The Documentation Problem
Your reality:- 📧 Same questions every week
- 🔁 Writing same answers repeatedly
- 📚 No time to create help docs
- ❓ Customers can’t find answers (if docs exist)
- 😰 Ticket volume growing
What You’ll Achieve
With BuildBetter:- ✅ Identify most common questions from support history
- ✅ Find best answers you’ve already written
- ✅ Generate help article drafts automatically
- ✅ Prioritize documentation by ticket volume
- ✅ Reduce ticket volume 15-25% with targeted docs
Time investment: 2-3 hours to create your first 5 help articles. Impact: Each article prevents 5-20 tickets/month.
Prerequisites
BuildBetter account
Support ticket history or call recordings with customer questions
Place to publish docs (help center, Notion, Confluence, etc.)
Step 1: Find Your Most Asked Questions
[screenshot: Signals filtered for questions]1
Import Support Data
If you haven’t already:
- Connect your support platform (Zendesk, Intercom, etc.)
- OR upload support tickets as CSV
- See Reduce Support Tickets guide for details
2
Filter for Questions
- Go to Signals
- Filter by Type: Question
- Sort by frequency or count
- “How do I export to Excel?” (47 tickets)
- “Where is the API key?” (32 tickets)
- “Can I change my email address?” (28 tickets)
3
Ask Chat for Top Questions
Query: “What are the top 10 most frequently asked customer questions? Group similar questions together.”[screenshot: Chat showing top questions with counts]Example output:
Step 2: Extract Your Best Answers
[screenshot: Support ticket with good answer highlighted as signal]1
Find Successful Resolutions
For each common question:
- Click on the question signal
- View all instances across tickets
- Look for answers that:
- Resolved the issue quickly
- Customer said “thank you” or “that worked”
- Are clear and complete
2
Generate Article Draft
Query Chat: “Create a help article for [question topic]. Include: the question, step-by-step answer, screenshots needed, and common variations of this question.”[screenshot: Chat generating help article]Example output:
3
Add Visual References
For each “screenshot needed”:
- Take actual screenshot from your app
- Highlight the button/area with arrows
- Add to your help center
Step 3: Prioritize Your Documentation
[screenshot: Documentation priority matrix]1
Create Priority List
Query: “For each common question, tell me: ticket volume, deflection potential (% of tickets we could prevent), and effort to document (low/medium/high).”Example output:
2
Start with Quick Wins
Focus on:
- High volume questions
- Questions with clear, existing answers
- Low effort to document
Step 4: Publish & Measure
[screenshot: Help center with new articles published]1
Publish to Help Center
- Copy article content to your help center
- Add screenshots
- Tag with keywords for search
- Link from related articles
2
Add In-App Links
Even better: Link to docs from your app
- Add tooltip near export button → “How to export”
- Show help article in error messages
- Contextual help where confusion happens
3
Track Impact
After 30 days:Query: “Are we getting fewer tickets about export functionality since we published the help article?”[screenshot: Ticket trend showing decline after doc published]Measure:
- Ticket volume for that topic (should decrease)
- Help article views (should increase)
- Customer self-service rate
Real Example: Taylor’s Documentation Win
Background: Taylor is solo support at 40-person startup. 300 tickets/month. No help docs. Week 1: Import 6 months of tickets to BuildBetter Analysis:- Top question: “How do I add team members?” (38 tickets)
- Second: “How to integrate with Slack?” (29 tickets)
- Third: “Export doesn’t work” (24 tickets)
- Used Chat to generate drafts
- Added screenshots
- Published to Notion (their help center)
- Added links in app UI
- “How to change plan/billing”
- “Reset password troubleshooting”
- Tickets: 300/month → 235/month (-22%)
- Top 3 questions: 91 tickets → 21 tickets (-77% on those topics)
- Time saved: ~15 hours/month
- Customer satisfaction: Up (faster self-service)
- Was 38 tickets in 6 months
- Now: 2-3 tickets/month (95% deflection)
- Article views: 450/month (helping customers proactively)
Best Practices
Start with top 5: Don’t try to document everything. Focus on highest volume questions.
Use actual customer language: Title articles how customers ask, not how you’d describe it.
Include variations: “How do I…?” + “Can I…?” + “Where is…?” all in same article for SEO
Add screenshots: Visual guides work 3x better than text-only
Link from app: Contextual help beats searchable help docs
Update regularly: When you fix a bug or change UI, update docs immediately
Common Questions
What if customers still ask even with docs available?
What if customers still ask even with docs available?
Common reasons:
- Can’t find docs: Improve SEO, add in-app links
- Docs unclear: Rewrite with more screenshots/steps
- Don’t want to read: Add video walkthrough
- Problem is different: Update article with edge cases
Should I document edge cases or just common scenarios?
Should I document edge cases or just common scenarios?
Start with common (80% of tickets).Add edge cases later if:
- Still getting tickets about it
- High-value customers affected
- Easy to document
How do I keep docs up to date?
How do I keep docs up to date?
When product changes:
- Update docs BEFORE releasing feature
- Review related support tickets after release
- Update docs based on new questions
What if answer requires engineering fix, not docs?
What if answer requires engineering fix, not docs?
Flag it for product/engineering:
- “Export fails” → Not a docs problem, it’s a bug
- “Can’t find export button” → Docs can help
Next Steps
Reduce Support Tickets
Full support ticket deflection strategy
Update Existing Docs
Keep documentation current as product evolves
Customer Health Tracking
Proactive customer success workflows
Knowledge Base Automation
Auto-generate docs from patterns
Stop answering the same questions. Write it once, help hundreds.
Every repeated support question is an opportunity to create a help article. Document your top 10 questions and watch ticket volume drop 20-30%.
Based on analysis of 300+ support teams. Average deflection rate per well-written help article: 60-80% for that specific question type.