Average Handle Time

What it means

AHT = total active handle time across all tickets ÷ number of tickets handled. Active handle time includes time spent reading the conversation, composing responses, and any wrap-up tasks (tagging, categorizing, noting outcomes). It excludes time the ticket sits in queue waiting for assignment. In ecommerce support, AHT varies sharply by ticket type. Order-status inquiries handled by a human take 2–4 minutes. Return and refund disputes can take 8–15 minutes. Complex fraud or damaged-goods cases may take 20+ minutes across multiple interactions. Mixing all ticket types into a single AHT figure obscures where agent time is actually going. AI assistance directly compresses AHT for human-handled tickets. When an AI drafts the first reply, surfaces the relevant order data, and pre-fills tags, agents spend less time on each interaction without sacrificing quality. The net effect is more tickets handled per agent per hour without adding staff. AHT should always be evaluated alongside CSAT and FCR. An agent with low AHT who frequently sends incomplete answers that trigger follow-up tickets is creating more work, not less.