The database has 178 "replacers", and most of them are not replacing something you cannot buy

About 4 minutes.

When you see "sandalwood oil replacer" in a catalogue, the intuitive reading is that sandalwood is too expensive or unavailable, so somebody built a cheaper version.

That reading can be checked.

87 pairs you can line up

There are 178 records with replacer in the name, and 144 of them have suppliers. Strip "replacer" off the name and look up the original, and 87 of them match.

(The other 57 do not match — presumably the original goes by a different name in the database, or the replacer stands in for an accord rather than a single material.)

Two hypotheses, tested

Hypothesis one: replacers exist because the original was banned.

Of the 87 pairs, how many originals have cosmetic_uses reading "not used anymore"?

Three. Dihydrocoumarin, costus root oil, musk ambrette.

Three out of 87.

Hypothesis two: the replacer is easier to get than the original.

How many replacers have more suppliers than their original?

Also three. Civet absolute (17 → 31), treemoss resinoid (1 → 4), karo karounde absolute (0 → 2).

Both intuitive explanations land three pairs each. Together that is under a tenth of the list.

What the 87 pairs actually look like

Sorted by the replacer's supplier count, the top of the list:

Original	Original	Replacer	Ratio
Civet absolute	17	31	1.82
Lemon oil	74	31	0.42
Sandalwood oil	81	27	0.33
Geranium oil	38	25	0.66
Lavandin oil	51	25	0.49
Bergamot oil	58	21	0.36
Lavender oil	99	21	0.21
Oakmoss absolute	42	21	0.50
Patchouli oil	148	21	0.14
Castoreum absolute	23	19	0.83
Agarwood oil	26	16	0.62
Musk ambrette	18	15	0.83
Costus root oil	21	14	0.67

Patchouli oil has 148 suppliers — among the best-supplied materials in this database — and it still has a replacer with 21.

Lavender at 99, grapefruit at 95, lemon at 74. None of these is hard to buy.

So what is a replacer replacing?

At least four different reasons are visible in these 87 pairs, all sharing one word:

One: the original is banned. Musk ambrette, costus root oil, dihydrocoumarin. Only three pairs, but the only situation where a replacer is the sole option.

Two: the original is animal-derived. Civet, castoreum. This is the only situation where the replacer genuinely wins — civet's 17 against 31 is the only ratio above 1 in the whole set.

Three: the original is expensive or unreliably supplied. Sandalwood, agarwood, rose absolute, orris absolute. This group matches the intuition.

Four: the original is cheap and legal, but you want something else. Lemon, lavender, patchouli, grapefruit, rosemary, mandarin — and this group is most of the list.

The database does not record the motive for the fourth group, and it is the one needing your own judgement: batch consistency, predictable cost, avoiding a specific constituent of the natural, or simply wanting something that smells similar and behaves differently. Citrus has an extra layer — phototoxicity is specific to natural citrus oils, and a replacer does not necessarily carry it.

(Batch-to-batch variation in naturals themselves is measured in How much does a natural vary.)

What a replacer's record looks like

These replacer records share a shape: category is almost always "fragrance agents", and cosmetic_uses is usually empty.

They are compounded bases, not single molecules — no CAS, no formula, no assay. Looking up a replacer gets you far less than looking up its original. The four reasons a trade name returns nothing are in Four kinds of nothing found.

What to do

• Do not assume a replacer means the original has a problem. Only three of 87 pairs work that way.
• Look up the original's cosmetic_uses and regulatory first. If the original is not flagged as retired, the replacer is solving price, consistency or something else — not legality. (How to read the retirement flags: Those 80 "not used anymore" materials.)
• The two animal-derived cases are the exception. For civet and castoreum, the replacer is the mainstream option.
• A replacer always carries less data than its original. No CAS means no IFRA mapping, no toxicology, no assay. Get the specification sheet from the supplier.
• Patchouli, lavender and lemon are freely available as originals. Using a replacer needs a specific reason; the word itself does not make it safer or cheaper.

What this doesn't establish

• This is a naming audit, not a market study. I counted records in this database whose names contain "replacer", not what anyone actually sells.
• The 87 pairs were matched by "same name minus replacer". I did not chase down the originals of the other 57, so their distribution may differ.
• Supplier count describes coverage, not availability or sales. That trap is in Deciding what to buy from how many formulas use it.
• The four reasons are my grouping. No field explains why any individual replacer exists, and the fourth group in particular is inference.
• I did not compare odour fields or composition for any pair. Whether a replacer smells like its original is entirely outside this article.
• Some originals carry restrictions in other fields — an IFRA use limit, for instance — while cosmetic_uses sits empty. "Only three are banned" is about that one field, not the regulatory picture.