What is average handle time?
Average handle time (AHT) is the average amount of time a human support agent spends actively working on a single ticket — from first reading through to close, including any follow-up. It's a capacity and efficiency metric: lower AHT means agents can handle more tickets per hour, which directly affects staffing costs and queue time.
AHT is measured differently across channels. For email, it's the total time an agent has the ticket open and active. For live chat, it's the duration of the conversation plus any post-chat wrap-up. For phone, it's call time plus after-call work. The benchmarks in this post focus on email and chat, which are the dominant ecommerce support channels.
AHT is most useful as a per-category metric, not just an overall average. An AHT of 8 minutes overall might be hiding a 20-minute average on complex return disputes that's dragging up an otherwise efficient operation.
Ecommerce support AHT typically ranges from 5–15 minutes per ticket across email and chat. Simple, data-grounded tickets (order status, basic product questions) run 3–6 minutes. Complex tickets (disputes, multi-item returns, escalations) run 12–25+ minutes. Strong performers average 5–8 minutes overall.
AHT benchmarks by ticket type
WISMO is the most automatable — not just because of volume, but because handle time is low to begin with. Automating 500 WISMO tickets at 4 minutes each saves 2,000 agent-minutes (33 agent-hours) per month. Complex disputes have the highest AHT and are harder to automate, but improving processes for them (templates, investigation workflows) yields significant time savings even without AI.
| Ticket type | Typical AHT | Strong performer AHT |
|---|---|---|
| Order status / WISMO | 3–6 min | 2–4 min |
| Return eligibility questions | 5–9 min | 4–6 min |
| Shipping issues | 7–12 min | 5–8 min |
| Product questions | 5–10 min | 4–7 min |
| Billing / refund requests | 8–15 min | 6–10 min |
| Account issues | 6–11 min | 4–8 min |
| Complex disputes / damaged goods | 15–30 min | 12–20 min |
| Overall blended AHT | 6–15 min | 5–8 min |
What drives AHT up
High AHT is usually caused by one of four factors: missing information that agents have to go find, unclear policies that require escalation for decisions, complex ticket types that are inherently time-intensive, or inefficient workflows (context-switching, tool friction).
Information lookup time
Agents who have to navigate between multiple systems to find an order, check a shipping status, or look up a policy spend a significant portion of their time on information gathering rather than resolution. A unified helpdesk that surfaces order data in the same view as the conversation eliminates this friction.
Policy ambiguity and approval chains
When agents can't make a decision without checking with a manager, AHT inflates — especially if managers aren't immediately available. Clear policies with empowered agents (can issue refunds up to $X without approval) dramatically reduce this delay.
Response drafting
Writing a fresh, personalized response to every ticket from scratch is slow. A comprehensive response template library for the top 15–20 ticket types is one of the highest-ROI investments for reducing AHT without any other tooling.
Ticket complexity mix
Blended AHT is sensitive to the mix of easy and complex tickets. If complex disputes are 15% of your volume but 40% of your agent-time, that's where the leverage is — either automate the upstream resolution or improve the process for handling disputes.
How AI reduces AHT
AI affects AHT through two mechanisms: deflection (removing tickets from the human queue entirely) and augmentation (making human agents faster on the tickets they do handle).
The deflection effect on AHT is indirect: when AI handles 50% of tickets, human agents work only on the harder 50%. Their AHT goes up on a per-ticket basis — because they're handling more complex cases — but the total agent-hours required falls dramatically.
The augmentation effect is direct: AI can draft a first response for human agents to review and edit, surface the relevant order data in the same interface, suggest the applicable policy, and flag similar resolved tickets for reference. All of these reduce the time agents spend on information gathering and response drafting.
When Bookbag is in the workflow, the tickets that reach human agents arrive with full context — conversation summary, order data, suggested action — already surfaced. Agents spend their time on judgment and empathy, not on lookup and drafting. Human AHT typically falls 20–35% on the tickets that remain in the human queue.
| Mechanism | AHT impact | How |
|---|---|---|
| AI deflection (ticket removed from queue) | 100% reduction for that ticket | AI resolves completely — no human time at all |
| AI drafting (human reviews AI draft) | 30–50% AHT reduction | Reviewing/editing is faster than drafting from scratch |
| Context surfacing (order data in view) | 15–25% AHT reduction | Eliminates system-switching for information lookup |
| Policy suggestion | 10–20% AHT reduction | Agent doesn't need to search for the applicable rule |
How to lower AHT without AI
AHT can be improved significantly without AI through process and tooling changes. If you're not yet ready for AI automation, these tactics have the highest ROI:
- 1Build a response template library for your top 20 ticket types. Every common question should have a polished, editable template. This alone typically reduces AHT by 20–30%.
- 2Surface order data in your helpdesk. Whether through a native integration or a sidebar extension, agents should see the relevant order when they open a ticket — not navigate to Shopify admin separately.
- 3Clarify and document your policies so agents can answer without escalating. Every policy question that requires a manager adds 5–15 minutes of AHT.
- 4Empower agents with action authority. Agents who can issue refunds up to $50 and initiate returns without approval resolve those tickets in one step instead of three.
- 5Review AHT by ticket type monthly and focus improvement effort on the highest-AHT, highest-volume category — that's where the most total time is being spent.
- 6Reduce context-switching: use a helpdesk that keeps the conversation, order data, and knowledge base in a single view.
Key takeaways
- Ecommerce AHT benchmarks: 6–15 min typical overall, 5–8 min for strong performers.
- WISMO tickets run 3–6 min and are nearly fully automatable — freeing significant agent-hours.
- Complex disputes run 15–30 min and are harder to automate; process improvements yield the most AHT reduction here.
- AI reduces AHT through deflection (removes tickets entirely) and augmentation (drafts, context, policy suggestions).
- A response template library is the highest-ROI AHT improvement available without AI tooling.