Customer Support Automation Audit

1. Support channels — map what you actually run

List every channel where customers reach you. Mark active volume and current response time.

• Email — primary inbox(es), average volume per week, current SLA.
• WhatsApp — number(s), volume, who answers, after-hours coverage.
• Live chat — platform, volume, hours of coverage, deflection rate.
• Phone — who picks up, voicemail handling, ticket creation flow.
• Other — DMs, contact form, in-product, partner channels.

2. Audit questions

Score each item from 1 (rarely a problem) to 5 (constant pain).

• Repetitive requests — how many of your tickets are the same 10 questions?
• Response delays — how often do customers wait > 4 business hours for a first reply?
• Escalation issues — how often do tickets bounce between people before resolution?
• Documentation gaps — how often do agents answer from memory vs. an SOP?
• After-hours coverage — what happens between 6pm and 9am?
• Quality drift — do answers vary by agent / day / channel?

3. Automation opportunities

For each opportunity, note the channel, the volume it touches, and the % of tickets it could deflect or accelerate.

• FAQ handling — AI answers top 20 questions with an easy escape to human.
• Triage — AI classifies every ticket (billing / tech / sales / refund) and tags urgency.
• Routing — auto-assign tickets to the right person / team / queue.
• Reminders — automated nudges for missing info, paused tickets, expiring quotes.
• Status updates — proactive 'your ticket is being looked at by X' messages.
• Reply drafting — AI writes the first draft, humans approve and send.

4. Scoring system

For each opportunity, compute Impact × Ease, then sort. Ship the top 3 this quarter.