Support Tool Stack

Support Tool Stack

Overview

Your support stack determines whether customers feel heard or abandoned. The mistake: picking the tool your last company used instead of the tool that matches your current stage, team size, and support philosophy. A 3-person startup on Zendesk Enterprise ($150/agent/mo) is burning money. A 50-person team on a shared Gmail inbox is losing customers to slow responses. This skill covers platform selection, setup, and optimization across every stage and support philosophy.

Authoritative Foundations

• Intercom — Conversational Support Framework — Conversational Support Framework
• Plain — API-first headless support and BYOAI via MCP — Model Context Protocol — tool servers for agent-safe CRM and enrichment access.
• Zendesk — Omnichannel CX Maturity Model — Omnichannel CX Maturity Model
• Help Scout — Support-Driven Growth (Bezos: customer obsession) — Support-Driven Growth (Bezos: customer obsession)
• Klaus/Maestro QA — Support Quality Framework — Support Quality Framework

When to Use

Trigger phrases: "choose support platform", "Intercom vs Zendesk", "help desk setup", "support tool stack", "migrate support platform", "set up ticketing", "customer support software", "Front vs Help Scout", "AI support tools", "help desk comparison", "Plain vs Intercom", "BYOAI support", "headless support stack"

Headless / BYOAI path

API-first stack (often with Attio CRM): Plain for support infrastructure — GraphQL API, embedded headless portal, Slack/email channels, native MCP for Jesse/Claude agents. BYOAI instead of vendor AI (Fin). Full pattern: skills/customer-success/headless-support/references/byoai-headless-stack.md. MCP config: mcp-setup.

Choose Intercom when you need Messenger + Fin without engineering a custom portal.

Platform Selection by Stage

Stage 0: Pre-Launch / $0-500K ARR (Founder Doing Support)

You need: Shared inbox. Basic knowledge base. Minimal cost.

Recommendation: Front if team has Slack workflow. Help Scout if docs-heavy. Plain if engineering team owns support UX and uses Attio-style composable stack.

Stage 1: $500K-3M ARR (First CS Hire)

You need: Ticketing, SLAs, basic reporting, knowledge base, chat widget.

Recommendation: Intercom for product-led SaaS. Help Scout for service-heavy.

Stage 2: $3-20M ARR (CS Team 2-10 people)

You need: Advanced routing, CSAT surveys, SLAs, integrations, analytics.

Recommendation: Intercom if product + support integrated. Zendesk if multi-channel (phone, social, email).

Stage 3: $20M+ ARR (CS Team 10+)

You need: Omnichannel, workforce management, QA, AI deflection, enterprise SSO.

Recommendation: Zendesk Enterprise. It's the standard for a reason.

Platform Deep Dives

Intercom — Best for Product-Led SaaS

Core modules:

• Inbox: Shared inbox with team assignments, collision detection
• Messenger: In-app chat widget with AI Fin agent (answers from help center)
• Articles: Knowledge base, public help center
• Product Tours: In-app onboarding flows, feature announcements
• Series: Email/chat automation for onboarding, engagement, expansion
• Tickets: Back-office ticketing for complex issues
• Reports: CSAT, volume, resolution time, team performance

Setup checklist:

1. Connect product events → auto-trigger relevant articles in Messenger
2. Set up Fin AI agent with help center content (3+ articles to activate)
3. Configure business hours, auto-reply for after-hours
4. Build 5-10 macros for top ticket types
5. Set up CSAT survey (post-resolution, 5-point scale)
6. Create saved replies library (50+ for common questions)
7. Set up assignment rules (round-robin or skill-based)
8. Integrate with CRM (HubSpot/Salesforce — see tickets in contact timeline)

Anti-patterns to avoid:

• Messenger on EVERY page (overwhelming — put on pricing, help, post-signup)
• Fin AI answering questions it can't handle (test thoroughly before launching)
• No macros (CS team retypes the same answers 50x/day)

Zendesk — Best for Enterprise and Multi-Channel

Core modules:

• Support: Ticketing, SLAs, macros, triggers, automations
• Guide: Help center, knowledge base, community forums
• Chat: Live chat widget, AI Answer Bot
• Talk: Voice/phone support with screen pops
• Explore: Analytics and reporting (pre-built dashboards)
• Sunshine: CRM platform (custom objects, workflows)

Setup checklist:

1. Define ticket fields (type, priority, product area, customer tier)
2. Set up SLA policies (first response time, resolution time by priority)
3. Create triggers (auto-assign, auto-respond, escalate based on keywords)
4. Build views (My Open Tickets, Unassigned, High Priority, All Open)
5. Set up macros (15-20 for common responses)
6. Configure business hours and schedules
7. Set up CSAT survey (automated post-resolution)
8. Create Guide help center with 20+ articles
9. Integrate with CRM, product analytics, and Slack
10. Set up Explore dashboards for team performance

Anti-patterns to avoid:

• Too many ticket fields (agents spend more time categorizing than solving)
• Overly complex triggers (hard to debug, unexpected side effects)
• No CSAT follow-up (negative scores without follow-up = churn risk)

Front — Best for Team Collaboration

Philosophy: Email-first, with internal comments and shared drafts replacing the "forward to colleague → they reply → forward back" dance.

Key features:

• Shared inbox with internal-only comments
• Collision detection (two people can't reply simultaneously)
• Rules engine for auto-tagging, assignment, routing
• Sequences for automated follow-ups
• Analytics: response time, resolution time, volume by channel

Best for: Teams that get most inquiries via email and need collaborative reply workflows. Not great for chat-first or phone-heavy support.

Help Scout — Best for Documentation-First Support

Philosophy: "If you do docs right, support volume drops." The Beacon widget surfaces relevant articles BEFORE the customer messages you.

Key features:

• Docs (knowledge base) tightly integrated with Inbox
• Beacon widget: contextual article suggestions in-app
• Saved replies, collision detection, workflows
• Lightweight CRM (customer profiles, conversation history)

Best for: Startups that want clean, affordable support without Intercom's complexity. Companies with strong documentation culture.

Plain — Best for API-First and BYOAI Support

Philosophy: API-first support infrastructure — build your portal, wire your agent. UI is optional; GraphQL API and MCP are primary interfaces.

Key features:

• Headless customer portal (embed in your product)
• Threads, customers, tenants, labels, help center
• Native MCP server for Jesse, Claude, ChatGPT
• Slack, email, in-app forms, live chat into one queue

Setup checklist:

1. Publish 30+ help center articles (same deflection bar as Fin)
2. Connect Plain MCP in agent IDE (mcp-setup — OAuth, read-first)
3. Configure draft → approve → send workflow for replyToThread
4. Embed headless portal or in-app form for ticket creation
5. Webhook → n8n → Attio/HubSpot for tier and account sync
6. Escalation labels for billing, security, enterprise (human-only)

Best for: Dev tools, API products, teams on Attio + Jesse already. Not a shortcut for non-technical CS — still need KB, SLAs, and human escalation.

AI Support Tools

Output Format

SUPPORT TOOL STACK — [Company]

Current Stage: [0/1/2/3]
Current ARR: $X M
CS Team Size: X

Selected Platform: [Intercom / Zendesk / Front / Help Scout / HubSpot Service]
Monthly Cost: $X/month ($X/seat × X seats)

Setup Checklist:
[Platform-specific setup items — 8-12 items]

Integrations:
- CRM: [HubSpot / Salesforce / Attio] — sync tickets to contact timeline
- Product: [Segment / Amplitude / Mixpanel] — see product usage alongside tickets
- Slack: ticket creation + alerts
- Knowledge Base: migrate existing docs to [platform] help center

Macros / Saved Replies (top 15):
1. [Type] — [response template]
2. ...

SLAs:
- First response: X hours (business hours)
- Resolution: X hours/days (by priority)
- P1 (critical): X min/hours
- P2 (high): X hours
- P3 (normal): X hours/days
- P4 (low): X days

CSAT Survey:
- Timing: [post-resolution / periodic]
- Scale: [5-point / CSAT / NPS]
- Follow-up: auto-escalate scores < X to manager

Implementation Checklist

• Platform matches stage (don't overbuy — Front at $19/seat may suffice for stage 0)
• SLAs defined with specific time targets, not "ASAP"
• 15+ macros/saved replies for top ticket types
• CSAT survey configured (post-resolution, not manual)
• Knowledge base has 20+ articles before launching chat widget
• AI agent tested with real questions before customer-facing
• CRM integration shows support tickets in contact timeline
• Business hours and auto-reply configured for after-hours
• Team trained on all macros, collision detection, and internal notes

Quality Check

Before delivering, verify:

• Output matches the user's stated request
• Named frameworks or sources are reflected in the recommendation
• The deliverable is specific enough for an agent to execute
• Any assumptions, risks, or dependencies are explicit
• No unsupported claims, invented facts, or private/internal references are included

Common Pitfalls

1. Overbuying for stage. Zendesk Enterprise at $115/seat for a 2-person CS team is lighting money on fire. Fix: Match platform to stage. Upgrade when you need the features, not before.

2. No macros at launch. CS team retyping the same 10 responses 50x/day is slow, inconsistent, and demoralizing. Fix: Build 15+ macros before go-live. Iterate weekly from actual tickets.

3. Chat widget on every page. Messenger/Beacon on every page = noise. Fix: Put on pricing page (pre-sales questions), help center (support), and post-signup (onboarding). Not on blog, homepage, or logged-out marketing pages.

4. AI agent without training data. Launching Fin or Answer Bot with 3 help articles produces wrong answers and angry customers. Fix: 20+ articles minimum. Test with 50+ real customer questions before launch.

5. No CSAT follow-up. Collecting "1/5" ratings without human follow-up sends the message: "We don't actually care." Fix: Auto-escalate scores < 3 to manager for personal follow-up within 24 hours.

6. Ignoring the knowledge base. Most support volume comes from the same 20 questions. If they're all documented, ticket volume drops 30-60%. Fix: Write articles from actual tickets. Every ticket that gets the same answer 3x becomes a help center article.

Execution Artifacts

• references/framework-notes.md — Named frameworks and reference tables
• templates/output-template.md — Deliverable shell for agent output
• scripts/check-output.py — Lightweight deliverable validator

Related Skills

• cs-playbooks — Onboarding, health scoring, CSQLs, churn intervention
• sla-management — SLA design, ticket routing, escalation, priority matrices
• cs-analytics-dashboards — CS metrics, NPS, CSAT, health scoring
• customer-onboarding — Structured onboarding, time-to-value, activation
• automation/tool-selection-stack — Stage-appropriate GTM tool stacks