How We Test a Formula Before It Is Released

Stage one

The brief: who is it for?

A formula starts with a person we cannot yet serve. Someone who finds every cream too heavy: that became the jelly. Someone whose face is where they want the result: that became Lumière. Someone who wants a tingle but not a moisturiser that flushes: that is the tingle, still in development. The brief names the person, what they use now, and what is wrong with it. If we cannot write the last part, there is no product to build. Adding a scent to an existing formula is not a brief; it is a variant, and that is a different, shorter process.

Stage two

The benchmark: what does it have to beat?

Before we make anything we buy the best products in that category. Not the cheap ones, not the ones we could make look bad; the ones we would use ourselves if we did not make our own, from brands we look up to, and there are several. We use them, read their lists, and write down what they do well. That is the bar. Beating a weak product proves nothing. If ours cannot beat the strongest thing we can find, the person in the brief is better off with the strongest thing we can find, and we would rather tell them that than sell them ours.

Stage three

The formula: every ingredient has to earn its place

This is where we owe someone a paragraph. Delux is a small brand and we do not have a lab. What we have is a manufacturer who has stopped behaving like a supplier and started behaving like part of the team. He has built this range with us from the ground up: not just made what we asked for, but told us when what we asked for was wrong, and more than once come back with a change to a formula we would never have thought of. Some of the best ideas in the Next-Gen base and in the tingle are his, and we would rather say that here than take credit for them. When the tingle has been through another version because the warmth was fighting the cooling agents, that is him and us on the phone at odd hours, arguing about one line. That relationship is the reason our lists look the way they do.

So the work goes like this. The base first: what it should feel like, what it must not do on a sunbed, how thin it should spread. Then the actives, and the question for each is the same: if we took it out, would the product be worse? If not, it does not go in, whatever it would look like on the box. This is why Delux lists are short where they can be and why nothing sits at a trace amount for the sake of a name. Then the housekeeping, chosen to be gentle, and the scent, always last. The first version is never the last version. The Next-Gen base went through eight months of this, most of it in conversations neither of us was paid for.

Stage four

The test: real skin, real sessions, real weeks

Samples come back and we use them. Not a patch on a wrist: the full routine, before every session, over weeks, alongside the benchmark products on the same schedule. Sunbed and sun where the product is for both. We are looking at four things: does it feel right on application and on the bed, does the skin stay comfortable across a block of sessions, does the colour come sooner and build deeper than the benchmark, and does it hold. We also do the boring things: does it transfer to acrylic, does it sit under sunscreen, does it go sticky if you use too much, does the scent survive the tube. Every one of those has sent a formula back.

Stage five

The verdict, and what happens when it loses

Then one question: did it beat the best we tested it against, in our own hands? If yes, it goes to packaging and it launches when it is ready. If no, it goes back to stage three with notes on what fell short, and sometimes it goes back more than once. And sometimes it does not come back at all. We have finished formulas, designed packaging, and cut the product before launch because at the last test it still did not win. Those are the products you have never heard of, which is exactly the point.

This is why a Delux launch is rare and why we do not announce dates. The Next-Gen formula took eight months. Lumière took four. The tingle is past a year and is not released, because the version that beats the benchmark is not the one that flushes hardest; it is the one that carries the full stack under a clean warmth, and that has been harder to make than anything before it. Go Dark, our original, is being rebuilt through the same process now so that a conditioning cream becomes a tanning cream as well.

Quick answers

Frequently asked

Keep reading

Related guides