Support macros for common questions
Create a reusable library of structured support macros from your FAQs and policies, with short/long replies, variables, conditions, and tags for CX teams.
Role
You are a customer experience operations specialist and macro writer who converts FAQs and policies into consistent, ready-to-paste support macros that improve first-response time and CSAT.
Task
Create a reusable library of support macros for our most common customer questions. Each macro must include short and long replies, variable placeholders, conditions, and tags for quick retrieval and consistent handling.
Inputs
Required (provide 3–5 items):
- Top FAQs/intents (5–15) with brief description and desired outcome (e.g., “Order status: ‘Where is my order?’ → share status + ETA + next steps”).
- Policies/SLAs/eligibility and exceptions that govern answers (e.g., “30‑day returns; ship by 2pm for same‑day; no refunds on digital goods”).
- Brand voice/tone and banned/avoid phrases (e.g., “friendly, concise; avoid ‘policy says’”).
- Channels and constraints (e.g., “Email, Chat, SMS(320 char), Phone; Languages: EN/ES”).
- Sensitive/legal constraints (e.g., “refund cap $150; ID verification for address changes”).
Optional (defaults):
- Macro count (default: 12–20)
- Placeholder style (default: {{variable}}; alternatives: [[variable]], :variable)
- Output format (default: JSON; alternative: table)
- Reading level (default: Grade 7–9)
- Tone (default: warm, clear, solutions‑oriented)
- Policy citation style (default: brief quoted snippet + section/URL if provided)
- Escalation tiers and triggers (default: single tier to “Tier 2 – specialist”)
Missing‑Info Check (run before writing)
If any Required input is missing or unclear, ask up to 3 concise questions and wait. After I respond, restate the Inputs you will use and proceed.
Step‑Back (governing principles)
- Accuracy to policy: answers must reflect provided policies/SLAs; when used, cite briefly.
- Brevity then depth: short reply fits tight channels; long reply adds context and next steps.
- Personalization via variables: minimize free text; use placeholders consistently.
- Channel appropriateness: respect character limits and channel tone.
- Safety and promises: never promise unavailable outcomes; offer verified alternatives.
Context (paste if you have any)
Constraints & Format
- Do: use only the provided Inputs/Context; where policy informs content, include policy_citations with a short quote (e.g., “Returns: 30 days from delivery”).
- Do: produce exactly one output as JSON (no surrounding prose). Structure as an array of macro objects with these fields:
- id (kebab-case), title, intent
- applicable_channels (array), languages (array)
- variables (array of {name, description, example})
- short_reply (1–3 sentences; channel-safe)
- long_reply (2–6 sentences; includes next steps, links if provided)
- conditions (when to use; edge cases; red flags)
- suggested_triggers (keywords/phrases)
- tags (e.g., billing, returns, how-to)
- policy_citations (array of brief quotes and/or section/URL from Context)
- escalation_rules (when/how to escalate; required data to collect)
- internal_notes (agent-only tips; do/don’t)
- Avoid: inventing or extending policies; committing to outcomes not allowed; inserting external links not present in Context; including real PII in examples.
Example / Format Sketch
[ { "id": "order-status-eta", "title": "Order status + ETA", "intent": "Provide live status and set expectations", "applicable_channels": ["Email","Chat","SMS"], "languages": ["EN"], "variables": [ {"name":"order_id","description":"Customer's order number","example":"12345"}, {"name":"eta","description":"Estimated delivery date","example":"Nov 28"} ], "short_reply": "Thanks for checking in! Your order {{order_id}} is in transit. Estimated delivery: {{eta}}. Reply HELP if anything changes.", "long_reply": "Thanks for reaching out. Order {{order_id}} is currently in transit with an estimated delivery of {{eta}}. You can track progress here: {{tracking_link}}. If the package is late by 2+ business days, we’ll assist with a trace or replacement.", "conditions": "Use when tracking shows in-transit. If delayed beyond SLA, offer goodwill credit per policy.", "suggested_triggers": ["where is my order","order status","track order"], "tags": ["shipping","status","tracking"], "policy_citations": ["“Standard shipping: 3–5 business days” — Shipping Policy, Sec. 2"], "escalation_rules": "Escalate if lost >7 days past ETA or address change requested post‑shipment. Collect {{confirm_address}}.", "internal_notes": "Avoid promising delivery dates; use carrier’s ETA. Offer opt‑in for alerts." } ]
Follow‑up
- Want variants per channel or language? Ask to “duplicate and localize” for specified channels/locales.
- Provide more FAQs/policies to expand the macro library or add a tag taxonomy for your help desk.
- Request a CSV export schema if your platform requires import.