What is ticket deflection?
Ticket deflection happens when a person resolves a support question through self-service, such as a help article or automated assistant, instead of opening a ticket. Measuring it requires a clear rule for what counts as a resolved question.
How deflection is measured
Start by defining a support question and what counts as resolution. A confirmed answer or completed task is stronger evidence than a help center visit. If you infer resolution because no ticket followed, label that measure as an estimate and state the observation window.
For a rate based on tracked questions, divide confirmed self-service resolutions by all questions in the same period. Avoid counting an escalated question once in self-service and again in human support.
A help-center-sessions-to-tickets ratio measures a different thing. It describes channel activity, so it should not be presented as the share of questions resolved. Compare ticket volume per active user as an additional signal, with changes in the customer base and product in mind.
Finding questions suited to self-service
Review a sample of tickets and group them by task. Setup instructions, configuration questions, and requests to find a feature may have repeatable answers. Bugs, access disputes, and account decisions can require investigation or a person with authority to act.
Our guide to how to deflect support tickets describes a way to organize that review.
Improving self-service resolution
Check whether people can find the relevant answer, whether it is accurate, and whether they can complete the task after reading it. Update instructions when the workflow changes and offer guidance for tasks with several steps.
Review failed answers and repeat contacts. A lower ticket count is not enough if people abandon the task or cannot reach support. Make escalation available and pass along the context needed to continue.
How Frigade fits
Frigade Assistant answers product questions and guides users through tasks using learned product knowledge and connected documentation. A connected helpdesk can receive a confirmed support request with conversation context when someone needs your team.
The support deflection use case explains the workflow. Our customer support AI comparison covers related tools.
Common questions
- How is ticket deflection rate calculated?
- If you track questions across self-service and human support, one rate is confirmed self-service resolutions divided by all tracked questions, multiplied by 100. Use the same period and count each question once. State your resolution rule: a help center visit or the absence of a ticket does not, by itself, confirm that the person got an answer.
- What is a good ticket deflection rate?
- A useful target depends on the questions, the audience, and the resolution rule. Compare the same categories over time and check answer quality alongside the rate. Keep a route to human support for issues that need investigation, an account decision, or a correction your self-service tools cannot make.