support-toolkit

Support Toolkit

Overview

Support tools define how customers experience your company when things go wrong — and when they need help getting things right. The mistake: buying the tool your last company used instead of the tool that matches your stage, team, and support philosophy. This skill covers deep configuration across the support stack.

Authoritative Foundations

• Intercom — Conversational support, Fin AI, Product Tours — Conversational support, Fin AI, Product Tours
• Zendesk — Omnichannel CX, AI Agents, Explore analytics — Omnichannel CX, AI Agents, Explore analytics
• Front — Collaborative inbox, rule-based routing — Collaborative inbox, rule-based routing
• Help Scout — Docs-first support, Beacon widget — Docs-first support, Beacon widget

When to Use

Trigger phrases: "support platform setup", "Intercom configuration", "Zendesk deep setup", "support tool comparison", "AI support agent setup"

Platform selection: references/platform-comparison.md · SLAs: templates/sla-matrix.md

Platform Deep Configuration

Intercom — Best for Product-Led SaaS

Setup checklist:
1. Fin AI agent: train on 30+ help center articles → test 50 real questions
2. Messenger: contextual article suggestions before "talk to human"
3. Macros: 15+ for top ticket types (reset password, billing question, etc.)
4. Product Tours: onboarding flow (guided setup → first value)
5. Series: 3 automated email sequences (onboarding, engagement, expansion)
6. Reports: CSAT (target 4.2+), resolution time (target < 4hrs), volume

Zendesk — Best for Enterprise and Multi-Channel

Setup checklist:
1. Ticket fields: type, priority, product area, customer tier
2. SLA policies: P1 (15min FRT), P2 (1hr), P3 (4hr), P4 (8hr)
3. Triggers: auto-assign, auto-respond, escalate on keywords
4. Views: My Open, Unassigned, High Priority, All Open
5. Macros: 20+ for common responses
6. Guide (KB): 30+ articles published before launching chat
7. Explore: team performance dashboard, CSAT trends, volume by channel

Implementation Checklist

• Knowledge base: 30+ articles before launching chat/AI agent
• AI agent trained and tested with 50 real questions (95%+ correct)
• Macros: 15+ for top ticket types
• SLA policies documented with FRT + resolution targets
• CSAT survey: post-resolution, scores < 3 auto-escalated
• Escalation path: L1 → L2 → L3 documented with triggers

Common Pitfalls

1. AI agent launched without training data. 5 help articles → 70% wrong answers → frustrated customers. Fix: 30+ articles. Test with 50 questions.
2. Chat widget on every page. Noise. Distraction. Fix: Pricing, help center, and post-signup only.
3. No SLAs. "We respond quickly" = no commitment. Fix: Specific FRT targets by priority. Measured. Reported. Reviewed weekly.

Output Format

The agent delivers a support platform configuration guide matched to the user's tool and team stage:

• Platform Recommendation (when selection is requested): scoring across stage fit, team size, and support philosophy — with a 1-line rationale for Intercom vs. Zendesk vs. Front vs. Help Scout
• Platform Setup Checklist: prioritized configuration steps for the chosen tool — AI agent training (article count target + 50-question QA protocol), macros for top ticket types, SLA policy definitions, view configuration, and reporting setup
• Knowledge Base Plan: article count target (minimum 30 before launching chat), topic coverage map by product area, and Fin/AI training QA pass rate threshold (95%+ on test set)
• SLA Matrix: Priority tiers (P1–P4) with First Response Time target, resolution target, escalation trigger, and measurement method
• CSAT & Reporting Dashboard: metric targets (CSAT ≥ 4.2, resolution time by tier), review cadence, and auto-escalation rule for scores below threshold

Quality Check

Before delivering, verify:

• All required sections complete
• Output matches the user's stated need
• No vague or unsupported claims
• Frameworks cited where applicable

Execution Artifacts

• references/framework-notes.md — named frameworks, citation anchors, and operating assumptions
• templates/output-template.md — copy-paste deliverable structure for the user
• scripts/check-output.py — local checklist validator for required sections This skill includes lightweight artifacts the agent can load on demand:
• references/platform-comparison.md — Intercom vs Zendesk vs Front vs Help Scout
• templates/sla-matrix.md — P1–P4 FRT and resolution targets Use the artifacts when the user asks for an implementation-ready deliverable, a repeatable workflow, or a quality check rather than generic advice.

Related Skills

• support-tool-stack — Platform selection by stage
• headless-support — AI support agents and KB architecture
• sla-management — SLA design and escalation
• cs-analytics-dashboards — CS metrics and health scores