Start Here

New to Candlegraph? Here’s your guide to navigating the site.

What is Candlegraph?

Candlegraph is a data-driven candle review site. Instead of just describing scents, I measure actual performance—burn time, cost efficiency, and throw potency—to calculate objective value indexes (SAV and PAE) for each brand.

Think of it as “candle reviews meet data science.”

First Time Visitors

Start with these three posts:

Welcome to Candlegraph - Quick intro to what this site is about (2 min read)
How Candlegraph Started - The journey from Excel hobby to systematic performance analysis (8 min read)
Methodology - How the SAV and PAE indexes work, explained with actual math (10 min read)

After those, you’ll understand the performance map and what the numbers mean in reviews.

Looking for Reviews?

Published Reviews:

Diptyque Narguile - Luxury tobacco and honey candle
View all reviews - Complete archive

Understanding the Metrics:

SAV Index - Scent Adjusted Value (throw potency relative to collection average)
PAE Index - Price-Adjusted Efficiency (burn hours per dollar relative to collection average)
100 = Collection Average - Higher is better, lower is worse

Check the methodology page for the full formulas.

Interested in the Technical Side?

Behind the Scenes:

Playing with R - My first attempts at data visualization
View all technical posts - R code, data pipelines, learning journey

The Stack:

Google Sheets → R (data_sync.R) → Hugo (static site) → Netlify
Performance indexes calculated from live data
All code available in reviews

By Content Type:

Reviews - Systematic brand performance analysis
Technical Posts - Data science, R tutorials, project notes
About Candlegraph - Origin story, project updates

By Brand:

Diptyque
More brands coming soon…

Full Archive:

All posts - Complete chronological list

What’s Next?

I’m actively burning candles and building the dataset. Expect new reviews every few weeks as candles finish burning and data is collected.

Currently measuring:

Erewhon - Smoked Oud & Incense

Coming soon:

Interactive Tableau performance map
More luxury brand reviews
Deep dive on the “hot throw measurement problem”

Questions?

“Why these specific brands?” - Mix of what I’ve purchased (supporting local/small businesses) and what I’m curious about testing
“How long does it take to review a candle?” - 30-50 hours of burn time spread across multiple sessions, plus data analysis
“Can I request a brand?” - Not currently taking requests, but I’m always open to suggestions!