I Produce Diwali Candles Months in Advance — How Early Can I Manufacture Without Risking Noticeable Fragrance Changes Before Customers Receive Them?

The cure is the floor, and it is the only hard date

There are two ends to a production window and only one of them is defined. The late end is defined by cure, and cure is a published expectation rather than a guess: the Luxury Soy Wax Chunks page states "allow candles to cure for a minimum of 48 hours (ideally 1–2 weeks) before performing burn tests", and most CSI fragrance-oil pages state a cure of 48–72 hours for soy before assessment. The series convention across these guides, and the practice of most makers selling soy containers, is 14 to 21 days. Whatever figure you settle on, the important consequence is the same: a candle you have not burn-tested at full cure is a candle you cannot describe accurately, and dispatching it is a commercial risk in a way that holding it is not.

The early end has no defined limit at all. This is not an oversight by CSI, and it is not a secret — it is a genuine absence. Across the 107 fragrance-oil product pages, the number stating a shelf life is zero, the number stating an expiry or best-before date is zero, and the number giving any storage instruction at all is zero. There is nothing to be found by looking harder. So "how early is too early" cannot be answered from a page; it can only be answered from units you have poured and held yourself.

Two general things are worth saying because they are uncontroversial rather than invented. Soy is a soft wax that contracts as it cools, which is why a candle's surface and its contact with the glass can keep changing after the pour — the mechanisms behind frosting and wet spots are both structural. And a candle that has been through a heat cycle has more thermal history than one that has not; that is a reason to keep a cooler control, not a reason to predict an outcome. Neither of those gives you a week count, and this page will not pretend otherwise.

Counting back from 8 November 2026

Here is the same decision as a table, with the two things that actually differ between the options: whether the cure completes before dispatch, and how old the candle is when the customer takes it home. The age column is the one that usually gets forgotten, because the maker's clock stops at dispatch and the candle's does not.

Read it as two zones rather than a ranking. Below about three weeks of lead time the plan fails for a reason nobody disputes: the cure is incomplete and the burn test has not happened. Above about four weeks you are choosing how much hold to accept, and hold is the part where honest guidance runs out and your own pilot units take over. Today is 29 September 2026, which puts you 40 days out — five weeks and five days. That is one clean cure cycle, and it is enough if you start this week.

Too early and too late both cost money

It is tempting to treat lead time as a single risk that grows with weeks. It is two opposing risks, and the cheapest plan is not the shortest one. Costs below are materials at CSI live pricing for the 500-candle example, so they understate the real loss — they exclude labour, packaging and the order you could not fill.

The asymmetry is the useful part. The late-pour failures are certain and self-inflicted: an uncured candle is uncured, and an untested wick is untested. The early-pour failures are uncertain and manageable: they may not happen at all, and a pilot batch plus a control pair costs ₹459.50 to insure against. Given a choice between the two, most festive producers should pour earlier and instrument it, rather than pour later and hope.

Pouring in dated sub-batches

This is the whole method, applied to a 500-candle Diwali run on a five-week runway. Sub-batch sizes of about a hundred are a method choice, chosen so that each is one working session and one carton stack; use whatever unit matches your melter and your shelving.

What a 500-candle Diwali run actually buys

Four lines cover a festive run of 500 × 200 g candles at 8%: 92 kg of wax, 8 kg of oil, 500 jars and 500 wicks. Each card shows the listed pack ladder, the price per kilo or per piece, and how far one pack goes. Where a size table shows ≈0 in its candles column, that is the 15 g trial bottle: it holds less than one 200 g candle's worth of oil at 8%, which is exactly what a trial bottle is for. Add a 3–5% wax allowance for spills and top-ups when you order, and build about 5% more candles than the order so you have something to reject with.

The pre-dispatch inspection

A week before dispatch, open the oldest cartons and work through a fixed list. The point of a list is that it separates the things that are genuinely not negotiable from the things that are your own tolerance to set — and it stops the inspection turning into a debate about whether a candle smells the same as it did in August.

What this year's run should leave behind

The uncomfortable truth about lead time is that this year's decision has to be made without this year's evidence. The only way out of that loop is to leave evidence behind deliberately. Four units and a notebook, and next Diwali the question stops being rhetorical.

One per cent of the run, and next year the lead-time question is answered from your own log rather than from a forum post. The next step from here is the formal version of the same idea — checkpoints, arms and a record sheet that supports an actual shelf-life statement — which is set out in how to establish a realistic fragrance shelf life. And if the surplus from this run is still sitting there in the spring, the question becomes a slow-moving-inventory question: see formulating for a nine-month hold.

Frequently asked questions