Queue-wait distribution
2,159 of 3,679 inbound calls had queue time| Wait | Calls | ShareShare of calls with any queue time |
|---|---|---|
| 0–10s | 1,780 | 82.4% |
| 10–30s | 308 | 14.3% |
| 30–60s | 51 | 2.4% |
| 1–5m | 12 | 0.6% |
| 5m+ | 8 | 0.4% |
| Total | 2,159 | 100.0% |
The 0–10s bucket sits below the rulebook §E queue gate: those calls never enter the Answer-% denominator, and most of that bucket is fast service, not hang-ups — measured 967 of 1,343 with a real disposition on the parity window (FACTS.md L23). Percentiles above are computed on the ≥10s population only, so they describe members who actually waited.
Abandon waits — how long members held on before hanging up
434 abandoned calls in the window| Wait | Calls | ShareShare of abandoned calls |
|---|---|---|
| 0–10s | 384 | 88.5% |
| 10–30s | 36 | 8.3% |
| 30–60s | 7 | 1.6% |
| 1–5m | 4 | 0.9% |
| 5m+ | 3 | 0.7% |
| Total | 434 | 100.0% |
An abandon under 10s is a quick hang-up, deliberately outside the rulebook’s Abandon-% denominator (§E). The buckets past 30s are the ones staffing can actually move — a member who held for a minute wanted to be answered.
Hour-of-week shape
all history · ignores the date filter (program-grain rollup)| 12a | 1a | 2a | 3a | 4a | 5a | 6a | 7a | 8a | 9a | 10a | 11a | 12p | 1p | 2p | 3p | 4p | 5p | 6p | 7p | 8p | 9p | 10p | 11p | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mon | · | · | · | · | · | · | 4s | 4s | 6s | 1,144s | 14s | 1,024s | 5s | 2s | 7s | · | · | · | · | · | ||||
| Tue | · | · | · | · | · | · | · | · | 3s | 4s | 5s | 5s | 9s | 6s | 4s | 1s | 2s | · | · | · | · | |||
| Wed | · | · | · | · | · | 1s | · | 12s | 5s | 4s | 5s | 907s | 10s | 8s | 5s | · | 1s | · | · | · | · | · | ||
| Thu | · | · | · | · | · | · | 2s | 3s | 3s | 4s | 7s | 7s | 4s | 1s | · | · | · | · | · | · | ||||
| Fri | · | · | · | · | · | · | · | 3s | 3s | 3s | 19s | 33s | 4s | 15s | · | 24s | · | · | · | · | · | |||
| Sat | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | ||||
| Sun | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · | · |
Mean queue wait per inbound arrival in the cell (sum ÷ count at read time). Thin cells swing hard — each tooltip carries its n; treat single-digit-n cells as anecdotes. This rollup is program-grain and whole-history by construction — the date filter above deliberately does not apply, the same convention as list-lifetime figures.
Queue waits over 2 hours — data suspects, not service events
campaign grain · call detail ships with sign-in| Campaign | Calls | Worst waitThe longest recorded queue wait on this campaign in the window | Dates |
|---|---|---|---|
| ACA00003 | 1 | 17.5h | 2026-07-06 |
| ACA00004 | 1 | 19.3h | 2026-07-20 |
| ACA00007 | 1 | 19.3h | 2026-07-29 |
| Total | 3 | — | — |
A 17–19 hour “queue wait” is a telephony logging artifact, not a member experience. These rows are excluded from nothing — they sit inside the certified totals — but they are flagged here and on the Trust tab so the parallel run can decide what to do with them deliberately.