I’m Choosing Fragrance Oils for a Commercial Candle Collection — Should I Evaluate Discolouration, Wick Behaviour and Wax Compatibility Alongside Scent Quality and Hot Throw?

Why scent and throw are not enough for a collection

For a single candle, choosing by scent and hot throw is reasonable. For a collection you intend to produce repeatedly, it leaves out most of what goes wrong in practice. Complaints rarely say a candle smells wrong. They say it arrived darker than the photograph, it tunnelled, it sooted the jar, or this month's batch does not look like last month's. Those are colour, wick and wax problems, and they belong on the same sheet as scent.

A collection also multiplies every weakness. One oil that needs its own wick, its own wax or its own storage rules adds complexity to every production run. Selection is the cheapest point at which to notice that, because nothing has been bought in bulk yet.

This page is the master view for the vanilla, gourmand, woody and heavy cluster. The individual problems (vanillin darkening, oud wicking, amber pool depth, musk diffusion) each have their own pages. Here they become columns on one scorecard.

The two-step scorecard: gates, then weighted scores

Step one: gates

Gates are pass-or-fail. An oil that fails a gate leaves that product line, however well it scores elsewhere. Typical gates include:

Colour stability, for white or pale lines. An undyed aged jar must stay within your written tolerance compared with an unfragranced blank. Oils whose pages state 0% vanillin, such as British Rose, Jasmine, Sandalwood and Vetiver, are the natural starting candidates; vanillin-bearing oils such as Dark Vanilla usually belong in a warm-toned line instead.

Process safety. The oil's published flashpoint must sit comfortably above your addition temperature. Most CSI pages list 132–206°C, but Neroli lists 85°C and Gardenia 98°C, and Royal Oud states above 93°C. Those oils can still be used, but only if your process adds them below those figures.

Documents. CSI's oil pages state IFRA certification, with IFRA and MSDS documents available on WhatsApp request. If your stockists require them, collect them before scoring, not after.

Step two: weighted scores

Score each surviving oil from 1 to 5 on seven criteria, using your own test results and the published specs. Multiply by the weight, add up, and divide by 5 to get a total out of 100. The weights below are a suggested starting point to validate.

Worked example: four candidates for a mixed collection

A maker is choosing four oils for a warm-toned collection in a 200 g jar at 8% (16 g oil). The candidates are CSI Lavender, Dark Vanilla, Sandalwood and Citrus Lemon. Because the collection is warm-toned, colour is scored rather than gated. Cost is scored on bands the maker wrote down before testing, which is a suggested starting point: the cheapest band scores 5 and each dearer band loses one point. Here Citrus Lemon sits in the top band, Lavender and Sandalwood two bands lower, and Dark Vanilla three bands lower.

The remaining scores below are illustrative results of one maker's own tests, not CSI data. They follow the published specs where those apply: Dark Vanilla's page says it will yellow white or light wax, so it scores low on colour; Sandalwood and Citrus Lemon state 0% and approximately 0% vanillin; Sandalwood's 50 g and 100 g packs were out of stock on the pull, which lowers its supply score.

On hot throw and scent quality alone, Dark Vanilla leads with 225 of a possible 225 weighted points (5 × 25 + 5 × 20). Once wick behaviour, colour and cost are counted, it drops to third. That does not mean Dark Vanilla should be dropped. It means it should enter the collection knowing its costs: a separate wick check, a warm finish, and a higher oil spend per candle.

Reading the scorecard honestly

A scorecard is a decision aid, not a verdict machine. Use it to make trade-offs visible.

Close totals mean the weights matter. Lavender at 82 and Citrus Lemon at 80 are effectively tied. If your brand is about room-filling scent, raise the hot-throw weight; if it is about price, raise cost. Decide the weights before scoring, not after you see which oil wins.

A low score on one criterion can be fixable. Sandalwood loses points on supply because two pack sizes were out of stock on the pull; that may change. Dark Vanilla loses points on colour; a warm-toned finish turns that into a design feature. Note the fix next to the score.

Reviews are context, not scores. A verified Lavender buyer wrote: "Quality of fragrance is very good. Cold through and hot though both are perfect." Another wrote: "Fragrance is awesome. Hot throw is also very good after burn candle." Useful signals, but one buyer's candle is not your jar, wax and wick. Score only what you tested.

Scoring blends and repeat batches

If a collection candle is a blend of two oils, score the blend as its own candidate rather than averaging the parts. Two oils that each pass on wick behaviour can still produce a blend that carbons up, and a pale-safe oil paired with a vanillin-bearing one inherits the colour problem. The scorecard works on the finished candle, not the ingredients list.

Repeatability deserves a thought at this stage as well. The supply score above covers pack sizes and documents, but it does not prove that a second bottle behaves like the first. Before a finalist becomes a permanent product, pour it again from a fresh purchase and check that its hot-throw and colour scores land close to the first round. Formula and process drive throw as much as the oil does, so a reorder test also catches drift in your own method, not only in the supply.

Running the scorecard across a collection

Where this page fits

CSI already has guides that cover parts of this question. This page answers the specific oil-in-wax question above; these go deeper on the rest:

Three things to buy for this job

Frequently asked questions