Foundation — a swamplink research property
Expensive foundations are sold on exotic-sounding ingredients. Measured against twelve FDA filings, price buys a longer list and better-placed actives — but not rarer chemistry.
Paying more buys a longer ingredient list and more marketing-facing actives — but not rarer chemistry. Mean ingredient rarity is essentially flat from the cheapest bottle to the dearest (0.62 at budget, 0.59 at mass, 0.65 at prestige, 0.65 at luxury). Prestige formulas are more elaborate, not more exotic.
| Measure | Budget | Mass | Prestige | Luxury |
|---|---|---|---|---|
| Ingredients in the formula | 27.0 | 35.0 | 47.5 | 36.0 |
| Mean ingredient rarity | 0.62 | 0.59 | 0.65 | 0.65 |
| Share that are named actives | 0.00 | 0.06 | 0.18 | 0.11 |
| Where those actives sit in the list | — | 0.78 | 0.37 | 0.76 |
Rarity is measured inside the corpus: an ingredient declared by one product of twelve scores 0.92, one declared by all twelve scores 0. That is a prevalence measure, not a supplier-price measure — a cheap commodity used by only one brand still scores as rare here.
Labelling law requires descending order of concentration down to 1%. So the position of an ingredient is evidence about how much of it there is — and a named active near the bottom of a forty-item list is present at well under 1%, however prominently it appears on the box.
Mass-market brands bury their actives; prestige brands don't. When a drugstore foundation carries a named active it sits around 78% of the way down the list. In prestige formulas the median active sits at 37% — materially higher up, and the only measure in this study where the price gap buys something unambiguous.
| Brand | Tier | Ingredients | Mean rarity | Exclusive to it | Named actives | Median active position |
|---|---|---|---|---|---|---|
| e.l.f. | budget | 27 | 0.62 | 7 | 0 | — |
| Maybelline | mass | 29 | 0.58 | 5 | 4 | 88% |
| Rimmel London | mass | 50 | 0.69 | 12 | 4 | 80% |
| Revlon | mass | 27 | 0.60 | 6 | 2 | 78% |
| L'Oréal Paris | mass | 34 | 0.54 | 2 | 0 | — |
| L'Oréal Paris | mass | 35 | 0.55 | 0 | 1 | 66% |
| MAC | prestige | 38 | 0.61 | 8 | 6 | 36% |
| Clinique | prestige | 37 | 0.66 | 14 | 7 | 38% |
| Estée Lauder | prestige | 50 | 0.66 | 11 | 9 | 38% |
| Lancôme | prestige | 36 | 0.55 | 0 | 5 | 42% |
| Lancôme | prestige | 51 | 0.68 | 16 | 11 | 37% |
| Dior | luxury | 36 | 0.65 | 10 | 4 | 76% |
| Estée Lauder | prestige | 73 | 0.75 | 35 | 16 | 30% |
"Named actives" are matched by pattern — extracts,
ferments, peptides, vitamin derivatives and the like — and the full pattern list is printed in
tools/analyze.py and in data/analysis.json, so the classification can
be argued with rather than taken on trust.
Where this is weak. Rarity here is corpus prevalence, not supplier availability or price, because credible bulk-price data for cosmetic raw materials is not freely published at a standard this project would accept. Position is evidence of concentration, not a measurement of it. And Lancôme's Teint Idole scores as one of the least rare prestige formulas precisely because it shares L'Oréal's parts bin — which is a finding, not an error.