Cost Per Ticket

Accurate CPT calculation requires fully-burdened cost accounting, not just agent salaries. Full formula: CPT = (Total Support Department Cost) / (Total Tickets Handled per Period). The numerator includes: agent salaries and benefits (the largest component, typically 60–70% of total cost), manager and ops team salaries, helpdesk software licensing fees, quality assurance tools, training costs, office space (for in-office teams), and overhead allocation (IT support, facilities as a % of headcount). The denominator should exclude tickets that are handled by bots or self-service with no agent touch — CPT measures human-handled contacts. Industry averages: phone support CPT $25–50; email/chat CPT $8–15; AI/chatbot CPT $0.50–2. Product Ops tracks CPT quarterly and models the cost reduction opportunity of deflection investments using these baselines.

CPT reduction should never compromise quality — reducing support headcount through understaffing simply shifts cost to longer resolution times, lower CSAT, and ultimately customer churn. Legitimate levers: (1) Self-service deflection — every ticket prevented by a good help center article is a ticket whose cost falls to near-zero. Even at a $10 average CPT, a single relevant article that deflects 100 tickets per month saves $1,000/month. (2) Handle time reduction — knowledge base tools embedded in the agent workflow, AI-suggested replies, and streamlined processes directly reduce AHT. A 2-minute AHT reduction across 1,000 tickets/week = 33 agent-hours saved per week. (3) Tier 0 / chatbot handling — common, simple tickets (password resets, billing FAQ, plan comparisons) handled by a chatbot at < $2/ticket vs. $12/email. (4) First-contact resolution improvement — each ticket that requires two or more agent interactions has a CPT multiplier. Improving FCR from 70% to 80% effectively reduces functional CPT by 10–15%.

CPT and CSAT are the primary efficiency/quality tension in support leadership. Cutting costs by reducing agent headcount, shortening handle time through rigid time limits, or deflecting conversations to self-service before they're ready all risk lowering CSAT. The resolution is: invest in smart deflection (that genuinely helps customers before they need to contact support), not coercive deflection (forcing customers to self-serve when they need human help). Measure CSAT for self-service experiences separately from human-handled experiences to detect quality degradation early. Track recontact rates (customers who self-served and then submitted a ticket anyway): high recontact rates prove self-service is not solving the problem, and CPT calculations that exclude these recontacts are artificially flattering. The ideal support leader owns both CPT and CSAT targets simultaneously, using both as guardrails in every investment decision.