Average Handle Time (AHT): Formula, Causes, and How to Reduce It

The AHT formula and its components

The standard formula is: AHT = (Total Talk Time + Total Hold Time + Total ACW) ÷ Total Handled Interactions. Each component is summed across all interactions in the measurement period before dividing — not averaged and then combined. For example: 150 minutes total talk + 25 minutes total hold + 50 minutes total ACW = 225 total minutes ÷ 50 interactions = 4.5 minutes (4 minutes 30 seconds) AHT.

Talk time is the active conversation between agent and caller — from when the call connects to the agent until the call ends, minus any hold time. It is what most people picture when they think of call length, but it is only one component of handle time.

Hold time is the duration the caller is placed on hold during the interaction. This is tracked separately because hold time is not productive conversation — it is time the caller is waiting while the agent researches, consults a colleague, or navigates a system. High hold time within AHT is a specific efficiency problem with its own causes and interventions.

ACW (after-call work) is everything the agent does after hanging up — writing call notes, updating CRM records, selecting a disposition code, scheduling a follow-up, or completing a form — until the agent changes their status back to available. ACW is frequently underestimated as an AHT driver. On high-volume queues, even 30 extra seconds of ACW per call multiplies into significant daily capacity impact.

Why lower AHT is not always better

AHT measures efficiency, not quality. A low AHT achieved by ending calls quickly rather than resolving issues completely drives the outcomes that cost more than the time saved: repeat contacts, escalations, and lower CSAT scores. When agents are coached or incentivized toward a specific AHT target without regard to resolution quality, the number improves while operational costs increase.

The right approach is to set AHT targets per call type and per complexity tier, not a single contact center-wide number. A complex troubleshooting call has a legitimately higher handle time than a simple address update. Comparing the two — and penalizing an agent for high AHT on a complex queue — is measuring the wrong thing.

What causes high AHT

Nine common drivers account for the majority of elevated AHT across contact center types:

1. Long hold times searching for information. Agent places the caller on hold while navigating systems, consulting policies, or looking up account details that should be surfaced automatically.
2. Complex issue types. Some call types genuinely require more time. This is not a problem to solve — it is a driver to isolate so it doesn't distort AHT benchmarks for simpler queues.
3. Poor first-call routing. Calls landing on the wrong team or agent type generate extended handle time as the agent either struggles with an unfamiliar issue type or transfers the call.
4. Knowledge base gaps. Agents can't find the answer quickly, so they hold the caller or improvise — both increase handle time and reduce consistency.
5. CRM and system slowness. Application performance directly affects AHT. Agents waiting for screens to load accumulate hold time and silence while the customer waits.
6. Inefficient ACW. Manual logging, unstructured note-taking, and unclear disposition taxonomies all extend after-call work beyond what the interaction requires.
7. Agent skill and training gaps. Less experienced agents take longer to navigate issues, find information, and communicate resolution — training and coaching are the direct interventions.
8. Lengthy verification procedures. Multi-factor identity verification at the start of every call adds consistent overhead — appropriate for security-sensitive interactions, worth reviewing for low-risk ones.
9. Unclear disposition categories. When agents spend time choosing between ambiguous disposition codes, ACW extends. Rationalizing the disposition taxonomy is a quick win for ACW reduction.

How to reduce AHT without hurting quality

The constraint in every AHT reduction effort is that speed improvements should not come at the cost of resolution completeness. The following nine methods address the structural causes of high AHT rather than the symptom:

1. Screen pop and CRM auto-population. Surfacing the customer record, open cases, and recent interaction history at the moment the call connects eliminates the opening verification and account lookup phase — one of the most consistent AHT reductions available.
2. Knowledge base improvements. A well-structured, searchable KB with accurate content reduces hold time during complex calls. Measure how often agents put callers on hold versus successfully resolving in-call to identify KB contribution to AHT.
3. Skills-based routing. Routing calls to the agent most qualified for that call type reduces handle time for both agent and caller — the right agent finds the answer faster and escalates less often.
4. Disposition code rationalization. Consolidate or clarify disposition taxonomies so agents spend seconds, not minutes, completing ACW categorization.
5. AI call summaries and auto-ACW. AI tools that generate draft call summaries and auto-populate CRM fields from transcript data can cut ACW by 40–60% on compatible call types.
6. Coaching on navigation efficiency. Some AHT is lost to agents navigating slowly between systems or using inefficient keyboard/application sequences — this is coachable with screen recording and observation.
7. IVR self-service deflection. Deflecting high-volume, low-complexity transactions (balance inquiries, appointment confirmations, status checks) to IVR reduces the agent queue and improves average AHT by removing the easiest calls while preserving agent capacity for complex ones.
8. Process and form streamlining. Audit the steps agents are required to complete during and after a call. Required fields, mandatory scripts, and multi-step confirmation workflows all add to handle time — some may be reducible without compliance risk.
9. Call type segmentation with separate AHT targets. Set AHT benchmarks per queue and call type. This doesn't reduce AHT directly but stops high-complexity queues from creating pressure to rush agents on straightforward work.

AHT vs customer experience trade-off

The core tension in AHT management is that the metric rewards speed and penalizes thoroughness. An agent who spends an extra two minutes ensuring the customer fully understands the resolution will show higher AHT and lower throughput than an agent who ends calls quickly — but their repeat contact rate will likely be lower, and their CSAT higher.

First-call resolution (FCR) is the counterbalancing metric. High FCR means the customer's issue was resolved without requiring a follow-up contact — no callback, no repeat call, no escalation. FCR and AHT are often in tension when contact centers are optimized purely for efficiency: lower AHT targets frequently produce lower FCR. The total cost of a call is not AHT × cost-per-minute — it is (AHT + (probability of repeat contact × AHT of repeat contact)) × cost-per-minute.

AHT × repeat contact rate is more meaningful than AHT alone. A contact center with a 4-minute AHT and a 35% repeat contact rate is doing more work per issue resolved than one with a 5-minute AHT and a 15% repeat contact rate. See FCR vs repeat contact rate for how these two metrics interact and how to measure them together.

AHT in workforce management and Erlang C

AHT is a primary input to Erlang C staffing calculations. The Erlang C formula takes call volume (calls per hour), AHT (in seconds), and target service level (e.g., 80% of calls answered within 20 seconds) to produce required agent count. Because AHT feeds directly into the formula, even small changes in average handle time produce meaningful staffing impact at scale.

On a queue handling 1,000 calls per hour, a 30-second reduction in AHT from 5 minutes to 4 minutes 30 seconds reduces total handling work by approximately 8.3%. At high volumes, this translates to measurable reduction in required agent count — or equivalent capacity to handle more volume with the same headcount. This is why operations teams treat AHT improvement as both a quality metric and a capacity planning lever.

For the full treatment of staffing models and how AHT flows through workforce management, see call center workforce management.

AHT segmentation analysis

A single contact center-wide AHT number obscures more than it reveals. Meaningful AHT analysis requires segmenting across at least the following six dimensions:

1. By agent. Individual AHT variance identifies coaching opportunities and high performers whose process efficiency is worth understanding and replicating.
2. By call type. The most important segmentation dimension. Different call types have fundamentally different handle time profiles — comparing across types produces noise rather than signal.
3. By queue or team. Teams handling different product lines, customer segments, or escalation tiers will have different AHT baselines. Queue-level benchmarks are more actionable than center-wide ones.
4. By time of day. AHT often varies across shift patterns — peak-period calls may involve different issue distributions; end-of-shift behavior may show compression pressure. Time-of-day analysis surfaces scheduling and staffing insights.
5. By channel (multi-channel operations). Voice, chat, email, and messaging interactions have structurally different handle time profiles. If your AHT report rolls up across channels, segment them before drawing conclusions.
6. By customer segment. High-value or complex customer segments may require longer handle times by design. Segmenting by customer tier identifies whether AHT variation reflects legitimate complexity or operational inefficiency.

AHT vs related metrics

For a full contact center metrics reference, see contact center KPIs.

Frequently asked questions