How to Test and Approve a New Fragrance for Your Range

What “approved” has to mean before it means anything

Most ranges grow by accident. An oil arrives, somebody pours two candles, everyone in the room says it smells lovely, and three weeks later it is on the website with a photograph and a price. Then it underperforms — weak when lit, or it yellows the white wax, or it needs so much oil that the candle cannot be sold at your price point — and nobody can say what changed, because nothing was ever fixed in the first place.

An approval is a fixed test, a fixed pass mark and a written record. Fixed test means the same wax, the same wick, the same jar, the same load, the same cure and the same room every time you judge a candidate, so that the only variable is the oil. Fixed pass mark means you wrote down what good enough is before you smelled it. Written record means that when the same oil comes back in a different pack six months later, you are comparing it against a number rather than a memory.

The gates in this page are ordered by what a rejection costs. Reading the product page costs nothing. A 15 gm bottle costs ₹93.22 for Apple Cinnamon. Three votives cost ₹536.90. Three candles in your real jar cost ₹876.74. A ten-candle confirmation batch takes the total to ₹4,473.38. A production order for one festive scent is ₹9,487.20 of oil before wax, wicks or jars. Every gate you skip moves the rejection downstream, where it is dearer.

Gate 1: the paper check that costs nothing

Four published facts decide whether a candidate is even testable in your formula. The candle usage range: most CSI oil pages print one — Sandalwood gives up to 10% of total wax weight for soy, Apple Cinnamon gives soy 6–10%, Japanese Cherry Blossom gives 6–13%, Gilli Mitti Patchouli gives 6–10% across wax types. Where a page does not print one, it is not published — ask on WhatsApp rather than guessing, and do not put an unpublished number on your own batch sheet.

The two ceilings, and the lower one governs. Your wax has a stated maximum and so does the oil. On CSI 464, whose page states up to 10%, a 6–13% oil is capped at 10% — the extra headroom on the oil page is unusable until you change wax. On Luxury Soy Wax Chunks, which states up to 13% with 8% as a starting point and 10% recommended, the same oil could be tested higher. Write the governing figure on the test sheet as a number, not as a range, so the person pouring does not have to interpret it.

Vanillin and discolouration. Pages that carry a figure are telling you something useful about your white candles: British Rose states 0% vanillin and no discolouration, Sandalwood states 0% vanillin, Cinnamon Vanilla states 2.5% vanillin with moderate discolouration or yellowing. Yellowing is cosmetic, not a fault, but it is a decision: if the candidate is going into a white wax gift set photographed for Instagram, it belongs on the appearance row of your score sheet before it is poured, not after a customer mentions it.

Flash point, read correctly. It is a transport, storage and handling property. It does not rank hot throw and it is not the temperature at which fragrance burns off. The usual 80–90°C add window sits below the flash points CSI pages print where they print them — Sandalwood 203°C, British Rose 192°C — but a few are much lower, and for those the sensible instruction is to add at the cool end of the window and not to hold the wax hot. Where no flash point is published, ask; do not estimate one for a safety document.

Gates 2 and 3: the bottle screen and the votive pour

Start with the bottle. A 15 gm pack — ₹68.44 to about ₹216.00 depending on the oil — answers one question: is this the scent the name and the published family led you to expect. It is a surprisingly good filter. Oils are named for what they evoke, and a page that publishes a family and notes is describing a composition, not a single material; “sandalwood” from two houses can sit in two different places. If the bottle is wrong, you have spent under ₹100.00 and saved a fortnight.

Do not, however, judge strength from the bottle. Neat oil on a blotter tells you almost nothing about how much of that molecule will leave a soy melt pool three weeks later, and it tells you nothing at all about how it behaves in a different medium. One CSI review makes that point unintentionally: a maker who liked a fragrance reported weak throw from a reed diffuser formula built on solvents rather than wax. That is a different physics problem from a candle, and a good candle oil can read flat in the wrong carrier.

Gate 3 is the screen pour: three 100 g votives at your standard wax, wick and load. Three, not one, because one jar tells you nothing about variation. At 8% each votive takes 8 g, so a 50 gm pack covers the round with room to spare, and the whole thing costs ₹536.90 with Clear Glass Votive Jars at ₹31.86 each — 38.8% less than the production round. Its job is to catch the failures that only appear in wax: separation or beading on the surface, a scent that turns, a colour shift, or a throw so faint that no wick will rescue it.

Gate 4: the blind panel and the score out of 18

Six criteria, scored 0 to 3 by each of three panellists, gives a maximum of 18 per nose and 54 for the panel. The criteria are cold throw at day 14 with the lid off at arm's length, hot throw at 1 hour, hot throw at 3 hours, room fill at 3 hours judged from the doorway, appearance (surface finish and any colour shift against the wax's own colour) and burn behaviour (pool reaching the wall by about 3 hours, steady flame, no soot). Score on entering the room, not after five minutes in it.

The pass rule has three parts, and all three have to hold. Each panellist scores 12 or more out of 18. Nobody scores 0 on hot throw at 3 hours — that single row is what a customer buys the candle for. And the spread between the highest and lowest panellist is no more than 3 points; a wider spread means one nose is reading something the others are not, and the honest response is to burn the second jar with a fourth person rather than to average the disagreement away.

Read the rows, not just the total. A candidate that is strong cold and weak hot is a shelf-appeal scent that will disappoint on burning. A candidate strong at 1 hour and fading by 3 is a top-note-led composition that may need a heavier base in the range around it. Weak room fill with good hot throw at arm's length is a small-room scent. None of those are failures in themselves; they are instructions about which product the oil belongs in, and they are the reason to keep the sheet.

Gate 5: the cost gate most candidates fail

Once a candidate has a score, cost it at the load that actually performed — not at the page's minimum, and not at the load you hoped for. At the series unit of 200 g of wax, every percentage point is 2 g of oil, so the arithmetic is direct: oil cost per candle = grams × the per-gram rate of the pack size you will really buy. Use 1 kg pricing if you will buy kilos and 100 gm pricing if you will not, because the per-gram gap between those is the difference between a viable scent and a vanity one.

The 13% on the Japanese Cherry Blossom page is a good worked example of the two-ceiling rule doing real financial work. On CSI 464 you cannot use it, so the oil costs you at most ₹96.84 a candle. Move the same scent onto Luxury Soy Wax Chunks, whose page states up to 13%, and the ceiling rises to 26 g — ₹125.89 a candle — which may be the right decision for a hero product and the wrong one for your everyday line. The oil did not change. The governing ceiling did.

Then apply the load sensitivity to the volume you actually pour. On Sandalwood at 1 kg pricing, moving a candle from 6% to 10% adds ₹40.21; across a 400-candle run that is ₹16,085.76. If the panel says the candidate only works at the top of its range, that figure is part of the approval decision, not an implementation detail to be discovered in November.

The go/no-go sheet, and what a “no” is worth

Put the decision on one page with six lines and a verdict: paper gate (governing load, vanillin note, flash point, documentation available), bottle gate (is it the scent expected), screen result, production-round panel score, oil cost per candle at the approved load, and range fit. Four verdicts are available, and only one of them is “no”.

The range-fit line is where disciplined brands say no most often. If the candidate scores 14 and your existing warm-spice bestseller scores 15, you have not found a new product, you have found a second version of one you already sell — and two similar scents on a shelf split the same demand while doubling your stock, your testing and your dead inventory. Approve into a gap, not into a crowd.

Keep rejections. One line each — date, oil, governing load, panel score, reason — turns into the most useful page in your file within a year: it stops you re-testing the same candidate because a different photograph caught your eye, and it tells you what your panel systematically dislikes, which is usually more informative than what it likes.

When an approved fragrance has to be approved again

Four changes invalidate an approval, and all four are things brands do casually. Changing the wax grade changes the ceiling and the release: moving from CSI 464 at a stated 10% to Premium Coconut Soy CSI 468 at a stated 13% also changes the melting point and the pour temperature the page gives, so the same oil at the same percentage is a new candle. Changing the jar changes pool diameter and therefore hot throw. Changing the wick changes the pool as well. And changing the load — even downward, to save money — is a new formula that has never been scored.

A re-approval does not need the whole ladder. Pour three jars in the new combination, cure 14 days, score against the retained reference from the original approval, and accept the change only if the score holds within 2 points. That is ₹876.74 and a fortnight to protect a product customers already recognise, which is the cheapest insurance in this article.

Keep one sealed jar from every approved combination, labelled with the batch code and the date, and keep about 15 g of the oil itself in a small sealed bottle out of the light. When a later pack of the same oil smells different to you, that retained pair — the candle and the neat oil — is the only way to tell whether the oil moved, your nose moved or your process moved. Ask for documentation on WhatsApp at the same time and file it with the sheet; a corporate buyer will ask for it long after you have forgotten which scent it belonged to.

Frequently asked questions