The Cure
Why the same candle smells different on day one and day fourteen.
A candle on day one is not the candle you made. The fragrance molecules need time to bind into the crystalline structure of the wax, and until that has happened you are burning a wet blanket.
I have thrown out perfectly good recipes because I judged them at 24 hours and thought the throw was weak. It was not weak, it was young.
The cure calendar
Cure times depend on wax and fragrance, but a workable calendar looks like this. Day one to three, ignore. Day four to seven, first test burn allowed for wick calibration only, do not judge the scent. Day fourteen, first honest scent test. Day twenty-one, the candle is representative of what it will be.
Paraffin cures faster, around 7 days. Coconut blends cure in about 10 days. Soy takes the full 14 to 21. Beeswax barely needs a cure at all, but it barely throws either, so that is a different problem.
The one-hour-per-inch rule
When you finally do the first burn, let the melt pool reach the edge of the vessel before you extinguish. The rule of thumb is one hour of burn per inch (25 mm) of vessel diameter. A three-inch (75 mm) jar wants three hours on its first burn.
This is where most makers accidentally condemn a good candle. They light it, burn it for 45 minutes, blow it out, and then wonder why it tunnels for the rest of its life. The first burn is not a taste test, it is calibration.
What honest looks like
The good-pour-versus-defects reference on this spread is not aspirational. Those are candles I made in a two-month run when I was tuning a coconut-soy blend for a wholesaler. Some are close to perfect. Others are informative in a way that only failure can be.
By the end of Act 1 you have poured a candle, read its defects, cured it fairly, and made your first honest judgement. Act 2 turns that single candle into a range you can repeat.
Repeatable beats spectacular. A candle you can pour ten times is a business. A candle you nailed once is a story.