Prose.

Somewhere in Brooklyn, a machine is reading your zip code's humidity.

It is 9 a.m. in a 43,000-square-foot room that smells faintly of rosemary and warm plastic. A bottle moves down an automation line. It does not contain "Shampoo for Normal Hair." It contains a formula calculated from one person's answers about their curls, their tap water, the stress they have been under, and the city they happen to live in. A label prints. A name goes on. The bottle is centrifuged - twice - and shipped to exactly one address.

This is Prose on an ordinary Tuesday. A beauty company that does not make products until you ask for one, and never makes the same one twice if it can help it. In an industry built on producing a million identical things and hoping you are average, Prose decided to produce one thing and assume you are not.

Mass production solved scarcity. It also flattened everyone into a "type."

Arnaud Plas was a marketing executive at L'Oreal, moving between Paris, Dusseldorf and New York, when he noticed something inconvenient about the business he was in. The hair-care industry sorts billions of distinct human heads into a handful of boxes - dry, oily, color-treated, normal - and sells each box a product designed for the statistical middle of it. The system is efficient. It is also, for most people, slightly wrong.

"Normal" hair does not exist. Your scalp in humid August is not your scalp in dry February. The water in one city strips differently than the water in another. The standard shelf cannot account for any of this, because the standard shelf was stocked months ago in a factory that never met you.

The tension, stated plainly

Personalization is easy to promise and brutal to deliver. You can hand someone a quiz and a marketing story - or you can rebuild the supply chain so that a genuinely different product comes out the other end. Most brands choose the quiz. Prose chose the supply chain.

Four beauty-industry veterans walked out of the establishment to compete with it.

In 2016, Plas left L'Oreal. He had already, in a plot twist worthy of a heist film, hired his future co-founder Paul Michaux there the year before. The two were joined by CTO Nicolas Mussat and R&D advisor Catherine Taurin - a founding team with roots at L'Oreal and Phyto and a combined résumé deep in beauty, science and technology.

Their wager was that the hard part was not the chemistry, which the industry had spent a century perfecting, but the logistics of one: how to take a single person's data and reliably produce a single, made-to-order product, then do it again two million times without building two million separate factories. The answer would have to be software, hardware and a lab that thought like a database.

Prose launched online in December 2017. Products designed in Paris, bottled in New York. The team raised early money from Forerunner Ventures and Lerer Hippeau, then an $18 million Series B led by Insight Partners in late 2018, with later funding reportedly pushing total backing well past that. The pitch was not "better shampoo." The pitch was a new manufacturing model wearing a beauty brand's clothes.

A consultation, an algorithm, and a library of 185+ ingredients.

It starts with questions - a lot of them. Hair type, scalp condition, lifestyle, diet, stress, and the climate where you live or are about to travel. More than 50 factors go in. A proprietary algorithm then selects from a library of over 185 ingredients and composes a formula meant for your specific combination, not your category.

The order goes to one of two customization centers - the original in Brooklyn, and since 2025, a new one in Commerce, California. There, "formulation building blocks" are assembled to the calculated recipe, labeled, and finished. Then, crucially, you tell them how it went. The Review & Refine system feeds your feedback back into the next batch, so the formula evolves with you. The bottle you get in year two is not the bottle you got in year one.

The skeptic's question: does personalization actually scale?

For years the knock on made-to-order was that it could not. Custom meant slow, expensive and small. Prose's numbers argue otherwise - the company says it turned profitable in 2024 while still mixing every product to order.

The proof is not only financial. Prose's Custom Shampoo won Good Housekeeping's 2024 Sustainable Innovation Award. The skincare line was tested across 2,000+ skin types and tones and is dermatologist-approved. And the company built all of this while keeping its B Corp and carbon-neutral commitments intact - the kind of footnotes that are easy to drop the moment growth gets hard.

Personalization and sustainability were supposed to be a trade-off.

Made-to-order has a quiet sustainability advantage: you mostly make what people will actually use, instead of warehousing oceans of product that expire on a shelf. Prose leaned into it - climate-neutral certification, a Public Benefit Corporation structure, ethical sourcing, and US manufacturing. The mission is not a tagline bolted onto the brand; it is partly a consequence of how the factory works.

And then there is Singular, the part that looks past Prose itself. By packaging its consultation, data and formulation engine as a platform, Prose is positioning the thing it is best at - turning a person into a product - as infrastructure the rest of the industry might one day rent. The beauty brand, it turns out, was building a beauty-tech company the whole time.

Back to that bottle on the line.

It has a name on it now. It is not "Shampoo for Normal Hair," because Prose never believed in normal hair. It is a record of one consultation, one set of answers, one local climate, refined by one customer's feedback - and it is one of millions moving through two American factories that treat scale and specificity as the same project instead of opposing ones.

The wager Plas made in 2016 was that people are tired of being sorted into types, and that technology had finally made it possible to stop. Two million customers, $165 million in revenue, and a platform with ambitions beyond its own shelves suggest he was onto something. Prose did not make a better average. It got rid of the average.

Watch & explore

Five things that amuse and inform

Sources: Prose, BeautyMatter, Glossy, Good Housekeeping, Fashionista, PR Newswire, Crunchbase, B Lab. Figures are company-reported and approximate.