Ticket Deflection

Definition

Ticket deflection is the reduction in inbound support contacts achieved when customers find answers in self-service content — typically a knowledge base, help center, or AI chatbot — before escalating to a human agent. It’s measured as a count (how many contacts were avoided) or as a percentage of total potential contacts.

The term is often used interchangeably with deflection rate, though strictly speaking they’re different: ticket deflection is the absolute count, deflection rate is the percentage.

Example

A SaaS support team handles 1,000 tickets per month in January without a knowledge base. In February, they launch a help center with 50 articles and a Beacon widget. By June, they’re handling 750 tickets per month — 250 contacts are now self-resolving through the KB. Their ticket deflection for that month is 250 contacts. Their deflection rate is 25%.

Why it matters for buyers

Deflection per dollar is the only honest ranking of knowledge base software. If a tool costs £500/mo and deflects 200 tickets per month, and each ticket costs your team £8 in agent time, the tool is saving £1,600/mo in labour — a 3.2x ROI before accounting for CSAT improvements.

The problem: vendors advertise “up to 70%” deflection in their marketing materials. The audited median for B2B SaaS teams with a serious content investment sits at 15-30% by month 6 (per benchmarks from Supportbench, Corebee, and supp.support, 2026). Anyone selling you 70% in 90 days is selling you cherry-picked B2C FAQ data from a high-volume consumer account.

Realistic planning targets for a new KB deployment:

• Month 1-2: 8-15% deflection (content is thin, search is unoptimised)
• Month 3-6: 15-30% deflection (core articles published, Beacon tuned)
• Month 6-12: 25-40% deflection (content gaps filled, AI layer tuned)
• Year 2+: 35-50% deflection for teams with editorial investment + AI

How vendors measure it (and why the numbers vary so much)

Different vendors measure deflection differently — and some definitions are much more flattering than others:

Pessimistic (honest) measure: Count tickets where Beacon/AI was shown AND user did not submit a ticket. Divide by total contacts.

Optimistic (marketing) measure: Count all sessions where the KB was accessed and no ticket was submitted. Divide by all KB sessions.

The optimistic measure inflates numbers by 3-5x because it counts users who visited the KB to confirm they needed to contact support — they found no answer, but they still “deflected” in the vendor’s metric.

When evaluating vendor deflection claims, ask: “What’s the denominator? Is it total contacts or total KB sessions?”

Tools that publish honest deflection data

• Help Scout: Honest about the Beacon deflection metric (contacts avoided ÷ total contacts)
• Document360: Dashboard shows “articles surfaced before ticket” with click-through tracking
• Zendesk Guide: Suite Pro analytics includes deflection tracking with clear methodology

Tools that don’t publish median deflection (they publish “up to” or case study numbers only): Document360 (in marketing), Guru (marketing claims 80%), most other vendors.

Related terms

• Deflection rate — the percentage form of ticket deflection
• Self-service rate — the % of users who resolve without any support contact (includes phone, email, chat)
• First contact resolution (FCR) — % of tickets resolved in a single interaction; KB quality feeds this
• Search-failure rate — % of searches that return no results; the most actionable KB analytics metric