The most common questions in my inbox lately aren't "which factory should I pick" — they're "how much of this active do I actually need, and can I legally say that on the label." Active skincare is the highest-margin and highest-risk category in OEM. Almost every failure traces back to two things: ingredient concentration and efficacy claim compliance. Here are the 8 questions I hear most, answered with an evidence-first lens.
Q1: If I claim "5% niacinamide," does the formula really have to contain 5%?
Not higher is better, and you can't just make up a number. Niacinamide's evidence-backed effective range is 2%–5% — repeatable human studies support this for brightening, oil control, and barrier repair (evidence grade A–B). Above 10% it becomes more likely to irritate and redden, with diminishing returns.
The red line: the advertised concentration must match the filed formula. In the US, MoCRA requires FDA facility registration and product listing; in the EU you notify via CPNP under Regulation (EC) 1223/2009. Claiming "5% niacinamide" when the formula holds 1% is misleading — a false-advertising risk that can trigger delisting or fines.
My advice: build around a "classic active + one trend active" instead of one big number. Niacinamide 3% + tranexamic acid 2% is steadier than a solo "10% niacinamide" story.
Q2: If an ingredient sits below position 8 on the INCI list, is it "added but not enough"?
This rule of thumb is directionally right, but has exceptions.
- For most conventional ingredients: the INCI list runs highest-to-lowest concentration, and anything under 1% may be listed in any order. So sitting below position 8 usually means the level has dropped below 1%. "Contains" is not the same as "at an effective level" — like dropping one sugar cube into a whole pot of soup.
- The exceptions are peptides and certain high-potency actives: palmitoyl pentapeptide-4 works at just 0.001%–0.005%, and acetyl hexapeptide-8 (Argireline) as a solution at 5%–10%. These are naturally low in the list — you can't flatly call them "underdosed."
The point isn't position; it's whether the effective concentration is reached. Every OEM checklist I hand clients includes this one line: verify each hero ingredient's minimum effective concentration, not just whether it "contains" it.
Q3: I want to add retinol — what concentration should I start at? Can I use it in autumn/winter?
Retinol is among the best evidence-backed anti-aging actives (grade A), but also the biggest source of blowups.
- Concentration: start at 0.025%, and only after tolerance build up to 0.1%, 0.3%, and at most 0.5%. Don't launch straight at 1% — that's not potency, that's a burn.
- Cadence: begin 2–3 nights a week, increase frequency as tolerance builds; night use only (photolabile), with sunscreen mandatory during the day.
- Contraindications: avoid in pregnancy and lactation; don't stack with high-concentration AHA/acids (stimulation doubles).
- Autumn/winter is not just fine — it's actually better: UV is milder and barrier stress is lower, making it a good window to build tolerance, as long as you pair it with hydration and repair (niacinamide, ceramides).
When I help OEM clients launch retinol, I always insist on a "repair companion" product rather than letting consumers apply raw retinol alone.
Q4: Is more ceramide always better? What ratio should a barrier-repair formula use?
No. It's about the biomimetic ratio, not the volume. In the "brick-and-mortar" model of the skin barrier, corneocytes are the bricks and intercellular lipids are the mortar — and the optimal molar ratio of ceramides : cholesterol : free fatty acids is 3:1:1 (Elias brick-wall model, industry consensus, evidence grade B–C). Dumping in a high concentration of ceramide alone throws the ratio off, spreads poorly, raises cost, and can actually slow repair.
- Effective ceramide level: 0.1%–1% is enough, when paired with cholesterol and free fatty acids at 3:1:1.
- Advanced stack: add madecassoside (0.1%–1%, anti-inflammatory and soothing), panthenol, and bisabolol for an "inflammation-first" soothing layer — calm first, then repair.
CeraVe and La Roche-Posay built their sensitive-skin franchises on "repair the barrier first, then claim efficacy" — the same principle I use when designing sensitive-skin OEM products.
Q5: How do I write efficacy claims without breaking the rules? Can I say "7-day brightening," "repairs sensitive skin," or "comparable to medical aesthetics"?
Mostly no. This is the biggest minefield in active skincare. My self-check asks four things: what's the evidence, are ingredients transparent, is the time promise realistic, and is the comparison exaggerated?
- Absolute/superlative and time-promise language: "7-day brightening," "28-day spot removal," "permanent," "best" — prohibited.
- Medical language: "cures," "repairs sensitive skin" (as a medical claim), "anti-inflammatory," "cosmeceutical" — cosmetics are not drugs; implying a medical effect is banned ("soothes" and "supports the barrier" are fine; "treats" and "cures" are not).
- "Comparable to medical aesthetics" / "replaces injections" — cross-dimensional exaggeration, prohibited.
Under MoCRA and EU 1223/2009 (Common Criteria, Commission Regulation 655/2013), every claim needs substantiation — brightening, anti-wrinkle, soothing, oil-control all require filed supporting data. The Ordinary and Paula's Choice print specific claims because a full substantiation file backs them, not because a copywriter dreamed it up.
Q6: Do brightening/sunscreen/acne products need special registration? How long does it take?
It depends on the claim — don't assume. Under China's 2021 Cosmetics Supervision and Administration Regulation, products split into two classes; the same "claim-first" logic applies when you map it to export markets:
| Type | Covers | Route | Typical timeline | | Special cosmetics | Whitening/spot, sunscreen, anti-hair-loss, hair dye, perm | Registration (NMPA in CN; in EU, stricter substantiation + CPNP notification) | usually 3–6 months |
| General cosmetics | Moisturizing, soothing, oil control, cleansing, acne (without whitening/spot claims) | Filing/notification | roughly 30 days |
Watch the word "acne." Plain oil-control/anti-acne uses general filing; but add a "spot fading" or "whitening" claim and it drops into special registration. Likewise "sunscreen" — even SPF 15 is special in most jurisdictions. Before you OEM, ask: "does my claim cross the special-cosmetics line?" That one question saves major rework.
Q7: How much does a human efficacy evaluation cost? Can I skip it?
If you claim a benefit, you can't skip the substantiation. Human efficacy evaluation is the evidence base behind any claim, and cost varies by test. A simple moisturizing/soothing instrument test runs a few thousand dollars; anti-wrinkle or brightening tests need human subjects and a study period, typically $1,000–$12,000, more for specialized claims.
- When it's mandatory: claiming brightening, anti-wrinkle, firming, soothing, oil control, or moisturizing — the notification requires a claim summary (efficacy substantiation).
- When you can defer: only "cleansing" / "basic moisturizing" without specific numbers, backed by literature or raw-material data — but that window keeps shrinking.
My advice: treat efficacy testing as a fixed product cost, folded into the OEM quote up front — not a last-minute add-on.
Q8: How do I balance concentration and cost? Does stacking actives equal a good formula?
Stacking ≠ good formula. A good formula is "precise and effective," not "more ingredients and bigger numbers."
- Three-layer structure: moisturizing base (occlusives + humectants + emollients) + active layer (effective concentrations met) + safety/stability layer (preservatives + buffering). All three must exist — drop one and something breaks.
- Cost logic: actives are usually the most expensive part. Blindly piling concentration can raise cost 30% while adding maybe 5% efficacy, and increase irritation risk. The best value is the "classic active + one trend active" structure, spending your budget where it counts.
- OEM parameters: MOQ 500–1,000 units, sampling 7–15 days, mass production 30–45 days, with a reference sample sealed once the formula is locked. Concentration, formula, and claims stay locked together — change one and update the other two.
⚠️ 3 High-Frequency Failure Alerts
- "High concentration on the label, low concentration in the jar": claimed vs. filed mismatch is misleading advertising — the single heaviest penalty.
- "Anti-acne" that drags in "spot removal": one word drops you from general filing into the special-registration channel, costing 3 months of waiting.
- "Claims built on copy alone": writing "fades fine lines in 28 days" with no human efficacy data fails notification — and is a time bomb even if it lists.
✅ 10-Point New-Brand Checklist
- Verify each hero active's minimum effective concentration (niacinamide 2–5%, retinol 0.025–0.5%, ceramide 0.1–1%).
- Claimed concentration = actual amount in the filed formula, to the letter.
- Compose ceramides at 3:1:1 (ceramide : cholesterol : free fatty acids).
- Sensitive-skin products get a soothing layer first (madecassoside/panthenol/bisabolol), then actives.
- Run the four hype questions before claiming — cut absolute, medical, and time-promise language.
- Ask "does this cross the special-cosmetics line?" — whitening/spot, sunscreen, anti-hair-loss → registration.
- Fold human efficacy evaluation costs into the OEM quote; don't discover the missing report at launch.
- Retinol products get a repair companion, labeled pregnancy-contraindicated and night-use only.
- Lock concentration, formula, and claims together — no mid-stream changes.
- Vet factories with the five-certificate check (production license + ISO 22716/GMP + third-party QC + safety assessment + filing certificate).
🔗 Ready to build an efficacy product that's "dosed right and compliant"? → https://www.quickoem.com/en/contact
Frequently Asked Questions
If I claim "5% niacinamide," does the formula really have to contain 5%?
Not higher is better, and you can't just make up a number. Niacinamide's evidence-backed effective range is 2%–5% — repeatable human studies support this for brightening, oil control, and barrier repair (evidence grade A–B). Above 10% it becomes more likely to irritate and redden, with diminishing returns. The real red line: the concentration you advertise must match what's actually in the filed formula. In the US, MoCRA (Modernization of Cosmetics Regulation Act of 2022) requires facility registration and product listing with the FDA; in the EU, you notify through CPNP under Regulation (EC) 1223/2009. Claiming "5% niacinamide" while the formulation only contains 1% is a misleading claim — a false-advertising and non-compliance risk that can trigger delisting, corrective action, or fines. My advice to brand owners: use a "classic active + one trend active" structure instead of chasing one big number. Niacinamide 3% + tranexamic acid 2% working together is steadier, cheaper, and far less likely to backfire than a solo "10% niacinamide" story.
If an ingredient sits below position 8 on the INCI list, is it "added but not enough"?
This is a widely-shared reverse-engineering rule of thumb — directionally right, but with exceptions. For most conventional ingredients: the INCI list runs highest-to-lowest concentration, and anything under 1% may be listed in any order. So sitting below position 8 usually means the level has dropped below 1%. "Contains" is not the same as "at an effective level" — like dropping one sugar cube into a whole pot of soup. The exceptions are peptides and certain high-potency actives: palmitoyl pentapeptide-4 works at just 0.001%–0.005%, and acetyl hexapeptide-8 (Argireline) as a solution at 5%–10%. These are naturally low in the list — you can't flatly call them "underdosed." The point isn't position; it's whether the effective concentration is reached. Every OEM checklist I hand clients includes this one line: verify each hero ingredient's minimum effective concentration, not just whether it "contains" it.
I want to add retinol — what concentration should I start at? Can I use it in autumn/winter?
Retinol is among the strongest evidence-backed anti-aging actives (grade A, RCT-supported), but it's also the biggest source of blowups. Concentration: start at 0.025%, and only after tolerance build up to 0.1%, 0.3%, and at most 0.5%. Don't launch straight at 1% — that's not potency, that's a burn. Cadence: begin 2–3 nights a week, increase frequency as tolerance builds; night use only (photolabile), with sunscreen mandatory during the day. Contraindications: avoid in pregnancy and lactation; don't stack with high-concentration AHA/acids (stimulation doubles). Autumn/winter is not just fine — it's actually better: UV is milder and barrier stress is lower, making it a good window to build tolerance, as long as you pair it with hydration and repair (niacinamide, ceramides). When I help OEM clients launch retinol, I always insist on a "repair companion" product rather than letting consumers apply raw retinol alone.
Is more ceramide always better? What ratio should a barrier-repair formula use?
No. It's about the biomimetic ratio, not the volume. In the "brick-and-mortar" model of the skin barrier, corneocytes are the bricks and intercellular lipids are the mortar — and the optimal molar ratio of ceramides : cholesterol : free fatty acids is 3:1:1 (Elias brick-wall model, industry consensus, evidence grade B–C). Dumping in a high concentration of ceramide alone throws the ratio off, spreads poorly, raises cost, and can actually slow repair. Effective ceramide level: 0.1%–1% is enough, when paired with cholesterol and free fatty acids at 3:1:1. Advanced stack: add madecassoside (0.1%–1%, anti-inflammatory and soothing), panthenol, and bisabolol for an "inflammation-first" soothing layer — calm first, then repair. Brands like CeraVe and La Roche-Posay built their entire sensitive-skin franchise on "repair the barrier first, then claim efficacy" — the same first principle I apply when designing sensitive-skin products for OEM clients.
How do I write efficacy claims without breaking the rules? Can I say "7-day brightening," "repairs sensitive skin," or "comparable to medical aesthetics"?
Mostly no. This is the number-one minefield in active skincare. I use a "four marketing-hype questions" self-check: What's the evidence source? Are the ingredients transparent? Is the time promise realistic? Is the comparison exaggerating across dimensions? Absolute/superlative and time-promise language: "7-day brightening," "28-day spot removal," "permanent," "best" — prohibited. Medical language: "cures," "repairs sensitive skin" (as a medical claim), "anti-inflammatory," "cosmeceutical" — cosmetics are not drugs; implying a medical effect is banned ("soothes" and "supports the barrier" are fine; "treats" and "cures" are not). "Comparable to medical aesthetics" / "replaces injections" — cross-dimensional exaggeration, prohibited. Under both MoCRA and EU 1223/2009 (with the Common Criteria in Commission Regulation 655/2013), every claim must have substantiation — brightening, anti-wrinkle, soothing, oil-control claims require supporting data filed at notification. Brands like The Ordinary and Paula's Choice can print specific claims precisely because a full substantiation file sits behind them, not because a copywriter dreamed it up.
Do brightening/sunscreen/acne products need special registration? How long does it take?
It depends on the claim — don't assume. Under China's 2021 Cosmetics Supervision and Administration Regulation, products split into two classes; the same "claim-first" logic applies when you map it to export markets: Watch the word "acne." Plain oil-control/anti-acne goes through general filing; but the moment an anti-acne product also claims "spot fading" or "whitening," it drops into the special-registration channel. Likewise "sunscreen" — even SPF 15 is special in most jurisdictions. Before you OEM, ask one question: "does my claim cross the special-cosmetics line?" That single question saves a huge amount of rework.
How much does a human efficacy evaluation cost? Can I skip it?
If you claim a benefit, you can't skip the substantiation. Human efficacy evaluation is the "evidence base" of any efficacy claim, and cost varies by test. A simple moisturizing/soothing lab-instrument test can run a few thousand dollars; anti-wrinkle or brightening tests need human subjects + instruments + a study period, typically in the $1,000–$12,000 range, with specialized claims costing more. When it's mandatory: claiming brightening, anti-wrinkle, firming, soothing, oil control, or moisturizing — the notification requires a claim summary (efficacy substantiation). When you can defer: only "cleansing" / "basic moisturizing" without specific numbers, backed by literature or raw-material data — but that window keeps shrinking. My advice: treat efficacy testing as a fixed cost of the product, and fold it into the OEM quote up front — not something you "remember" right before launch.
How do I balance concentration and cost? Does stacking actives equal a good formula?
Stacking ≠ good formula. This is the point I repeat most. A good formula is "precise and effective," not "more ingredients and bigger numbers." Three-layer structure: moisturizing base (occlusives + humectants + emollients) + active layer (effective concentrations met) + safety/stability layer (preservatives + buffering). All three must exist — drop one and something breaks. Cost logic: actives are usually the most expensive part. Blindly piling concentration can raise cost 30% while adding maybe 5% efficacy, and increase irritation risk. The best value is the "classic active + one trend active" structure, spending your budget where it counts. OEM parameters: typical MOQ 500–1,000 units, sampling 7–15 days, mass production 30–45 days, and seal a reference sample once the formula is locked. Concentration, formula, and claims must all be locked together — never change the concentration mid-stream and forget to update the claim.
Yes. For every claimed benefit we can arrange the supporting test file — an instrumental moisturizing or soothing test, or a full human study — and package the claim summary for FDA MoCRA listing and EU CPNP notification. We flag any claim that lacks evidence before it reaches your label.
Several. Salicylic acid is capped at 2% in leave-on acne products in many markets (the exact band varies by region), and retinol is capped or gated in some jurisdictions. Where a limit is unclear, we mark the claim "subject to local regulation" and confirm it during formulation.
We connect brands with factories that hold the right export credentials and support NMPA filing (China), CPNP notification (EU), and MoCRA facility registration plus product listing (US), with INCI, MSDS, stability, and microbial reports supplied.
Lock the claimed concentration, the filed formula, and the label claim together at briefing, and keep one reference sample. We treat the three as a single approved spec — changing one means re-approving all three.