Average resolution time (ART)

Average Resolution Time (ART) measures how long it takes for a customer issue to be fully resolved, from the moment the request is created until the case is closed. Unlike metrics that focus only on active handling time, ART accounts for everything that happens between first contact and final resolution, including internal escalations, follow-up messages, waiting periods, and asynchronous communication. ART captures the entire lifecycle of an issue, so it overlaps with the performance of AI systems that support routing, triage, customer self-service, and workflow automation. 

How average resolution time (ART) works

ART is calculated by averaging the total time it takes to resolve all closed tickets over a given period. For instance, if ten issues collectively required 500 hours from open to close, the ART would be 50 hours. The calculation includes:

• Time between customer messages and agent responses
• Delays caused by escalations or handoffs
• Internal research or verification periods
• AI-driven self-service time before human involvement
• Any waiting period before the ticket is officially closed

Average resolution time captures every contributing factor, offering a more holistic view of performance than metrics like Average Handle Time (AHT), which measure only the agent’s active work. This metric is especially helpful in environments with asynchronous communication or multiple channels, where resolution may unfold across calls, emails, chats, and automated interactions.

The role of ART in AI-supported customer journeys

AI significantly shapes ART by determining how efficiently issues are understood and resolved. Well-trained AI models can reduce resolution time by:

• Automatically categorizing and prioritizing incoming requests
• Providing customers with accurate self-service answers
• Suggesting next-best actions to agents in real time
• Summarizing previous interactions so agents don’t need to review long histories
• Identifying knowledge gaps that slow down teams

These capabilities help reduce unnecessary back-and-forth and improve consistency across channels.

In contrast, slow or inaccurate AI can increase ART. If a model misclassifies intent, returns outdated information because of high latency, or takes too long to generate responses due to slow inference time, issues may require additional agent intervention. This can prolong the overall resolution window.

ART is also affected by cross-team dependencies. AI can route tasks to the correct internal groups or automate parts of validation workflows, but unresolved bottlenecks like slow approvals or manual verification steps still extend the time to resolution.

Key drivers of average resolution time (ART)

Average resolution time is shaped by several operational and technical factors that determine how quickly an issue moves from initial contact to full resolution. Key drivers include:

• Routing accuracy: Misrouted or misclassified tickets require correction, adding time before the right team begins work.
• Knowledge quality: Outdated or incomplete information slows both AI systems and agents, extending resolution cycles.
• Workflow clarity: Complex or unclear escalation paths introduce delays, especially when multiple reviews or approvals are required.
• System responsiveness: High latency or slow inference time delays automated responses and agent-assist suggestions.
• Channel switching: When customers move between channels without unified context, agents spend additional time reconstructing the issue.

Together, these factors determine how efficiently AI and human teams can work toward closing a case.