hour during business hours. Aim for 95th-percentile within target rather than median because outliers are where incidents go wrong; per-service tier customer-critical services tighter, internal services looser.

• Sev 1: under 5 minutes. The headline target; the most urgent class.
• Sev 2: under 15 minutes. Less urgent but still time-bounded.
• Sev 3: under 1 hour business hours. Routine work; no off-hours requirement.
• 95th percentile within target. Median is too forgiving; outliers are where incidents go wrong.

How to consistently hit it

Three mechanisms drive consistent MTTA. Multi-channel paging covers phone DND, dead battery, no signal as real failure modes; backup on-call gets paged when primary doesn’t ack within 5 minutes; quarterly synthetic pages verify the chain end-to-end so surprises during real incidents are caught in advance.

• Multi-channel paging. Phone, app, SMS, voice escalation; phone DND, dead battery, no signal are real failure modes.
• Backup on-call. If primary doesn’t ack in 5 minutes, secondary gets paged; tested monthly.
• Quarterly synthetic pages. Verify the chain end-to-end; surprises during real incidents are expensive.
• Per-rotation chain test. The full ack chain tested per rotation; supports rotation-specific issues.

Reading MTTA data

The MTTA trend tells different stories. Trend up means paging tool degrading, rotation understaffed, or on-call fatigue, and deserves investigation; trend down is usually good but watch for ack-without-action and track MTTR separately; the 99th percentile is the worst-case signal often more useful than the median.

• Trend up. Paging tool degrading, rotation understaffed, on-call fatigue; investigate.
• Trend down. Usually good; watch for ack-without-action; track MTTR separately.
• 99th percentile signal. Worst-case response; often more useful than the median.
• Per-week trend dashboard. Trend visible to the team; supports continuous awareness.

When MTTA is bad

Bad MTTA has three causes in order. First: tool reliability, because lost pages mean lost MTTA, and paging tool dashboards show delivery rates. Second: rotation health (burnout, vacation gaps, off-hours coverage), which a survey plus staff-up addresses. Third: process, where routing fixes are usually high-leverage.

• Tool reliability first. Lost pages mean lost MTTA; paging tool dashboards show delivery rates.
• Rotation health second. Burnout, vacation gaps, off-hours coverage; survey, staff up.
• Process third. Right person paged, alert properly classified; routing fixes are usually high-leverage.
• Per-cause investigation playbook. Each cause has a specific check; supports fast diagnosis.