Q&A Question Discovery

This drill finds high-opportunity questions across Stack Overflow, Quora, Dev.to, and other Q&A platforms where your expertise can produce standout answers. It produces a ranked queue of questions to answer, prioritized by visibility, relevance, and competition.

Input

• ICP definition: what problems your product solves, for whom, in what context
• Product expertise areas: specific topics, technologies, or domains you can speak on authoritatively
• Competitor names: so you can find "vs" or "alternative to" questions
• Target platforms and tags (e.g., Stack Overflow tags: python, api, rate-limiting)

Steps

1. Define your tag and keyword maps

Build three keyword maps:

Tag map (Stack Overflow / Dev.to):

• Primary tags: 3-5 tags directly in your product domain (e.g., javascript, react, testing)
• Adjacent tags: 5-10 tags where your ICP hangs out but your product is not the direct topic (e.g., node.js, typescript, ci-cd)
• Use the qa-platform-api-read fundamental to pull related tags:

  GET /tags/PRIMARY_TAG/related?site=stackoverflow
  

Keyword map (Quora / general search):

• Pain-point queries: "how to solve X", "why does X happen", "best way to handle X"
• Comparison queries: "X vs Y", "alternative to X", "should I use X or Y"
• Recommendation queries: "best tool for X", "what do you use for X"

Competitor map:

• "[competitor] problems", "[competitor] not working", "switch from [competitor]"

2. Pull candidate questions from Stack Overflow

Using the qa-platform-api-read fundamental, run these queries for each primary and adjacent tag:

# Unanswered questions (highest priority)
GET /questions/unanswered?tagged=TAG&sort=creation&site=stackoverflow&pagesize=50

# Questions with no accepted answer
GET /search/advanced?tagged=TAG&accepted=false&sort=creation&site=stackoverflow&pagesize=50

# Questions with low answer counts
GET /search/advanced?tagged=TAG&answers=1&accepted=false&sort=votes&site=stackoverflow&pagesize=50

Also search for keyword-based matches:

GET /search/advanced?q=KEYWORD&sort=creation&site=stackoverflow&pagesize=50

Collect all unique question IDs. Target: 50-100 candidate questions per discovery run.

3. Pull candidate questions from Quora

Using the qa-platform-api-read fundamental's SerpAPI approach:

# For each keyword in your keyword map
curl "https://serpapi.com/search.json?engine=google&q=site:quora.com+KEYWORD&tbs=qdr:w&api_key=KEY&num=20"

The tbs=qdr:w restricts to the last week. Parse Quora question URLs and titles from results.

4. Pull candidate questions from Dev.to

# Recent help/discuss posts
curl "https://dev.to/api/articles?tag=help&top=7&per_page=30"
curl "https://dev.to/api/articles?tag=discuss&top=7&per_page=30"

Filter by titles that contain your keywords.

5. Score and rank all candidates

For each candidate question, compute a priority score (0-100):

| Signal | Weight | Scoring | |--------|--------|---------| | No accepted answer | 25% | 100 if no accepted answer, 20 if accepted but stale | | View count | 20% | Normalize: top 10% of views = 100, bottom 10% = 10 | | Recency | 20% | Posted < 6h = 100, < 24h = 80, < 72h = 50, < 1w = 20, older = 5 | | Low competition | 15% | 0 answers = 100, 1 answer = 70, 2 = 40, 3+ = 10 | | Tag relevance | 10% | Primary tag match = 100, adjacent = 50, keyword-only = 30 | | Asker reputation | 10% | High-rep asker = more visible question. > 1000 rep = 100, > 100 = 60, < 100 = 30 |

Sort by total score descending. Take the top 10-20 for the answer queue.

6. Build the answer queue

For each selected question, create a queue entry:

{
  "platform": "stackoverflow",
  "question_id": "12345",
  "url": "https://stackoverflow.com/questions/12345/...",
  "title": "How to handle rate limiting in Node.js?",
  "tags": ["node.js", "api", "rate-limiting"],
  "score": 87,
  "priority": "high",
  "answer_count": 0,
  "view_count": 1500,
  "created_at": "2026-03-30T10:00:00Z",
  "your_angle": "We built a rate limiter; can share the pattern without promoting product",
  "estimated_effort": "15 minutes",
  "status": "pending"
}

Store the queue in Attio using the attio-lists fundamental: create a list named "Q&A Answer Queue" with structured fields for each property above.

7. Log discovery run metadata

Using attio-notes, log each discovery run:

• Date, platforms searched, tags and keywords used
• Total candidates found, total after scoring
• Top 10 selected questions with scores
• Any new tags or keywords discovered during the search

Output

• Ranked queue of 10-20 questions to answer across all platforms
• Attio list with queue entries and metadata
• Discovery run log for tracking coverage over time

Triggers

• Run daily during Smoke and Baseline levels
• Run every 4 hours during Scalable level (automated via n8n)
• Feed into qa-answer-crafting drill for response generation