Skip to content

Latest commit

 

History

History
168 lines (119 loc) · 5.01 KB

SDK_REAL_USAGE_DATA.md

File metadata and controls

168 lines (119 loc) · 5.01 KB

Python SDK Real Usage Data

Query Date: 2025-11-25 Data Source: Production PostgreSQL database (api_requests table)

Summary

  • 113 API requests made via Python SDK (User-Agent: OilPriceAPI-Python/1.0.0)
  • 4 unique users actively using the SDK
  • 4 active days with SDK usage
  • First request: 2025-09-29 (launch day)
  • Last request: 2025-11-25 (today)

PyPI vs API Usage Comparison

PyPI Statistics (from pypistats)

  • 763 total downloads (with mirrors)
  • 256 real downloads (without mirrors)
  • Download pattern: Launch spike (Sep 29-30), steady 4-9/day

API Usage (from production database)

  • 113 API requests from Python SDK
  • 4 unique users making requests
  • Download-to-usage conversion: 4/256 = 1.56% activation rate

Most Popular Endpoints

Endpoint Request Count Unique Users
/v1/prices/past_year 77 2
/v1/prices/latest 36 4

Insight: Historical data (past_year) is more popular than latest prices.

Daily Usage Pattern

Date Requests Unique Users
2025-09-29 17 1
2025-10-02 10 2
2025-10-07 67 1
2025-11-25 19 1

Insight:

  • Sep 29: Launch day testing (17 requests)
  • Oct 2: Second user joined (2 unique users)
  • Oct 7: Heavy usage day (67 requests, likely batch processing)
  • Nov 25: Current testing (19 requests)

Performance Metrics

  • Average response time: 17,812ms (17.8 seconds)
  • Note: This is unusually high, likely due to the /past_year endpoint fetching large datasets

Key Findings

✅ Positive Signs

  1. Real users exist: 4 unique users (not just me)
  2. Production usage: 67 requests in one day suggests batch processing use case
  3. Historical data demand: /past_year endpoint is most popular (77 requests)
  4. Consistent activation: Users who download it are actually using it

⚠️ Areas for Improvement

  1. Low activation rate: Only 1.56% of downloaders become active users
  2. High response times: 17.8s average suggests performance issues or large datasets
  3. Sparse usage: Only 4 days of activity in 57 days since launch
  4. Small user base: 4 users is very early stage

What to Include in Reddit Post

✅ CLAIM THIS (Accurate & Verifiable):

"Early adoption: 250+ PyPI downloads since September launch, with 4 active users making 100+ API requests. Most popular: historical price data (past_year endpoint). Looking for feedback to improve it."

Why this works:

  • Honest numbers (verifiable in PyPI + production DB)
  • Shows real usage (not just downloads)
  • Shows what people actually use (historical data)
  • Invites engagement ("looking for feedback")

❌ DON'T CLAIM:

  • "Used by X companies" (only 4 users, likely individuals)
  • "Processing Y requests/day" (only 4 days of activity)
  • "Proven in production at scale" (113 requests total is modest)

Comparison: PyPI Downloads vs API Usage

Problem: 256 real PyPI downloads, but only 4 active users

Possible reasons:

  1. Testing/evaluation: People download to evaluate, but don't use yet
  2. No API key: Downloaded SDK but didn't sign up for API key
  3. Not needed yet: Downloaded for future project
  4. Bad first experience: Downloaded, tried, hit error, gave up

Action: Reddit post should invite feedback on onboarding experience

Updated Recommendation for Reddit Post

Line 66 Update (Current):

**Early adoption:** 20+ downloads in the past month from developers testing in production. Looking for feedback to improve it.

Line 66 Update (Recommended):

**Early adoption:** 250+ PyPI downloads since September, 4 active users making 100+ API requests. Most popular feature: historical price data. Looking for feedback on what to improve.

Why better:

  • More specific (250+ vs 20+)
  • Shows actual usage (4 active users, 100+ requests)
  • Shows what users want (historical data)
  • Lower pressure (4 users is honest, not inflated)

Action Items

  1. Update Reddit post with accurate stats (250+ downloads, 4 active users)
  2. Investigate performance - Why is avg response time 17.8 seconds?
  3. Improve onboarding - Why only 1.56% activation rate?
  4. Add caching - /past_year is popular, should be cached
  5. Monitor post-Reddit spike - Expect 5-10x download increase

Success Metrics to Track Post-Reddit

Before Reddit:

  • PyPI downloads: 256 real downloads
  • Active users: 4
  • API requests: 113 total

Target After Reddit (1 week):

  • PyPI downloads: 500+ real downloads (2x)
  • Active users: 15+ (4x)
  • API requests: 500+ total (5x)

Target After Reddit (1 month):

  • PyPI downloads: 1,000+ real downloads
  • Active users: 50+
  • API requests: 2,000+ total
  • Community contributions: 3+ PRs/issues

Status: ✅ Real data collected from production Confidence: High - all numbers verifiable Next Step: Update REDDIT_POST_FINAL.md with accurate statistics