Longest delay in queue
Definition
Longest delay in queue
Longest delay in queue (LDQ) is the longest time any single caller waits on hold before an agent answers. Contact centers watch it live because one stranded caller often signals a routing failure, a staffing gap, or a volume spike nobody forecast.
LDQ is a peak metric, not an average. Average speed of answer (ASA) can look healthy while one caller sits for 20 minutes. LDQ surfaces that outlier in real time — before it turns into a complaint or a churn event.
Most cloud contact center platforms print LDQ on the supervisor wallboard and fire an alert once a threshold is crossed. That gives team leads a window to reroute traffic, add capacity, or push the caller to a callback offer.
LDQ is also a leading indicator. When it climbs, abandonment rate and service level usually move next. Reading the three together tells you what kind of problem you actually have.
Key takeaways
- LDQ measures the single longest live hold time, not the queue average.
- It exposes routing gaps and staffing shortfalls that averages hide.
- Most centers read LDQ next to abandonment rate and service level.
- Common alert thresholds sit between 3 and 10 minutes, sector dependent.
- Callback offers and skills-based routing are the fastest LDQ fixes.
How it works
LDQ is calculated live. The platform starts a timer the moment a call enters the queue and stops it when an agent connects. The longest-waiting caller holds the reported value until they are answered, abandon, or are outlasted by someone newer.
Supervisors read LDQ on a wallboard beside calls in queue and longest available agent. When it crosses a preset threshold, a visual or audio alert fires so the floor can react while the caller is still on the line.
Targets vary by sector. The bands below reflect what operations leads commonly set on the floor, not a published standard, so treat them as starting points rather than compliance numbers.
| Contact center type | Typical LDQ target | Common alert threshold |
|---|---|---|
| Retail and e-commerce support | Under 2 minutes | 3 minutes |
| Financial services | Under 3 minutes | 5 minutes |
| Healthcare triage | Under 90 seconds | 2 minutes |
| Utilities and telco | Under 5 minutes | 8 minutes |
| Offshore BPO voice pod | Under 4 minutes | 6 minutes |
Underneath the wallboard, the math is queueing theory. The Erlang C formula predicts how many agents you need to hold LDQ under target for a given call volume and handle time.
Call Centre Helper’s worked Erlang C example sets service level at 80% of calls answered in 20 seconds, caps agent occupancy at 85%, and applies a 30% shrinkage allowance before it lands on a headcount.
That worked example needs 14 agents on the phones and 20 rostered once shrinkage is applied. Understaff that roster and LDQ blows out first — long before the daily service level average moves.
Track LDQ as one input inside a broader key performance indicator stack, never as a standalone score. On its own it tells you a caller waited. Read next to average handle time, it tells you why.
Examples
LDQ shows up wherever inbound voice volume is unpredictable. The scenarios below are composite but typical: a seasonal retail spike, an annual enrollment crush, an unplanned outage, and a steady-state offshore pod working to a client threshold.
- Retail returns line, Black Friday 2024. A mid-sized US retailer watched LDQ jump from 90 seconds to 14 minutes inside an hour after a carrier delay went viral. The supervisor triggered callbacks and pulled two agents off live chat.
- US health insurer, open enrollment. LDQ runs 8–12 minutes on the first Monday of December most years. The team pre-schedules 30% more agents for that day and still watches the wallboard hourly through the peak.
- Telecom outage handling. LDQ passes 20 minutes during regional outages. Most carriers push those callers to an interactive voice response (IVR) message and log them for an automated callback once service is restored.
- Offshore business process outsourcing (BPO) pod. A Manila team supporting a software client caps LDQ at 4 minutes with skills-based routing. Anything longer pages the client’s operations lead inside the same shift.
The pattern holds across all four. Waiting is the part customers remember — HubSpot’s customer satisfaction research reports that 79% of buyers pick live chat specifically because it is immediate. Voice queues carry the same expectation.
The International Customer Management Institute (ICMI) publishes the contact center benchmarking studies and workforce playbooks most operations leads consult when they set an LDQ alert threshold for the first time.
Related terms
LDQ sits inside the queue metric family, and it only reads correctly next to its neighbours. The terms below cover the same call from different angles: wait, abandonment, target attainment, talk time, and outcome. Agent scheduling sits outside this cluster.
- Average Speed of Answer: the mean queue wait across all answered calls.
- Abandonment Rate: the share of callers who hang up before an agent picks up.
- Service Level: the percentage of calls answered inside a target window, usually 80% in 20 seconds.
- Average Handle Time: the mean length of the agent interaction itself.
- First Call Resolution: the share of issues closed on the first contact.
- Call Center: the operational unit these queue metrics describe.
FAQ
These are the questions operations leads ask first when LDQ appears on a wallboard. Each answer assumes a voice queue, though the same logic applies to chat and messaging queues that hold customers in line.
What is a good longest delay in queue?
Most centers target LDQ under 3 minutes and set the alert between 5 and 8 minutes. Regulated sectors such as healthcare triage aim tighter, often under 90 seconds, because a delayed call carries a real safety cost.
How is LDQ different from average speed of answer?
ASA is the mean of every answered call, while LDQ is the single longest live wait. A healthy ASA can hide one caller sitting for 20 minutes, and LDQ makes that outlier visible in real time.
Which tools report LDQ?
Every mainstream cloud contact center platform surfaces LDQ on the supervisor wallboard, including Amazon Connect, Five9, Genesys, NICE CXone, and Talkdesk. Most let you set custom thresholds and route the alert to Slack, Teams, or email.
How do I reduce longest delay in queue?
The fastest levers are skills-based routing, callback offers at peak, and pulling agents in from non-voice channels when LDQ trends up. Over a longer horizon, better forecasting and a shrinkage allowance near the 30–35% band prevent most spikes.
Does LDQ include calls that abandon?
No, LDQ tracks the current longest-waiting live caller — once someone hangs up they drop out of the LDQ figure and roll into abandonment rate instead.
Outsourcing to the Philippines’ top outsourcing destination? Shortlist from the Top 40 BPO companies in the Philippines, size the pod with the Outsourcing Cost Calculator, then explore OA’s outsourcing hubs to compare providers reporting LDQ.
Independent
