This is the whole path written as a dependency order — what has to be decided before what, and which decisions become expensive once you are past them. Most launch guides are written as a checklist. The problem with a checklist is that it implies the items are independent and that you can work them in any order. They are not, and you cannot. Choosing a market changes the formula. Choosing a pack changes the preservative system. Choosing a launch date changes which manufacturing route is available to you.
What makes a late decision expensive is rarely the decision itself; it is the work already built on top of the earlier one it contradicts. It links out to the detail rather than repeating it.
Stage One: Decisions That Cost Nothing Now and Everything Later
Four things, and all of them are free to decide today.
Who is this for, in one sentence
Not the brand story. The functional description: what it is, who uses it, when, on what, at roughly what price. "A morning serum for oily and combination skin in humid climates, worn under sunscreen." Every clause rules something out — the climate narrows the texture, the position in the routine decides what it has to sit under — and ruling things out is what makes the sentence useful to a formulator. A description that excludes nothing gives a lab nothing to work against.
Every market you intend to sell in
Including the ones twelve months away. This changes the formula, not just the label — permitted ingredients differ, UV filters differ dramatically, some claims require an approved active in one market and are ordinary in another. Adding a market after the formula is locked can mean reformulating and repeating testing. A market named before the formula is a specification; the same market named afterwards is a second project. What each market requires is the place to start.
What you are actually differentiating on
Be honest, because this decision selects your manufacturing route — and it works in that direction only. Brands that choose a route first tend to end up describing their differentiation to fit the route they have already committed to.
- The formula — a specific complex, texture or claim you intend to substantiate → ODM
- Brand, packaging, channel, audience → private label, and this is a legitimate and frequently more profitable position rather than a lesser one
- You already own a formula → OEM
The comparison, including what each costs in time and what you own afterwards, is in OEM, ODM and private label compared.
What is fixed: date, budget or specification
They will conflict, and they will conflict late, when there is least time left to think about it. Deciding in advance which one gives way is the difference between a managed trade-off and a crisis. A date driven by a trade show or a retailer window is a genuine constraint and it often selects the route on its own.
Stage Two: The Product
One product, not a range
Five SKUs is five formulas, five stability programmes, five sets of artwork and five registrations per market, while splitting the same launch attention. None of that work is shared between the five. The attention is shared: the same launch and the same first customers now have to carry five stories instead of one. Brands that launch one product well and extend from real sell-through data generally reach a profitable range faster than brands that start with five. The arithmetic is in the development cost breakdown.
Choose the format before the active
Format determines cost structure, minimum order, packaging options and shelf presence. A sheet mask, a serum and a cream are different businesses at the same revenue — what a unit costs to make, what a first order has to be and what the pack can do are not the same in any two of them. Sheet masks in particular suit a first product because they need no tooling — see what actually decides a sheet mask.
Choose an ascending active, not an emerging one
Enough consumer awareness that you are not educating the market, not so much saturation that the ingredient no longer differentiates. Emerging actives carry regulatory and supply risk a first product is badly placed to absorb, because there is no second product behind it earning while the first one waits — where actives sit in the cycle sets out the arc.
Decide packaging with the formula
Not after. Preservation demand, fill behaviour and component lead time all depend on the pack, and challenge testing runs in the final pack — so changing it later means testing again, which puts the pack decision inside the development window rather than after it. Airless pump versus dropper and the rest covers what each format costs you.
Agree the claim set early
Claims determine classification in the US and must satisfy the common criteria in the EU. Rewriting a claim after artwork is printed is the expensive version of this conversation. Clean beauty claims covers the ones brands most often assume are free.
Stage Three: The Partner
Send the same brief to three or four candidates and compare how they answer, not just what they quote. A quote tells you a price; an answer tells you whether the question was understood, and that is the better predictor of how the project will run once it is under way. Ask what they control, who owns the formula, which facility holds which certification, and what happens when a batch fails specification. The full question list is in how to vet a Korean cosmetics supplier.
Two things to settle in writing before development starts:
Formula ownership and exclusivity. Three separate terms — ownership, scope, duration — and none is implied by the word ODM. Each can be agreed and each can be declined, but only if it is raised while there is still a negotiation to have.
What the quote includes. Development, sampling, testing, components, tooling, artwork, documentation, freight. Two quotes are only comparable once both are expanded.
Stage Four: Development
The brief is the highest-leverage document in the project. A precise brief converges in two or three rounds; a vague one runs to six, and each round is two to three weeks of calendar as well as cost. What separates the two is not the lab but how much of the target was written down before the first sample — which is why the brief is free to improve. How to brief a Korean formulation lab is the structure.
Send a physical benchmark for sensory. It is worth more than any adjective: two people can use the same word for different textures, and nobody misreads a sample.
Give feedback comparatively and in writing — "the same but 20% less slip", not "not quite right". Number your samples and evaluate side by side.
Then stability and challenge testing, which is a fixed block of calendar time that cannot be compressed by paying more. Plan around it rather than against it — what the testing covers.
Stage Five: Market Entry
Most destination markets require a locally established party — a responsible person, licence holder or notification holder — though the exact role differs by jurisdiction, and the US is the notable exception: under MoCRA the responsible person need not be US-established, as long as the label carries a US contact for adverse event reports. A Korean manufacturer cannot fill these market-entry roles for you. We coordinate the manufacturing documentation; your market-entry party handles the local filing or regulated role where one is required.
- United States — MoCRA registration, listing, responsible person, safety substantiation. What an imported product needs
- European Union — EU Responsible Person, Product Information File, safety report, CPNP. The EU route in order
- Elsewhere — the markets we cover in depth
Start this in parallel with development, not after it. It is the second most common source of a slipped launch, behind revision rounds.
Stage Six: The First Order
Order to your forecast, not to the price break. Beyond your forecast, a per-unit saving becomes unsold inventory, which is a worse problem to have — cash sitting in stock cannot be spent on the marketing that would move it, which is how a good unit price turns into a slow launch. What actually sets your minimum — format, packaging, route, formula, market — is in what determines cosmetic MOQ, and low minimum order manufacturing is the conversation if the first quote lands above budget.
The Five Most Expensive Mistakes
In order of how often we see them.
1. A vague brief. Costs three to four extra revision rounds, which is a season, and those rounds are spent discovering what the brief should have said.
2. Deciding packaging late. Costs a repeat of stability and challenge testing.
3. Adding a market after the formula is locked. Costs reformulation plus repeated testing plus a second registration.
4. Launching a range instead of a product. Multiplies every one-time cost while splitting attention that was never large enough to divide.
5. Assuming exclusivity that was never negotiated. Costs you the differentiation you paid ODM prices for.
All five are decisions, not accidents, and all five are cheapest to fix before development starts.
Frequently Asked Questions
How long does it take to launch a skincare product?
Private label from an existing base is substantially faster than ODM development, because the slow parts — formulation rounds and stability testing — have already happened. For ODM, the fixed blocks are revision rounds, stability and challenge testing, packaging lead time and per-market registration. Count backwards from your date through all four rather than forwards from today.
Should we start with one product or a range?
One, in almost every case. A range multiplies formulas, testing, artwork and registrations while dividing the same launch attention. Extend from real sell-through data — a second product briefed after six months of customer feedback is a far better product than a second product briefed on the same day as the first.
Can we launch in several markets at once?
You can, and it costs roughly per market rather than being shared. Registration, artwork and documentation do not amortise across markets. Two markets done properly usually outperform five done thinly, and the second market is much cheaper to add once the manufacturing documentation exists.
What is the single biggest thing we control?
The brief. It decides how many revision rounds the project takes, and rounds are the largest variable in both cost and calendar. It is free to improve, which makes it the only item on this page with an unlimited return on effort.
Do we need to know the formula we want?
No, and specifying one is usually counterproductive. What a lab needs is a direction: one hero active, a supporting direction, hard exclusions, a sensory benchmark, your markets and your claim intent. Choosing every ingredient yourself removes the lab's ability to solve stability and sensory problems, which is the thing you are paying for.