Skip to content

TODO-ANALYTICS.md

Latest commit

 

History

History
129 lines (102 loc) · 3.42 KB

TODO-ANALYTICS.md

File metadata and controls

129 lines (102 loc) · 3.42 KB
Raw

Analytics & Logging TODO

Current Blind Spot

Problem: We're getting organic traffic (75 images, 27 without marketing) but have no idea where it's coming from.

What we know:

  • 59 images (Feb 21 10:28 UTC) → 75 images (Feb 22 19:32 UTC)
  • 16 new organic requests in ~9 hours
  • Last activity: 27 minutes ago (20:00 UTC check)

What we DON'T know:

  • Referrer URLs (where did they come from?)
  • User agents (human vs bot?)
  • Which themes are popular?
  • Which titles/descriptions are being used?
  • Repeat users vs new users?
  • Geographic location?

Quick Wins (Add to server.js)

1. Basic Request Logging

app.get('/og', (req, res) => {
  const { title, description, theme } = req.query;
  const referrer = req.get('Referrer') || 'direct';
  const userAgent = req.get('User-Agent');
  
  console.log(JSON.stringify({
    timestamp: new Date().toISOString(),
    title: title?.substring(0, 50),
    theme,
    referrer,
    userAgent: userAgent?.substring(0, 100),
    ip: req.ip
  }));
  
  // ... existing image generation
});

2. Referrer Tracking

Log top referrers to understand traffic sources:

  • GitHub README views?
  • Search engines (Google, DuckDuckGo)?
  • Direct links shared on Discord/Slack?
  • Social media (Twitter, LinkedIn)?

3. Theme Popularity

Track which themes are actually being used:

const themeStats = {};
themeStats[theme] = (themeStats[theme] || 0) + 1;

This tells us:

  • Which themes to showcase in marketing
  • Which themes no one uses (maybe remove?)
  • User preferences (dark vs light, minimal vs bold)

4. Simple Analytics Dashboard

Create /admin/stats endpoint (protected by admin secret):

app.get('/admin/stats', requireAdmin, (req, res) => {
  res.json({
    total: stats.total,
    last24h: getLast24Hours(),
    topReferrers: getTopReferrers(),
    topThemes: getTopThemes(),
    recentRequests: getRecent(20)
  });
});

External Analytics Options

Option A: Plausible (privacy-friendly, $9/mo)

  • No cookies, GDPR compliant
  • Real-time dashboard
  • Referrer tracking, page views

Option B: Simple Analytics ($9/mo)

  • Similar to Plausible
  • Clean UI, privacy-focused

Option C: Google Analytics (free)

  • Most powerful, but overkill
  • Privacy concerns
  • Requires cookie consent

Option D: Self-hosted (Umami, free)

  • Privacy-first
  • Own your data
  • Requires setup/maintenance

Recommendation for now: Add basic logging to server.js (Option 1-3 above). Free, instant, sufficient.

Questions to Answer

Once we have analytics:

  1. Where is organic traffic coming from?

    • If GitHub → double down on README optimization
    • If search → focus on SEO content
    • If direct → someone's sharing the link (find where)

  2. What are people using SnapOG for?

    • Blog posts? (lots of text)
    • Product pages? (short, punchy titles)
    • Social sharing? (varied content)

  3. Free tier conversion potential?

    • How many unique users?
    • How many requests per user?
    • Are power users hitting the 50/hour limit?

  4. Best time for Reddit post?

    • If we see traffic spikes at certain hours → post then
    • If weekends vs weekdays matter → time accordingly

Next Steps

  1. Add basic logging to server.js (1 hour)
  2. Monitor for 48 hours
  3. Analyze traffic sources
  4. Optimize based on data
  5. Add external analytics if needed

Key insight: Can't optimize what we can't measure. Need visibility into organic discovery channels ASAP.