I Want to Launch a Candle Brand in India With Only Four Core Fragrances — How Do I Choose Profiles That Cover Different Customer Preferences?

What "different customer preferences" can honestly mean

There is a version of this question that cannot be answered: which four scents will Indian customers buy. Nobody holds that data at the level that would make it useful — not a supplier, not a marketplace, not a consultant. Preference varies by city, by price band, by season, by whether the candle is for a home or a gift, and by who happens to walk past your table. Anyone quoting you a ranking of families by popularity is offering an opinion with a number attached to it.

There is also a version of this question that is answerable, and it is the more useful one: how do I make sure my four SKUs are not four answers to the same question? That is a coverage problem, and coverage can be audited from the product pages before you spend anything. Two axes do almost all the work.

The first axis is warm against fresh — whether the oil's notes sit in the amber, wood, spice and vanilla register or in the citrus, green, mint and dewy-floral register. The second is sweet against dry — whether the finish reads as sugared or as clean. Four oils, one per quadrant, and your range has a shape. Eight oils spread across two quadrants, and it does not, however good each one smells.

The four corners, and the oil CSI lists in each

Here is the core set with everything the four product pages state. The vanillin column is the one most new brands skip and then discover eight weeks later, when a white-wax candle in a clear jar has gone the colour of weak tea. Where a page states nothing, the table says so.

Three things follow from that table and they are all operational. First, every one of the four caps at 10% in soy, so a single 8% house load works across the whole range and your recipes stay comparable. Second, only one of the four — Vanilla Bean — carries a published discolouration risk, which makes it the one scent in the core that should be photographed and sold in a coloured or opaque vessel, or sold with amber toning described as part of the product. Third, Rajnigandha publishes neither a flash point nor a vanillin figure, so you add it below your wax's stated add temperature and you check a cured jar yourself before you commit. A silent page is not a clean bill of health.

Reading the map: what each corner answers, and what it does not

A corner is useful when you can say in one sentence which customer request it satisfies and which it does not. The table below is the script. It is written from the pages' own descriptions and from nothing else — no claim about which request comes up most often, because we do not know.

Notice what the fourth column is doing. Each corner comes with a different production constraint, which means four SKUs is also four slightly different manufacturing recipes even at one load. Sandalwood is the only one of the four whose page states a separate pillar figure, so it is the only one you could extend into a moulded product without re-researching. Vanilla Bean is the only one that will visibly change your wax colour. These are the details that make four SKUs manageable and eight SKUs chaotic.

The four core oils, costed per candle

Each card shows every listed rung, the price per 100 g at that rung, and roughly how many 200 g-class candles the pack pours at 14.8 g of oil. Read the per-candle figure at the bottom as the oil line in your cost sheet — wax, wick, jar, lid, label and packaging sit on top of it.

What a four-SKU launch actually costs at each rung

The decision that sinks new candle brands is not which four scents. It is which rung to buy them at. The table below prices the identical four-SKU core at all five listed rungs, with the candle count and the cash tied up in oil alone. Read the right-hand column as the question you have to be able to answer before you click.

Look at what the step from 100 g to 1 kg actually buys. The total oil bill rises from ₹2,382.68 to ₹21,852.80 — ₹19,470.12 more — and the per-candle oil cost falls from ₹352.64 for a set of four candles to ₹323.43, a saving of ₹29.21 across four candles. To recover ₹19,470.12 of extra cash at ₹7.30 a candle you have to pour and sell about 2,670 candles. At twenty candles a month that is more than eleven years. The kilo is not a saving at launch; it is a bet, and the bet is on scents nobody has bought yet.

There is a second trap in that table. Sandalwood's per-gram price does not move at all between 100 g and 1 kg, so the kilo buys you zero saving — purely continuity of supply. And across CSI's wider fragrance range the ladder sometimes runs the wrong way: Sage and Cedarwood lists 500 gm at ₹2,584.20 and 1 kg at ₹6,867.60, so two 500 g bottles give you the same 1,000 g for ₹5,168.40 — ₹1,699.20 less than the kilo. Never assume the bigger pack is the cheaper pack on this store; divide the two prices yourself.

The four-SKU launch sequence

Six steps from an empty shelf to a four-scent range with a reorder rule attached. The sequence is deliberately slow at the start and fast at the end, because the expensive mistakes all happen in the first two steps.

When a fifth scent earns its slot

A fifth SKU is not free. It is a new wick test, a new cure observation, a new label, a new photograph, a new line on the stock sheet and a new bottle that may not get finished. The test for adding one is therefore not "would this be nice?" but "which specific request is my four unable to answer, twice?"

Palo Santo deserves one extra line, because it illustrates both halves of this guide. Its ladder is completely flat — ₹6.018 a gram at 50 g, 100 g, 500 g and 1 kg — so there is never a price reason to buy it big. And its stated flashpoint is 93°C, which is the lowest among the oils named on this page and below the 80–90°C add window several waxes publish; that is a formulation detail to respect, not a defect. Specification shapes how you make the candle. It never tells you whether anyone will buy it.

Frequently asked questions