Post-Dial Delay (PDD)
Understand Post-Dial Delay — the industry-standard metric for measuring the time from INVITE to first ringback, thresholds, and why PDD is the most important call setup KPI.
Post-Dial Delay (PDD) measures the time from sending a SIP INVITE to receiving the first indication of call progress — typically the 180 Ringing response. It represents the moment the caller transitions from waiting in silence to hearing a ringback tone. PDD is the single most user-visible call setup metric and a standard industry KPI.
Think of PDD as the caller's patience timer. From the moment they press "call," the clock starts ticking. Until they hear ringing, they do not know if the call is going through. Long PDD causes callers to hang up and retry — or worse, switch to a competitor.
One-shot metric
PDD is captured once per call. It appears in the Call Timing Overview cards on the endpoint detail page and in aggregate statistics (P50, P95, P99) on the test run summary page.
How It Works
CallMeter records the timestamp of the outgoing SIP INVITE and the timestamp of the first meaningful provisional response (typically 180 Ringing or 183 Session Progress with SDP):
PDD = first_ringback_indication - invite_sent_at
PDD encompasses the entire signaling chain: local proxy processing, carrier routing, international transit, and remote endpoint alerting. It is the sum of all delays in the call setup path.
Why It Matters
PDD is an industry-standard telecom KPI because it directly measures caller experience:
- Customer satisfaction — Studies show callers begin to perceive delay after 3 seconds and may abandon the call after 8 seconds
- Revenue impact — In contact center scenarios, every second of PDD reduces the chance the caller stays on the line
- Carrier SLAs — SIP trunk providers are often measured on PDD as part of service level agreements
- Regulatory requirements — Some telecom regulators mandate maximum PDD for certain call types
For enterprise VoIP testing, PDD is the headline metric for call setup quality. CallMeter aggregates PDD across hundreds or thousands of test calls to produce percentile distributions (P50, P95, P99) that reveal both typical and worst-case setup times.
Thresholds
| Level | PDD | Interpretation |
|---|---|---|
| Good | 3000ms or less | Normal call setup, caller experiences minimal wait |
| Warning | 5000ms | Noticeable delay, investigate routing chain |
| Critical | Above 8000ms | Excessive delay, callers likely to abandon |
P95 matters more than average
Average PDD can mask problems. If P50 is 2 seconds but P95 is 8 seconds, one in twenty calls has unacceptable delay. Always check percentile distributions, not just averages.
Common Causes of High PDD
| Cause | Explanation |
|---|---|
| Carrier routing delays | SIP trunk provider routing calls through congested or distant paths |
| International transit | Calls crossing multiple carriers and international gateways |
| Number portability lookups | Database queries to determine the correct carrier for ported numbers |
| Complex IVR processing | Destination PBX running IVR logic before generating ringing |
| DNS resolution chains | Multiple DNS lookups along the signaling path |
| Authentication round-trips | SIP digest authentication adding extra message exchanges |
How to Fix It
- Benchmark across carriers — Test the same destinations through different SIP trunk providers. PDD varies significantly by carrier and destination.
- Optimize routing — Minimize the number of SIP hops between caller and callee. Each hop adds processing time.
- Use direct routes — Where possible, establish direct peering with destination carriers to avoid transit hops.
- Pre-authenticate sessions — If your SIP infrastructure requires authentication, use registration-based auth rather than per-call challenge/response.
- Monitor P95/P99 — Focus on tail latency, not averages. Improve the worst cases first, as they have the highest abandonment risk.
Related Metrics
- Time to Trying — First-hop delay; a fast Trying with slow PDD means the delay is downstream
- Time to Ringing — Similar to PDD but specifically measures 180 Ringing arrival
- Call Setup Time — Total time including answer; PDD is the pre-answer portion
- Call Result — High PDD may correlate with timeout or cancelled call results
RFC Reference
Post-Dial Delay is measured using SIP provisional responses defined in RFC 3261 (SIP: Session Initiation Protocol), Section 21.1. The 180 Ringing (Section 21.1.1) and 183 Session Progress (Section 21.1.2) responses mark the end of the PDD measurement window.
Time to Ringing
Understand the Time to Ringing metric — how CallMeter measures the delay from SIP INVITE to 180 Ringing, and what it reveals about call routing and remote endpoint responsiveness.
Call Setup Time
Understand Call Setup Time — how CallMeter measures the total SIP signaling duration from INVITE to 200 OK, thresholds, and what affects the complete call establishment.