How to calculate cost per ticket
For a typical ecommerce setup, labor is 70–80% of total support cost. That means CPT is highly sensitive to headcount and agent efficiency — which is why deflection (fewer tickets reaching humans) is the biggest lever on CPT.
The calculation: Total monthly support spend ÷ Total tickets resolved that month = Cost per ticket. For a blended AI + human operation, include AI platform costs in the numerator but count all contacts in the denominator — including AI-resolved ones — to get the true blended CPT.
- Agent labor: salaries, benefits, and payroll taxes for support staff, prorated by time spent on support
- Platform costs: helpdesk software, chat tools, AI subscriptions — all support-specific tooling
- Management overhead: team lead and manager time spent on support operations
- Training and onboarding: amortized over the employment period
- Infrastructure: telephony, integration costs, any custom development
Industry-typical cost per ticket for ecommerce support ranges from $8–$25 for human-handled tickets. Strong performers with efficient agents reach $5–$10. With AI automation, blended CPT typically falls to $2–$6, depending on deflection rate.
Industry benchmarks
These ranges are based on industry-typical labor costs and team structures. Your actual CPT will depend on agent salaries in your region, your helpdesk platform costs, and your ticket volume. A small team in a high-wage market will have higher CPT than the same team in a lower-wage market.
| Setup | CPT range | Notes |
|---|---|---|
| Small team, email only (1–3 agents) | $12–$30 | High overhead per ticket at low volume |
| Mid-size team, email + chat (4–10 agents) | $10–$20 | Better leverage, some efficiency gains |
| Larger team (10+ agents) | $8–$18 | Economies of scale, more tooling |
| Self-service / help center deflection | $0.50–$2 | Near-zero marginal cost per deflected contact |
| AI agent (blended with human handling) | $2–$8 | AI lowers the average across all contacts |
| AI agent, high deflection (60%+) | $1.50–$5 | High deflection pulls blended CPT very low |
What drives cost per ticket up
Understanding the cost drivers lets you identify which interventions will move the number most. For most ecommerce stores, two or three factors dominate.
Low volume relative to headcount
Support economics favor scale. A team of three agents handling 500 tickets per month has much higher CPT than a team of eight handling 3,000 tickets. Small-volume stores are often better served by a lean AI-first setup than by maintaining dedicated agents.
High average handle time
Tickets that take 15+ minutes for a human to resolve cost significantly more than tickets resolved in 3–5 minutes. Complex tickets — disputes, multi-item return issues, damaged goods claims — are expensive regardless of volume. Identifying which ticket types have high AHT and either automating them or creating better templates is a direct CPT lever.
High repeat contact rate
When customers reopen tickets or contact again about the same issue (low FCR), that inflates ticket count without inflating revenue. Reducing repeat contacts by resolving issues completely the first time reduces CPT and improves CSAT simultaneously.
Expensive channels in the mix
Phone support costs 3–5x more per resolution than email or chat, because of the real-time agent time commitment and the 1:1 nature of voice calls. Stores that route most contacts to phone before trying chat will have systematically higher CPT.
How AI lowers cost per ticket
This model uses illustrative figures ($15 human CPT, $0.50 AI cost per contact) to show the math. The key insight is that each percentage point of deflection is worth roughly $0.14 in CPT reduction in this example — the benefit compounds as deflection rises. Stores using Bookbag with high deflection rates consistently report blended CPTs in the $3–$6 range, compared to $12–$20 before AI.
| Deflection rate | Human CPT | AI cost/contact | Blended CPT |
|---|---|---|---|
| 0% (no AI) | $15 | — | $15.00 |
| 20% deflection | $15 | $0.50 | $12.10 |
| 40% deflection | $15 | $0.50 | $9.20 |
| 55% deflection | $15 | $0.50 | $7.05 |
| 65% deflection | $15 | $0.50 | $5.75 |
| 70% deflection | $15 | $0.50 | $4.85 |
Building your own cost model
Before investing in AI support, build a simple cost model to project the return. It takes about 20 minutes and gives you a defensible number to share with stakeholders.
- 1Calculate your current CPT: total monthly support spend ÷ monthly ticket count.
- 2Estimate your likely deflection rate with AI — use the benchmark table in this post as a starting point. 40–55% is realistic for most ecommerce stores; adjust up if you have high WISMO volume.
- 3Calculate the new blended CPT using the formula: (deflection rate × AI cost per contact) + ((1 − deflection rate) × human CPT).
- 4Multiply the CPT reduction by your monthly ticket volume to get monthly savings.
- 5Compare monthly savings to the AI platform cost to get your payback period.
- 6Add qualitative benefits that don't show in CPT: 24/7 coverage, faster response time, CSAT improvement.
Key takeaways
- Industry-typical cost per ticket for ecommerce is $8–$25 for human-handled tickets; strong performers reach $5–$10.
- AI automation lowers blended CPT to $2–$6 at deflection rates of 50–70%.
- Labor is 70–80% of support cost, so deflection (removing tickets from human queues) is the primary CPT lever.
- Build a simple cost model before evaluating AI: current CPT × deflection rate × volume = projected savings.
- Each percentage point of deflection reduces blended CPT by a measurable, compounding amount.