Ask anyone inside a cosmetic business how long a product takes and you will get an answer in months, not weeks, and a wince. The sequence is well established and unglamorous, and every stage in it is a place where a commercial decision quietly overrides a technical one. Understanding the pipeline is the fastest way to stop reading launches as scientific events.
Stage one: the brief, where the constraints are set
A development brief is written by marketing, not by science. It typically fixes the category, the positioning, the retail price, the target cost of goods, the pack format, the desired texture, the claims wanted, the launch window and any ingredient exclusions the brand has committed to publicly.
Those exclusions matter more than most shoppers realise. A brand that has built its identity on a free from list has narrowed its formulator's options permanently, sometimes in ways that make the product harder to preserve or less stable. That is a marketing decision with a technical bill attached, and the bill is paid in the laboratory.
Stage two: bench work
Bench samples are made at very small scale, often a few hundred grams, on laboratory equipment that behaves nothing like a production vessel. A formulator will typically produce several variants that differ in emulsifier system, thickener, active level or sensory modifiers, and present them for assessment.
The assessment at this stage is largely sensory and aesthetic. Does it feel right for the price. Does it look like the category. Does it absorb without pilling under a foundation. These questions are not trivial, because they determine whether the product is used at all, but they are not questions about whether it works.
Stage three: the decision that is usually cost
Somewhere between bench and pilot, the cost of goods is confronted. The target was set in the brief against a retail price, and if the preferred sample misses it, something has to give. In practice the thing that gives is rarely the retail price and rarely the launch date.
| Lever | What it does to the product | Whether a shopper can detect it |
|---|---|---|
| Reduce the level of an expensive active | Lowers cost per unit directly, may reduce or remove the effect being relied on | Only if the pack states the level, and only if a previous level was known |
| Substitute a cheaper emulsifier or oil phase | Changes skin feel, absorption and sometimes stability | Yes, by feel, but usually only against memory of a previous formulation |
| Simplify the fragrance | Reduces cost noticeably, changes the experience of using the product | Yes, immediately, and it is the change shoppers complain about most |
| Change the pack to a cheaper component | Reduces cost and can change dosing, protection from light and air | Yes, visibly |
| Increase batch size | Reduces cost per unit with no change to the formula at all | No, and this is the lever brands prefer for exactly that reason |
The last row is the one worth holding onto. Volume is the cheapest way to reduce unit cost, which is why minimum order quantities shape the industry so heavily. That subject has its own piece: what a minimum order quantity decides.
Stage four: pilot and scale up
A formula that works in a beaker does not automatically work in a two thousand litre vessel. Heating and cooling take far longer, shear from an industrial mixer is different from a laboratory homogeniser, and the order and rate of addition can change the structure of the emulsion. Scale up failures are common and they are expensive, because they are usually discovered late.
This is one of the strongest arguments for the contract manufacturing model. A factory that developed the formula on its own equipment already knows it can make it. A brand that had a formula developed elsewhere and then places it with a factory sometimes discovers that the formula has to be adjusted to be manufacturable there, which raises the question of whose product it now is.
Launch windows and why they never move
Retail listings are agreed in range reviews that happen at fixed points in the year. Press and influencer schedules are booked months ahead. Pack components have long lead times and are ordered before the formula is finally signed off. Once those commitments exist, the launch date is the most expensive thing in the project to change and the formula is the cheapest.
So when a project runs late, the formula gets simplified. That is not a failure of care. It is the rational response to a set of costs that were fixed before the science started.
- Retailers. Get predictable ranges and negotiating leverage from a fixed calendar
- Pack suppliers. Get long, firm orders placed before development is complete
- Media and creator partners. Sell inventory against a date the brand cannot move
- The formulator. Absorbs the compression, and takes the blame if the product underperforms
A structural account of how launch calendars work in the sector, not a description of any specific project.
Stage five: safety, notification and the paperwork
Before a cosmetic can be placed on the UK market a safety assessment must be carried out and a product information file assembled, and the product must be notified. This is not a formality. The safety assessor can and does refuse, ask for a change, or require a warning on the pack. Where a product suddenly appears with a usage restriction on the label, this stage is often why.
The paperwork also fixes the label. Ingredients must be declared using the international nomenclature, warnings must appear, the responsible person's name and address must be shown, and durability must be indicated. What that adds up to is covered in how to read a cosmetic label in order.
Three years in development
That some period of three years elapsed between an early stage of the project and the launch. Development time is not a defined term and is not audited by anyone.
It can legitimately include periods where nothing was happening: waiting for real time stability data, waiting for a pack tool, waiting for a retail listing window, or pausing the project entirely.
It does not mean three years of laboratory work. It does not mean three years of testing. It does not imply any particular number of formulation iterations, any clinical work, or any original research. A long elapsed time can equally indicate difficulty, indecision or a delayed launch slot.
The brand would have to say what happened during the period. A meaningful version names the work: the number of formulation rounds, the tests run, the studies commissioned and their design. Elapsed time on its own is a number without a unit of effort attached.
Stage six: first production, and the version you actually get
The first production batch is checked against a retained standard: a reference sample of the approved formula. Colour, odour, viscosity, pH and microbiological quality are typically checked and recorded. This is the point at which the product becomes a repeatable industrial output rather than a project.
It is also the point at which small differences begin. Batch to batch variation within an agreed tolerance is normal and expected. A shade that looks slightly different, a cream slightly thicker, a fragrance marginally stronger: these are usually within specification, not evidence of a reformulation.
What to take from the pipeline
Three things, and they are all deflationary in a useful way. First, a launch is a scheduling event, not a scientific one. Second, the most consequential decisions about your product were made before any chemistry began, in a document about price and positioning. Third, the people best placed to explain a product are almost never the people presenting it to you.
None of that means products are bad. Most are competent and some are genuinely good. It means the correct question at a shelf is not what the story is, but what the product does over eight weeks on your own face, which is the only trial whose result belongs to you.