If Lavender Performs at 6% but Vanilla Needs 9%, Why Should I Use the Same Fragrance Percentage for Both?

What the uniform 9% costs, candle by candle

Fix the candle: 200 g (the series default; use your own) in Luxury Soy Wax CSI 464 at ₹507.40 a kilo, one Eco C2 wick, your own jar and box (labelled worked examples). The title's loads are yours: lavender passed at 6%, vanilla needed 9%. Fragrance is CSI's Lavender and Dark Vanilla at their 1 kg prices, ₹5.22 and ₹7.65 a gram. Load is a percentage of total weight: at 6% the candle is 12 g of oil and 188 g of wax; at 9%, 18 g and 182 g.

At 9%, fragrance is the biggest single line in the lavender candle, 34.7% of COGS. At 6% it drops to 25.8% and wax becomes the biggest. The difference is 6 g of oil worth ₹31.33, minus 6 g of extra wax worth ₹3.04: ₹28.29 a candle. In the vanilla candle nothing changes, because nothing should.

Why forcing one load hurts both candles

A uniform load has to be set by the oil that needs the most, or the candles that need more will fail. So the rule is really "every oil at vanilla's load". That has three costs on the lavender candle, only one of which is money.

1. Money: oil you pour and do not need

Covered above: ₹28.29 a candle at CSI prices. The saving is concentrated in the oil that sells and passes lower; how that sums across a whole range of eight or more oils is the range-wide per-oil load post.

2. Burn behaviour: a candle wicked for a different load

Higher loads can change how a wick behaves: the flame can get smaller or sootier, carbon can build up faster, and heavier oils such as vanilla commonly need a different wick from lighter herbals (tendencies, not rules). If your lavender has been wicked at 9%, the wick may be right for 9% and not for 6%, or the reverse. That is why a lower load is a new candle to re-check, not a free saving; whether vanilla and lavender in the same jar should share a wick at all is the vanilla-needs-a-larger-wick post, and the fragrance-chemistry side is the lavender-burns-well-but-vanilla-tunnels post.

3. Headroom: 9% sits close to some wax ceilings

CSI 464's page states "Fragrance Load Up to 10%". Your 9% of total weight is 9.89% of wax weight, inside the ceiling on either reading, but near it. Oil beyond what a wax holds tends to show up as sweating or pooling on the surface; at 9% in CSI 464 you are within the stated limit, but a lavender candle at 6% has far more margin for a hot storeroom or a slightly heavy pour. Never raise any oil above the lower of the wax's and the oil page's stated maximum to make a uniform rule work.

The formulation science behind "optimise each oil separately" is in the lavender-at-8%-but-vanilla-not post; this page owns the rupees and the bench.

The money over a year, and the compromise trap

At 200 lavender candles a month you stop buying ₹6,265.20 of lavender oil every month. That is cash, not a discount: you simply pour less. The vanilla line, at 150 candles a month in the worked example, is untouched.

The compromise trap: one "middle" load

Some makers try to keep one number by meeting in the middle, say 7.5% for both (15 g of oil). It looks like a saving and behaves like two problems. Lavender at 7.5% saves ₹14.14 a candle against 9%, leaving ₹14.15 on the table compared with its own 6%. Vanilla at 7.5% is below the load your test passed; it "saves" ₹21.42 of oil per candle and ships a candle you have not approved. A compromise load is still a uniform load, set by neither oil.

The case for one percentage, priced honestly

The best argument for a uniform load is operational. One number to remember, one line on the batch sheet, one way to train a helper. With two loads, someone will eventually pour the wrong one. The two mistakes are not equal:

Only the second mistake matters. Suppose it happens 4 times a year and you scrap every candle (worked example; most makers would catch it sooner). That is ₹25,153.60, and the lavender saving still leaves ₹42,742.40 a year. The lavender saving covers about 10.8 scrapped vanilla batches a year before the uniform rule wins on money. If your bench is so chaotic that you would lose more than that, fix the bench; that problem is costing you on more than fragrance.

There is one honest case for keeping a uniform load: very low volume. At 60 lavender candles a month the saving is ₹20,368.80 a year, and if your helper changes every season, the training cost may be real. Even then, a printed recipe card costs almost nothing.

When one load really is the right answer

A uniform load is correct in exactly one situation: when every oil's own test lands on the same number. If lavender, vanilla, coffee and rose all pass at 8% in your wax and jar, one card says 8% and you have lost nothing. The mistake is not having one number; it is choosing the number before the tests, or keeping it after the tests disagree. Two more cases deserve a pause before you split loads. First, if your 6% result came from a different wax lot, jar batch or wick from the one you produce on, it is not yet a result for this candle. Second, if lavender is a small seller and the saving per year is a few hundred rupees, spend the bench time on your biggest oil instead; the ranking logic for a whole range is in the range-wide per-oil load post. Everywhere else, the numbers favour two loads by a wide margin, and the risk that argues against them is one of the cheapest risks in the business to control: a card, a colour and a second pair of eyes on the scale.

The fix: one recipe card per SKU

A recipe card turns a percentage into weights. Nobody on the bench should calculate 6% of anything during a pour. Here is the format, filled in for the two worked SKUs at a batch of ten 200 g candles (2,000 g):

Three habits make the card work: the oil bottle and the card share a colour, the scale is tared in front of the second person, and the batch sheet records the load poured, not just the scent name. With twenty or more SKUs, the same discipline is what lets you change one oil's load without revalidating everything else; how much retesting each change needs is the quick-confirmation-versus-revalidation post.

Confirm the 6% before it ships

The title says lavender performs at 6%. Before you bank ₹67,896.00 a year, make sure that result came from a controlled comparison: 6% against a 9% control, everything else identical, judged blind, burned to the end. If it did, you are nearly done; if it came from one candle on one evening, run this first. Designing the ladder itself is the lower-load test design post, and what to retest when a formula is redesigned around a lower effective load is the 6%-instead-of-9% formula redesign post.

Cost of the confirmation: ₹469.89 of lavender, ₹563.21 of wax, ₹360.00 of jars, ₹56.64 of wicks and ₹600.00 of labour, about ₹2,049.74. At 200 lavender candles a month it repays in about 11 days; at 60 a month, in about 1.2 months. What the customer must notice: nothing. Same scent, same throw at the same distance, same burn time. If the blind test cannot tell 6% from 9%, neither can they.

The two oils, the wax and the wick rungs

The products in the worked candles, chosen on page statements, price and stock on 4 October 2026; CSI publishes no throw data, so nothing here is ranked on performance.

Where a bigger pack saves nothing: CSI 464 is flat per gram from 1 kg to 10 kg; C1 and C2 wicks are flat per wick at every pack; Lavender's 1 kg is barely cheaper than two 500 g packs. The saving on this page is grams you stop pouring, not a pack you start buying. Whether the dearer vanilla candle should sell at a different price from lavender is a pricing question for the fragrance-specific COGS and pricing post.

Frequently asked questions