I Used 8% Fragrance Oil but My Candle Still Smells Weak — What Should I Change?

Week 0: is your 8% really 8%, and where does it sit?

"I used 8%" can mean three different candles. Load is a percentage of the finished candle's total weight: a 150 g candle at 8% holds 12 g of oil and 138 g of wax. If you melted 150 g of wax and added 12 g of oil, your candle weighs 162 g and the true load is 7.4%. If you measured 12 ml with a spoon, you do not know the grams at all, because oils differ in density. Neither slip explains a candle that smells of nothing, but both make every later comparison unreliable. Weigh both on a scale, total-weight basis, from now on.

Where does 8% sit on your oil page?

Of the 102 oils in CSI's range, 50 pages state a candle load or range in some form (two of them, Lavender Oud and Amber Oud Royal, only as "Candles (Soy/Paraffin/Blends) 10%" in a perfume usage table). Where 8% falls inside that range decides whether load is even a candidate change.

Then the wax. The lower of the wax page's and the oil page's maximum governs: Luxury Soy Wax CSI 464 states up to 10%, Luxury Soy Wax Chunks and CSI 468 up to 13%; other waxes' pages do not state one in the text we hold. These are formulation limits, not regulatory ones, and the maximum is not the best load. For most readers the audit ends with one conclusion: at 8% the oil already has a fair load, so the next change is not more oil. How to choose between 6%, 8% and 10% from scratch is a separate question, answered in Should I Use 6%, 8% or 10% Fragrance Oil for Strong Soy Candles?.

The rules of the change log

A change log is a plain record of what you changed, what you held still, and what happened. It turns "I tried everything" into evidence. Five rules make it work.

This is deliberately narrower than a full isolation grid; that protocol, with control candles across every variable, is in How to Test Whether Weak Hot Throw Comes From the Oil, Wax, Wick or Process. The change log suits a maker who already has a working recipe that is slightly off, not one starting from nothing.

Weeks 1 to 4: one change each

Week 1 — the wick (pour 28 September; verdict 12 October)

Pour two 150 g candles at 8% from one melt: the reference with your current wick, the change with one size up (C1 to C2 if you use Eco wicks). Centre both. At day 7 and 14, log the pool at hours 2, 3 and 4. The winner is the wick whose pool reaches the glass by hour 3 to 4 and is roughly 0.5 to 1 cm deep, without soot or a tall flame; these benchmarks are commonly cited rules of thumb, not CSI data. At 8% an under-sized wick is the commonest reason, in our judgement, for a candle that smells fine cold and faint lit. If the jar is wider than about 8 cm, a C2 may still fall short; the Cotton Braided Wicks Ready Made page says they are "engineered for large-diameter candles exceeding 8 cm".

Week 2 — the cure (no pour; judge on 5, 12 and 19 October)

The second change costs nothing: judge later. Take the week-1 reference and burn it at day 7, day 14 and day 21 instead of the day 2 or 3 many makers use; do the same with any existing candle from your weak batch that you can date. Several CSI oil pages state a soy cure of 48 to 72 hours (Palo Santo's is one); treat that as the minimum. Many makers judge at day 7 and decide at day 14; that is practice, not a rule. If the day-14 reference is clearly stronger than the day-3 one was, your recipe was never weak; your calendar was.

Week 3 — the add temperature (pour 12 October; verdict 26 October)

Pour two candles with whatever wick won week 1: the reference made the way you always make it, with the temperature measured and written down, and the change with oil added at the temperature your wax page states. For CSI 464 that is its stated pour temperature of 80°C–90°C; for CSI 468, 58°C–60°C; for Eco Soy CSI 400, add at 85°C and stir gently for 2 minutes. Where a page states no temperature, ask the wax maker rather than borrowing another wax's figures. Too-cool addition can give poor incorporation, and long holding at high heat tends to drive off lighter top notes; both are tendencies that vary by oil. Temperature causes are covered in depth in Did I Add Fragrance Oil at the Wrong Temperature?. One customer review of the Mini Electric Wax/Soap Melter (₹2,360.00) describes it solving temperature drops in a cold region; it is a convenience purchase, not a throw fix.

Week 4 — the oil (pour 19 October; verdict 2 November)

Only now does the oil get its turn, and it gets a fair one: your oil and a candidate oil, both at 8%, both with the winning wick, both added at the logged temperature, both judged at day 14. This is the "is it actually the fragrance oil?" test. If the candidate fills the room and yours does not, your oil is weak in this candle at this load, and you have earned a replacement. If both are weak with a full pool at day 14, the oil was never the problem: look at the wax (see Should I Change Soy Wax If My Fragrance Oil Has Weak Hot Throw?) or ask us to look at your log. The evidence standard for calling an oil weak is set out in How Do I Know If My Fragrance Oil Is Too Weak for Candle Making?.

The change log sheet

Copy this into a notebook or spreadsheet. One row per pair. Fill in the last two columns only when the numbers are in.

Six candles use 72 g of oil in all: 60 g of yours (two per pair in weeks 1 and 3, one in week 4) and 12 g of the candidate. They also use 828 g of wax, 6 jars and 6 wicks (one of your current size in week 1, and five of whichever wins, if that is C2). Keep a spare reference candle from every pair; it is the only honest answer to "was the last batch stronger?"

What each change costs, in the test and in production

Two costs matter: what the change costs to test, and what it adds to every candle if you adopt it. The second is the one that decides whether a fix is worth keeping.

Materials consumed by all six candles, valued at CSI prices (wax as CSI 464 at ₹0.51 per gram, ₹420.13; six clear jars at ₹94.40 each, ₹566.40; one C1 and five C2 wicks, ₹53.10; 12 g of Palo Santo, ₹72.22), come to ₹1,111.84, or ₹185.31 per candle, excluding your own oil. New spend is ₹749.30 with Palo Santo as the candidate, ₹897.40 with Coffee Vanilla Tobacco, and ₹1,103.30 if you also need the scale. All figures exclude labour, freight, packaging and GST.

What to buy for the log

Two tools that make the log trustworthy, then two candidate oils for week 4. The wick for week 1 is in the buy strip above; Eco wicks are covered in detail on the cold-good, hot-weak page. The oils are chosen because 8% sits well inside the range their pages state and because every size is in stock on 25 September 2026, not because of any measured throw, which CSI does not have.

Other candidates, if neither suits your range: any oil whose page states a range with 8% in the middle will do, and the ranked soy shortlist in Which Fragrance Oils Should I Buy for Strong Hot Throw in Soy Candles? lists five with every size in stock.

If Diwali stock cannot wait six weeks

Diwali falls on 8 November 2026. The log's last verdict lands on 2 November, which is too late to change the oil for Diwali production. Be honest with yourself about that, and use the log's first rows. The wick verdict lands on 12 October: if the C2 wins, pour production on 13 October with the C2 and your current oil, cure to 27 October, and dispatch in time. Keep weeks 3 and 4 running alongside, on test candles only, for your Christmas batch (25 December 2026). Do not change two things in a production batch because Diwali is close; that is how a range goes from weak to unpredictable.

Frequently asked questions