Why I Trained at ISIPCA Versailles to Build a Solid Perfume Brand in India

Sonal Sahani, founder of SOSA Home & Body - 15 May 2026 - 13 min read

If you have ever bought an expensive imported perfume and felt it died in Indian heat, the perfume did not fail - it just was not built for where you live. This is the founder's diary entry I have been putting off writing for five years. It is the story of why I flew to Versailles in 2018 to train at ISIPCA, what I learned there, and the slow translation work that followed when I came back to Mumbai and realised the curriculum did not match the climate I was building for. Call it the Versailles-to-India Translation Problem.

ISIPCA taught me how to formulate perfume. India taught me how to translate it. The first part took 10 months. The second part took 5 years. SOSA's 9 variants are the output of that translation work.

The Versailles classroom moment

It was October 2018. Versailles autumn light, the kind that makes the parquet floor in the ISIPCA classroom look like it has been polished by 50 years of perfumery students walking across it. We were in a raw-materials session, smelling vetiver from Haiti against vetiver from Java. The professor - she had spent 22 years at one of the big four houses before teaching - was talking about base-note diffusion in cold weather. She said, almost in passing, "fragrance is built for the climate of the wearer."

I wrote that line down. Then I sat in the back of the room and went through every perfume I had ever owned in my head. The Chanel my mother gave me at 19. The Tom Ford I had bought myself after a difficult first job. The Diptyque candle in my flat in Mumbai. Every single one of them had been built for someone else's climate. They were optimised for 20C Paris autumn, not 38C Mumbai April. They were optimised for dry continental skin, not the oilier, more sebum-active skin most Indian women have. They were optimised for indoor air that smells of nothing in particular, not for air that often carries cumin, garam masala, agarbatti and frying onion.

I had bought, worn and loved fragrances my entire adult life. None of them had been built for where I lived. The professor moved on to the next slide. I stayed with that sentence for the rest of the term.

That is the moment SOSA was born, even though it would not have a name until ten months later in a small kitchen in Mumbai, and would not have a finished product until 2020.

ISIPCA, in 4 paragraphs - what the school actually teaches

Before I get to the translation problems, it is worth being precise about what ISIPCA is and what it teaches. Founded in 1970 by Jean-Jacques Guerlain (yes, of that Guerlain family), ISIPCA is housed in Versailles, in the same town as the Palace, and is administratively part of the University of Cergy-Pontoise. It runs degree programmes in perfumery, cosmetics and food aromatics. Almost every senior perfumer at Givaudan, IFF, Firmenich and Symrise - the four large fragrance houses that supply most of the perfume sold globally - has either trained at ISIPCA or hired from there.

The curriculum is built around four pillars. The first is raw-material chemistry - learning to recognise, by nose and by gas chromatography, roughly 1,200 individual fragrance molecules and natural extracts. The second is formulation, where you learn the structural grammar of top, heart and base notes, the use of fixatives, the ratio of accord to solvent. The third is stability testing - how a formula behaves at 4C, 25C, 40C, in UV light, against oxygen. The fourth is regulation, primarily IFRA (the International Fragrance Association code) and the EU Cosmetics Regulation 1223/2009.

What ISIPCA does brilliantly is teach you to think in molecules. A senior ISIPCA-trained perfumer can pick up an unknown fragrance, smell three drops on a strip, and tell you which musks, which citrus aldehydes, which isoamyl acetate top note is in it. That skill is the price of admission to professional perfumery anywhere in the world. Without it, you are guessing.

What ISIPCA does not teach, because it is calibrated for the European market, is how that molecular logic shifts when the conditions shift. The curriculum assumes 18-22C ambient, 50 percent humidity, European skin chemistry, European indoor air, and EU regulatory framework. India is none of those. India is a country where the same formula behaves like a different formula. The professor's line - fragrance is built for the climate of the wearer - is the entire point. If you take ISIPCA chemistry and run it on Indian skin in Indian air, the chemistry is right but the experience is wrong. The translation work is what closes that gap.

The 6 translation problems ISIPCA does not teach

Here are the six problems I had to solve between 2019 and 2024 to get from ISIPCA-trained formulator to a 9-variant range that actually performs on Indian skin in Indian conditions. Each problem cost me months and a lot of prototypes.

1. Heat stability above 30C

ISIPCA stability testing peaks at 40C, but it is designed to confirm shelf survival, not on-skin behaviour at that temperature. Most European perfumes are designed to evaporate optimally at 32C - the average European pulse-point temperature. Indian pulse-point temperatures sit closer to 36C in summer. That 4-degree gap is the difference between a top note that opens elegantly over 45 minutes and one that flash-evaporates in 15. To slow that evaporation, the carrier had to change. Alcohol was out. A heavier, oil-and-wax carrier was in. That took 14 prototype rounds.

2. Humidity-driven evaporation rate

Humidity is not neutral to fragrance. At 50 percent humidity, water molecules in the air do not compete with fragrance molecules for nasal receptors. At 75 percent humidity, they do - and they win. This means a fragrance that lands beautifully on Paris skin can land muddy on Mumbai skin, because the air is already saturated. The translation answer: build the heart and base notes heavier, and let the top notes do less of the work. ISIPCA teaches a 30/40/30 ratio for top/heart/base. SOSA's range runs closer to 20/40/40 - a redistribution forced by Indian humidity.

3. Indian skin chemistry vs European

European skin pH averages around 5.0. Indian skin pH averages around 4.8 - slightly more acidic, with a different sebum profile (higher olein, lower stearin) and a different microbiome. Acidic skin tends to amplify musks and dull aldehydes. This is why a European-formulated musk perfume can smell loud on Indian skin and a European-formulated citrus can smell dead. The translation answer: dial back synthetic musk loading by roughly 18 percent and run the citrus through a higher-end natural cold-press to keep the aldehyde lift alive.

4. Masala-vapor coexistence

This one is uniquely Indian. The air in most Indian homes carries trace levels of cumin, coriander, fenugreek, garam masala, mustard seed and frying oil. Most of those molecules sit in the heavier mid-range volatility band - the same band as fragrance heart notes. So an imported perfume's heart note has to compete with food vapour for the same olfactory space. The translation answer was either to lean into spicy heart notes that harmonise with the kitchen (cinnamon, cardamom) or to use heart notes deliberately above the masala band (florals high, citrus high). SOSA's Fire variant takes the first path. Lust and Velour take the second.

5. Pulse-point temperature differential

A perfume's projection curve is shaped by the temperature at which it warms on skin. European formulators design for 32C wrist temperature. Indian wrist temperature in summer is 35-36C. That difference compresses the fragrance arc. Top notes peak earlier, heart notes peak earlier, base notes start earlier. To stretch the arc back out, the base-note molecular weight had to rise (heavier amber, denser musk, longer-tail vanilla) to slow the warming-up phase.

6. Beeswax-matrix carrier choice

The last problem was format. After 18 months of trying to make a liquid carrier work for Indian conditions, the answer was that no liquid carrier really did. Alcohol flash-evaporates in heat. Oil bleeds through fabric. Water emulsifies but does not hold fragrance load. The translation answer was a solid beeswax matrix, where the wax slows evaporation, the carrier oil binds the fragrance load, and warm skin re-melts the surface layer on contact. The format itself became the solution. This is the choice that made the entire range possible.

How I rebuilt the formulation for India

The rebuild was not glamorous. It took five years and a small flat in Bandra that smelled, at various points, like burnt vanilla, over-extracted vetiver and once - very memorably - like a chemistry exam I was clearly failing. Here is the rough timeline.

2019. First six months back in Mumbai, working with an ISIPCA classmate's contacts at a small Grasse-based supplier. The first prototypes were straight ISIPCA-format alcoholic eaux de parfum. They smelled good in the lab and died within 90 minutes on Mumbai skin. Lesson: lab is not life.

2020. Pivoted to oil-based formats. Better longevity, worse application. The oil bled into clothes, oxidised faster than expected in Indian heat, and the top notes still flash-evaporated. Lockdown gave me time to read everything on solid perfume formats I could find, including most of the academic literature on beeswax-fragrance binding from cosmetic chemistry journals.

2021. First solid-balm prototypes. I burned through roughly 40 batches before landing on a beeswax-to-jojoba-to-fragrance ratio that held in Mumbai April heat without going either too firm in winter or too soft in monsoon. The ratio went into a notebook I still have.

2022. The first three variants - what would later become Beast, Lust and Velour - went through 100-customer panel testing across Mumbai, Pune and Bangalore. Sterling, Siren and Desire followed in late 2022.

2023. The IFRA-compliant fragrance load was recalibrated specifically for the lower Indian skin pH. Fire, Storm and Sway joined the range. The full nine were on shelf by year end.

2024-2026. Iteration since then has been small. The range has held. Sales data, repeat-purchase rates and customer reviews confirm what the translation work was supposed to do - the perfume holds on Indian skin in Indian air. That is the whole point.

What ISIPCA still gave me - the science backbone

None of the translation work would have been possible without ISIPCA. The school gave me four things I could not have got anywhere else.

First, the molecular vocabulary. I can read a fragrance like a sentence. I know what isoamyl acetate does in a top note, why ambroxan behaves differently from cashmeran in the base, when to use Iso E Super and when to refuse it. That vocabulary lets me trace a problem to its molecule, not just to its smell. When Sway was over-projecting in early prototypes, I could narrow it down to a single coumarin choice within an afternoon. Without ISIPCA, that diagnosis would have taken weeks.

Second, stability testing protocol. Every SOSA batch goes through a six-point stability matrix - 4C, 25C, 40C, 50C with UV, oxygen exposure, and a 14-day Mumbai monsoon ambient test that I added. The first five came from ISIPCA. The sixth came from India. Without the first five, the sixth would be guesswork.

Third, IFRA compliance literacy. Knowing what you cannot put in a perfume is at least as important as knowing what you can. ISIPCA's regulatory module is the reason SOSA's labels declare the 26 EU allergens and the reason we have never had a recall.

Fourth, the smell library. Ten months of daily blind-smelling raw materials trains the nose in a way that nothing else does. When a vendor sends me a vetiver sample, I know within three seconds whether it is Haiti, Java, or Bourbon, and roughly what season it was harvested in. That competence is what stops me from being talked into bad raw materials by a confident salesperson.

So when I say ISIPCA is what made SOSA possible, I mean it literally. The translation work is the visible bit. The science backbone is the invisible bit that holds the translation up.

Why I chose solid balm format

I get asked this often, usually phrased as "why not just make a spray?" The answer is that the format choice was the translation answer.

A spray needs a carrier. Every available spray carrier - alcohol, water-alcohol, oil, silicone - has a behaviour problem in Indian conditions. Alcohol flash-evaporates above 35C, taking the top notes with it. Water-alcohol blends precipitate over time in humid storage. Oil sprays bleed and stain. Silicone sprays are tolerated by some skin types and rejected by others. None of them solve the heat-humidity-skin chemistry triangle.

A solid balm in a beeswax matrix solves the triangle at the format level. The wax holds the fragrance below its flash-evaporation point. The carrier oil binds the molecules so they release slowly. The warm-skin activation curve matches the way Indian pulse points heat the surface on contact. And solid perfume travels - it does not leak in an autorickshaw, it passes airport security, it survives the Mumbai monsoon in a handbag. The format is not just a stylistic choice. It is the answer to the climate problem.

The 15g size, the round tin, the metal lid, the matt finish - those are stylistic. The wax matrix itself, with its specific ratios, is the translation work made visible.

If you want the longer version of this argument, the cluster blogs on solid perfume vs spray perfume and alcohol-based perfume was never built for Indian conditions go deeper.

The 9 SOSA variants - each solves one ISIPCA-vs-India problem

Here is the map. Each of the nine variants in the SOSA range is the answer to one of the translation problems above, or a combination. This is not marketing - this is how I think about the range when I sit down to do the next stability test.

Each variant is its own answer. The range as a whole is the full nine-part translation. Browse the complete SOSA solid body perfume collection to see all nine in context.

FAQ