The number of revision rounds a development project takes is set almost entirely by the brief, and almost none of it by the lab. This is not a comfortable thing to tell a client, but it is what the pattern shows: precise briefs converge in two or three rounds, vague briefs run to six or seven, and the difference in calendar time is a season.
The good news is that a better brief is free. It costs an afternoon of decisions you were going to have to make eventually, moved earlier. What follows is what a Korean formulation lab actually needs from you, in roughly the order it needs it, and what the common failure modes look like.
Why Vague Briefs Are Expensive
A revision round is not one email. It is a lab bench session, a new sample, a courier to your country, a week of you evaluating it, and a conversation about what to change. Two to three weeks, every time, and the cost is calendar rather than cash — which is exactly why it gets underestimated during planning and dominates the schedule later.
The mechanism is simple. Without a target, the lab has to guess. A guess has maybe a one-in-three chance of matching what you imagined, so the first sample becomes a way of discovering what you wanted rather than a proposal against a specification. That is a legitimate way to work, but it should be a decision you made, not an accident.
The worst version is a brief that is vague and confident: "a luxurious, fast-absorbing, lightweight but rich anti-ageing serum with clean ingredients at an accessible price." Every word means something different to different people, several of the requirements conflict, and none of it constrains anything.
The Six Things a Lab Actually Needs
1. The product concept in one sentence
Not the marketing copy. The functional description: what it is, who uses it, when, and on what. "A morning serum for oily and combination skin in humid climates, worn under sunscreen" tells a formulator more than three paragraphs of positioning, because every clause rules something out.
Include the price tier here too. A formula for a mass-market price and a formula for a prestige price are different formulas, and knowing which one you want in advance saves an entire round.
2. The sensory target
This is the single most valuable thing in a brief and the one most often left out. Sensory is where subjective language fails hardest, so use references instead of adjectives.
Send a benchmark product. A physical sample of something on the market whose texture you want to approximate is worth more than any description. You are not asking to copy a formula — you are establishing a shared vocabulary for what "lightweight" means.
If you cannot send a benchmark, specify along axes instead:
- Viscosity — water-thin, light gel, gel-cream, cream, balm
- Finish — matte, satin, dewy, glossy
- Absorption — instant, fast, slow with a nourishing after-feel
- After-feel — dry, cushioned, tacky-acceptable or tacky-unacceptable
- Slip and spreadability
- Colour and clarity — clear, opalescent, opaque, tinted
And say which of these you will not compromise on. Every formula trades one against another; knowing your priority order lets the lab make the trade in the direction you want.
3. The active direction
Not a full ingredient list — a direction. Which one active is the hero, which supporting actives you want present, and which ingredients you want excluded and why.
Exclusions matter more than people expect. "No fragrance," "no essential oils," "no animal-derived ingredients," "must be certifiable as vegan," "no silicones" — each of these narrows the formulation space substantially, and discovering one of them in round four means discarding rounds one through three.
Be explicit about which claims you intend to make about the actives, because a claim you intend to substantiate imposes requirements that a claim you do not intend to make does not. If you want to build around a specific active, an ingredient page such as PDRN serum and ampoule development is a reasonable starting point for what that choice implies for format and packaging.
4. The destination markets
List every market you intend to sell in, including the ones that are twelve months away.
This changes the formula, not just the label. Permitted ingredients and permitted concentrations differ by market. Preservative choices differ. Sunscreen filters differ dramatically. Some markets require specific certification for a product's claims, and some require it for its supply chain. Adding a market after the formula is locked can mean reformulation, repeated stability testing and a second round of registration.
If a market is genuinely uncertain, say so and say when you will know. A lab can often keep a formula compatible with a possible future market at little cost, if it knows to.
5. Packaging intent
Formula and package are one decision, not two, and briefing them separately is a common and expensive mistake.
Say what format you want and how firm that is. An airless pump, a dropper bottle, a jar, a tube and a glass ampoule impose different preservation requirements, different viscosity windows and different fill behaviour. A formula developed for an airless pump may not be preservation-adequate in a jar; a viscosity that fills cleanly on one line will not on another.
Also state your volume and price expectation for the component. Custom tooling and stock components have very different lead times and minimums, and component lead time frequently exceeds formulation time on the project plan.
6. Timeline and what is fixed
State your target launch date and say what is driving it — a trade show, a retailer window, a funding milestone. Then say which of budget, date and specification you would sacrifice first if the three cannot all be met. They frequently cannot.
Be realistic about the fixed blocks. Stability and challenge testing occupy a defined period that cannot be shortened by paying more. Any plan that assumes it can is not a plan.
What Not to Put in a Brief
A competitor's full ingredient list with "make this." An INCI list is ordered but not quantified, so it cannot be reverse-engineered into a formula, and asking for a copy raises questions no one needs. Send the product as a sensory benchmark and describe what you want different about yours — that is both more useful and more defensible.
Every idea the team has had. A brief with twenty-five requirements has no priorities, which means the lab assigns them for you. Rank them, or at minimum separate must-have from nice-to-have.
Marketing copy in place of specification. "Transformative," "clinically inspired," "next generation" — none of these constrain a formulation. Keep the positioning document and the brief as separate files and send both.
Claims you have not checked. If a claim is central to the product, raise it in the brief so the lab can flag whether it is achievable and substantiable in your markets. Discovering after development that the claim needs to change is the most expensive possible time to find out.
A Brief That Works
The shape below covers what matters, and fits on two pages.
| Section | What to write | |---|---| | Concept | One sentence: what, who, when, price tier | | Format | Serum, cream, mask, cleanser — and how fixed it is | | Sensory target | Benchmark product, plus your non-negotiable axes | | Actives | Hero active, supporting direction, hard exclusions | | Claims | What you intend to say, and which claims are essential | | Markets | Every destination, including the ones twelve months out | | Packaging | Format, stock or custom, volume, decoration intent | | Volume | Realistic first order and twelve-month projection | | Timeline | Target launch, what is driving it, what flexes first | | Constraints | Certifications, standards, retailer requirements |
Send it as one document. Split briefs — sensory over email, actives in a spreadsheet, markets mentioned on a call — are how requirements get lost.
How the Rounds Should Go
With a good brief, a normal development runs roughly like this.
Round one should be close on concept and approximately right on sensory. If it is not close, the gap is usually a brief problem rather than a lab problem, and it is worth diagnosing before ordering another sample.
Round two narrows sensory and confirms the active system. This is where you should be giving specific, comparative feedback — "the same but 20% less slip", not "still not quite right".
Round three should be a confirmation sample. If you are still discovering requirements at round three, something was missing from the brief and it is worth rewriting it rather than continuing to iterate.
Then stability and challenge testing, which is fixed calendar time and cannot be compressed.
Give feedback comparatively and in writing. "Softer than sample A but less rich than sample B" is actionable. "Not quite luxurious enough" is not. Number your samples, keep them, and evaluate them side by side rather than sequentially from memory.
When You Do Not Have a Brief Yet
Not every brand starts with this much clarity, and pretending otherwise wastes rounds too. There are two honest ways to proceed.
The first is to ask for a direction round — two or three deliberately different samples across the sensory space, understood by both sides as exploratory. You are buying vocabulary, not a candidate formula.
The second is to start with an existing base. Selecting from a private label catalogue puts a real product in your hands, which is a far better foundation for a brief than imagination. Many brands launch this way, learn what their customers actually respond to, and write a much sharper brief for an ODM hero product in year two. The trade-off between those routes is set out in OEM, ODM and private label compared.
Either way, say which mode you are in. Custom cosmetic formulation work is scoped very differently for exploration than for execution, and a lab told which one it is doing will serve you better than one guessing.
Frequently Asked Questions
How long should a formulation brief be?
Two pages. Long enough to cover concept, sensory, actives, claims, markets, packaging, volume and timeline; short enough that every line constrains something. A twenty-page brief usually contains less information than a good two-page one, because the priorities are buried.
Do I need to know exactly which ingredients I want?
No, and specifying a full ingredient list is usually counterproductive. What you need is a direction: one hero active, a supporting direction, and a clear list of exclusions. Choosing every ingredient yourself removes the lab's ability to solve stability and sensory problems, which is what you are paying for.
Can I send a competitor product as a reference?
Yes, as a sensory benchmark, and it is the most useful single item in a brief. What does not work is sending an ingredient list and asking for a copy — an INCI list is ordered but not quantified, so it cannot be reverse-engineered into a formula. Send the product, describe what you want different.
How many revision rounds should I budget for?
Two to three with a solid brief; five or more without one. Budget the calendar, not just the cost: each round is a bench session, a sample, shipping and your evaluation time, which is realistically two to three weeks. This is why the brief is the cheapest place to spend effort in the whole project.
What if my launch date is already fixed?
Say so in the brief, and say what flexes — specification, budget or the date itself. A lab that knows the date is immovable can propose a different route, such as starting from an existing base rather than developing from scratch. A lab that finds out late has fewer options, and none of them are good.