Cold Throw vs Hot Throw — Why Can the Same Fragrance Perform Differently?

Two instruments, one oil

Start with the definitions, stated carefully, because most of the confusion in this topic is definitional rather than technical. Cold throw is the scent an unlit finished candle releases into the air around it. Hot throw is the scent a burning candle releases into the room it is in. Both are properties of a finished candle — not of a bottle, not of a wax, not of an oil in isolation — and both depend on the wax, the load, the vessel, the wick, the cure and the room, not only on the fragrance.

What the diagram above sets out is that the two share an oil and differ in everything else. The cold arrangement is patient and enclosed: a fixed surface, ambient temperature, a tiny volume, and weeks in which a concentration can build. The hot arrangement is urgent and open: a small growing surface, a much higher temperature, a room-sized volume, and only the hours of a burn. That asymmetry has one non-obvious consequence worth stating: cold throw is a stock and hot throw is a flow. When a customer lifts a lid they receive everything that collected while the jar was closed. When they light the candle they receive whatever it can produce per hour, continuously, against ventilation.

This is also why the two failures feel so different. A weak cold throw is noticed in a shop, at a stall, or in the first ten seconds of unboxing, and it costs you the sale. A weak hot throw is noticed at home, hours in, weeks later, and it costs you the reorder and sometimes the review. Same candle, two commercial exposures, two different remedies — and if your notes only contain one number per candle, you cannot tell which exposure you are failing.

Why a pyramid answers two questions

The top-heart-base convention is a commercial shorthand rather than a measurement, but it encodes something real: the order in which materials leave. Light, highly volatile materials go first and define an opening. Middle-weight materials define the body. Heavy, slow materials define what is still detectable much later. That ordering is why a single blend can produce two unrelated verdicts from your two instruments — and why the sentences a page uses about each tier are a usable forecast.

Read the table as a set of predictions, not results. In each case the question is: which tier is the page putting its strongest language on, and therefore which of your two columns is likely to score higher? Then go and find out, because a forecast that is never checked is just a preference.

Two honest limits on this method. First, five of the 107 pages state no notes or profile at all and 14 more state a single-note profile without tiers, so the forecast is simply unavailable for some oils — Sultan's Smoke, for instance, states a full pyramid but no candle load percentage, so its dose comes from your wax page. Second, a pyramid is written to describe a fragrance on skin or in the air, not specifically in wax at 8% released from a 28 cm² pool. It is the best free predictor you have. It is not a substitute for the sheet below.

The four outcomes, and what each one tells you to do

Once you are scoring in two columns, every candle you test lands in one of four boxes. The value of the framework is that each box has a different action attached — and two of the four are commercial decisions rather than reformulation problems.

The two-column test sheet

The design below compares three structures under identical conditions. Two jars per oil, because one jar is a story; a 120 g candle at 8% in a Green Shiny Glass Jar with Golden Lid, whose page states 120 gm capacity, 6 cm diameter and 6.5 cm height; one wax, one wick code, one pour session, one cure length. The only difference between the arms is the oil, which is the only way the result means anything.

Which lever moves which throw

The practical payoff of separating the two columns is that your levers are not symmetrical. Some changes move both readings, some move one and leave the other alone, and one of them moves cold throw while doing nothing whatsoever to the burn. Knowing which is which stops you spending on the wrong side of the problem.

Three oils across the structure spectrum

These are the three arms of the comparison above, chosen to sit at different points on the pyramid rather than to be ranked. Oil cost is for a 120 g candle at 8% — 9.6 g — at 1 kg pricing. Where a page states no load or no flash point, this guide says so.

Which reading your business should optimise

The honest answer depends on how you sell, and it is worth deciding deliberately rather than by accident. If most of your revenue comes from markets, stalls and exhibitions, the cold reading is doing the selling: a customer lifts a lid, decides in two seconds, and may never tell you what the burn was like. Optimising cold throw there is rational — cure lids-on, keep stock boxed, and favour structures with a vivid stated opening.

If most of your revenue comes from repeat orders, gifting and corporate, the hot reading is the whole business, because the buyer and the judge are the same person and the judgement happens at home. There the base of the pyramid, the pool width and the cure length are what you are actually selling, and a candle that wins the lid test and loses the evening will show up as a flat reorder rate months before anybody writes you a complaint.

Most Indian candle brands are somewhere in between, which argues for the same thing the two-column sheet argues for: run both columns and let the range split the job. A citrus-led scent can carry the stall and the summer collection; an anchored woody or gourmand can carry the gifting line and the Diwali range. Working back from Diwali on 8 November 2026, a batch that needs a full 21-day cure has to be poured by mid-October, and testing both columns before that date is what turns a guess about the festive line into a decision. If your hot column is weak across every arm, the cause is more likely to be delivery than fragrance, and the ordered diagnosis is in the six-check guide rather than in a new bottle.

Frequently asked questions