COSMIC ROI walkthrough
Language
Total time saved so far23 hours2 days faster TTM
What COSMIC actually saves you, step by step.
COSMIC is a co-formulator: it gathers every relevant data point - INCI regulatory, botanical pathways, peptide stability, HLB/HLD, microbiome, claims - and proposes a complete picture to the formulator, who keeps full control of the decision. The expert validates and signs off in minutes instead of spending days on database lookups, supplier round-trips, and spreadsheet juggling.
Six tools, one pipeline. Each step shows the traditional workflow side-by-side with COSMIC, with honest ranges for both junior (0-3 years) and senior (5+ years) formulators. The total counter at the top updates as you scroll through.
Step 1
INCI Explorer
Single-INCI deep dive — function, regulatory, suppliers
Time saved this step
23h
TTM days saved
2d
Purpose
Research one INCI ingredient — function, regulatory status, max use levels, compatibilities, supplier options.
Time saved vs Traditional21h – 26h
Traditional
effective costHours
21h – 26h
| Sub-task | Time |
|---|---|
| Search CosIng database | 8 min |
| Search supplier TDS / specs | 20 min |
| Cross-check REACH on ECHA CHEM | 45 min |
| Read SCCS opinion if applicable | 60 min |
| Internal catalog + colleague consult | 40 min |
| Per-INCI subtotal × 8 INCIs | 23h |
With COSMIC
✓ savingHours
15min – 30min
↧Bundled into: 1 COSMIC INCI Explorer query
| Sub-task | Time |
|---|---|
| Search CosIng database | incl. |
| Search supplier TDS / specs | incl. |
| Cross-check REACH on ECHA CHEM | incl. |
| Read SCCS opinion if applicable | incl. |
| Internal catalog + colleague consult | incl. |
| Per-INCI subtotal × 8 INCIs | 25 min |
What COSMIC delivers hereclick to expand
Features used
Joined CosIng + REACH/ECHA + SCCS opinions in one query
Compatibility flags + ontological class + HLB / pKa
Supplier options surfaced from internal catalog
Data the formulator gets
- Function (CosIng) + canonical role tag (15 categories)
- EU Annex status + max % rinse-off / leave-on / oral
- REACH registration + ECHA dossier links
- SCCS opinion ref + restricted conditions
- PubChem CID, SMILES, MW, XLogP, TPSA, pKa
- HLB (Davies / Rule-of-Mixtures), Hansen δD/δP/δH
- Function tags (15 canonical roles), allergen / GHS flags
- Supplier options + product co-occurrence stats (~318k retail products)
Time saved this step
23h saved · 98% faster
Traditional
23h
With COSMIC
24min
Manually consolidating CosIng, ECHA REACH dossiers, SCCS opinions, and supplier TDS for one INCI is mostly tab-switching. INCI Explorer returns the joined record in one query, with the regulatory annex, max %, REACH status and supplier matches inline.
Step 2
Botanical Explorer
Botanical extracts — INCI, actives, suppliers, claims
Time saved this step
15h
TTM days saved
4d
Purpose
Identify and validate botanical extracts — INCI mapping, active compounds, claims potential, sourcing.
Time saved vs Traditional12h – 18h
Traditional
effective costHours
13h – 18h
| Sub-task | Time |
|---|---|
| Literature search (PubMed, Google Scholar) | 4h |
| Supplier outreach for spec sheets | 3d |
| INCI confirmation + synonym mapping | 2h |
| Active marker compound verification | 3h |
With COSMIC
✓ savingHours
30min – 45min
↧Bundled into: 1 COSMIC Botanical Explorer query
| Sub-task | Time |
|---|---|
| Literature search (PubMed, Google Scholar) | incl. |
| Supplier outreach for spec sheets | incl. |
| INCI confirmation + synonym mapping | incl. |
| Active marker compound verification | incl. |
What COSMIC delivers hereclick to expand
Features used
Curated taxonomy + INCI mapping (multilingual aliases)
Active marker compounds with PubChem / KEGG / Reactome links
Validated supplier shortlist + claim references (LOTUS, COCONUT)
Data the formulator gets
- Latin name, family, genus, plant part — full taxonomy
- Multilingual INCI aliases (camomilla, manzanilla, Kamille…)
- Active marker compounds with PubChem CIDs
- Compound classes (flavonoids, terpenoids, polyphenols…)
- KEGG pathways + Reactome gene sets
- ChEMBL / OpenTargets gene targets — claim back-up
- Geographic distribution (countries grouped by continent)
- LOTUS / COCONUT cross-link, supplier matches
Time saved this step
15h saved · 97% faster
Traditional
15h
With COSMIC
30min
Botanical work is mostly waiting on suppliers. The Botanical Explorer skips the round-trip: INCI mapped to genus/species, active markers pre-resolved to compound IDs, and supplier matches with claim references in one place.
Step 3
Peptide Explorer
Bioactive peptides — sequence, mechanism, dose, stability
Time saved this step
16h
TTM days saved
4d
Purpose
Select bioactive peptides — sequence, mechanism, efficacious dose, stability, supplier.
Time saved vs Traditional14h – 18h
Traditional
effective costHours
14h – 18h
| Sub-task | Time |
|---|---|
| Mechanism research (literature) | 4h |
| Efficacy dose from clinical studies | 4h |
| Stability / compatibility check | 3h |
| Supplier qualification + COA review | 2d |
With COSMIC
✓ savingHours
15min – 30min
↧Bundled into: 1 COSMIC Peptide Explorer query
| Sub-task | Time |
|---|---|
| Mechanism research (literature) | incl. |
| Efficacy dose from clinical studies | incl. |
| Stability / compatibility check | incl. |
| Supplier qualification + COA review | incl. |
What COSMIC delivers hereclick to expand
Features used
Sequence + mechanism cards with literature back-links
Validated dose ranges from clinical & vendor studies
pH/temperature stability + formulation interaction warnings
AMP cross-link to antimicrobial peptide database
Data the formulator gets
- Sequence (1-letter + 3-letter), MW, length, isoelectric pH
- Mechanism of action with PubMed back-links
- Validated efficacious dose range (% w/w)
- pH window + temperature stability bands
- Formulation incompatibilities (oxidants, surfactants, metals)
- AMP database cross-link (antimicrobial activity, MIC)
- Suggested pairings (carriers, peptide-peptide synergy)
- Suppliers + COA-grade specs
Time saved this step
16h saved · 98% faster
Traditional
16h
With COSMIC
20min
Peptide selection is bottlenecked by clinical-dose research and stability checks. Peptide Explorer pre-resolves sequence, mechanism, validated dose range and the formulation guard-rails (pH bands, oxidative incompatibilities) so the formulator only validates, not researches.
Step 4
Emulsion Design
Phase architecture + surfactant blend
Time saved this step
12h
TTM days saved
3d
Purpose
Design the emulsion architecture — phase composition, emulsifier system, % ranges.
Time saved vs Traditional9.0h – 15h
Traditional
effective costHours
13h – 18h
| Sub-task | Time |
|---|---|
| Literature search for benchmarks | 1h |
| Excel-based phase calc, iterations | 2h |
| Lab pre-tests for emulsion type | 3d |
With COSMIC
✓ savingHours
3.0h – 4.0h
↧Bundled into: 1 COSMIC Emulsion Design step
| Sub-task | Time |
|---|---|
| Literature search for benchmarks | incl. |
| Excel-based phase calc, iterations | incl. |
| Lab pre-tests for emulsion type — focused validation only | 2.5h |
What COSMIC delivers hereclick to expand
Features used
AI-suggested phase architecture given target product type
Auto-calculation of phase % ranges + viscosity class prediction
HLB-paired surfactant blends fed from the catalog
Data the formulator gets
- Suggested phase split (oil / water / emulsifier %)
- Recommended emulsion type (O/W, W/O, multiple, micro)
- HLB target band + paired surfactant blend
- Predicted viscosity class & sensory profile
- Pre-flagged incompatibilities (electrolyte tolerance, pH limits)
- Recommended preservative system pairing
Time saved this step
12h saved · 77% faster
Traditional
15h
With COSMIC
3.5h
Designing an emulsion from scratch eats lab time on type validation. Emulsion Design proposes the phase architecture and the surfactant blend up-front, so the lab work shifts from "will it emulsify?" to focused stability checks.
Step 5
HLB / HLD Calculator
Required HLB + HLD parameters
Time saved this step
6.8h
TTM days saved
1d
Purpose
Calculate required HLB of the oil phase OR HLD for surfactant systems.
Time saved vs Traditional5.6h – 7.9h
Traditional
effective costHours
6.0h – 8.0h
| Sub-task | Time |
|---|---|
| Manual Excel HLB calc per oil component | 2h |
| HLD parameter lookup (Cc, T, EACN values) | 2h |
| Iteration with formulation | 3h |
With COSMIC
✓ savingHours
9min – 24min
↧Bundled into: 1 COSMIC HLB / HLD step
| Sub-task | Time |
|---|---|
| Manual Excel HLB calc per oil component | incl. |
| HLD parameter lookup (Cc, T, EACN values) | incl. |
| Iteration with formulation | incl. |
What COSMIC delivers hereclick to expand
Features used
Auto-fetch HLB values from INCI database (Davies + Rule-of-Mixtures fallback)
HLD parameters pre-stored for common surfactant systems
Real-time recalculation as the formula changes
Data the formulator gets
- Required HLB of the oil phase (auto-pulled per component)
- Davies group-contribution prediction (when catalog gap)
- Rule-of-Mixtures fallback on fatty-acid profiles
- HLD parameters: Cc, T, EACN, electrolyte effect
- Recommended surfactant blend to match target HLB
- Live recalculation as the formula composition changes
Time saved this step
6.8h saved · 96% faster
Traditional
7.0h
With COSMIC
15min
Required HLB is a calculation, but it's a tedious one when oils have to be looked up one by one. HLB/HLD pulls each component's HLB from the catalog (or predicts it via Davies / Rule-of-Mixtures), and recalculates live as the formula moves.
Step 6
Coformulator
Brief → ranked formula + 11-engine assessment
Time saved this step
26h
TTM days saved
5d
Tip: numbers above reflect the "Start from zero" flow. Cloning has its own row below.
Purpose
Generate full formula draft from a brief.
Time saved vs Traditional20h – 33h
Traditional
effective costHours
30h – 40h
| Sub-task | Time |
|---|---|
| Concept ideation + benchmark survey | 6h |
| First draft formulation | 4h |
| Iteration with assessment (pH, stability, regulatory) | 20h |
| Documentation | 4h |
With COSMIC
✓ savingHours
7.0h – 10h
↧Bundled into: 1 COSMIC Coformulator run (brief → ranked draft + 11 engines)
| Sub-task | Time |
|---|---|
| Concept ideation + benchmark survey | incl. |
| First draft formulation | incl. |
| Iteration with assessment (pH, stability, regulatory) — human review | 5h |
| Documentation | incl. |
What COSMIC delivers hereclick to expand
Features used
Brief → ranked INCI suggestions via XGBoost ranker
TAG-generated full formula draft (composition + % + phases)
11 assessment engines validate output (pH, stability, regulatory, microbiome, …)
Auditable trail + technical claims dossier on export
Data the formulator gets
- Ranked INCI shortlist (XGBoost ranker over 200k formulas)
- Full draft: phases A/B/C/D, % w/w, mixing order
- Predicted pH, HLB target, viscosity class
- Stability flags (oxidative, photo, microbial)
- Regulatory check across EU / US / CN / JP / KR / IN / AU / BR
- Virtual-Toxicology (VTox) screen + comedogenic estimate
- Microbiome-safety scoring + preservative spectrum
- Antimicrobial protection against ISO 11930 panel
- Auditable claims dossier + token-cost ledger on export
Time saved this step
26h saved · 76% faster
Traditional
34h
With COSMIC
8.0h
Coformulator is exactly that — a CO-formulator. It returns the complete knowledge stack (composition, regulatory, safety, stability, microbiome, claims) for the human formulator to validate and decide. The expert keeps control; COSMIC removes the hours spent gathering and cross-referencing data.
Step 7
Cloning Formulator
Reverse-engineer a benchmark into a defensible formula
Time saved this step
37h
TTM days saved
8d
Purpose
Reverse-engineer a benchmark product into a defensible formulation.
Time saved vs Traditional28h – 47h
Traditional
effective costHours
40h – 55h
| Sub-task | Time |
|---|---|
| Reverse-engineer INCI list from label | 30 min |
| Source equivalent INCIs from internal catalog | 1h |
| Estimate concentrations from INCI order + lit. | 1h |
| Lab tests for matching texture/sensory | 7d |
| Iterate to match benchmark | 4d |
With COSMIC
✓ savingHours
8.0h – 12h
↧Bundled into: 1 COSMIC Cloning step (INCI ingest + ML pct + GNN compat.)
| Sub-task | Time |
|---|---|
| Reverse-engineer INCI list from label | incl. |
| Source equivalent INCIs from internal catalog | incl. |
| Estimate concentrations from INCI order + lit. | incl. |
| Lab tests for matching texture/sensory — focused validation only | 1d |
| Iterate to match benchmark — focused only | 4h |
What COSMIC delivers hereclick to expand
Features used
Input: benchmark INCI list (label or photo OCR)
Estimated concentrations via ML on similar formulas database
Suggested supplier matches + GNN-validated compatibility
Auto-paired HLB/HLD + microbiome + stability flags
Data the formulator gets
- Estimated % w/w per INCI (XGBoost ranker on benchmark library)
- Suggested catalog-equivalent INCIs (with cost / supplier)
- GNN-validated INCI–INCI compatibility map
- Reverse-engineered phase split (A/B/C/D)
- HLB/HLD pairing already aligned with the benchmark
- Stability + microbiome + antimicrobial guard-rails
- 11-engine assessment on the cloned formula (same as Coformulator)
Time saved this step
37h saved · 79% faster
Traditional
46h
With COSMIC
9.5h
Cloning a benchmark is mostly guessing concentrations and chasing texture matches in the lab. Cloning Formulator estimates concentrations from a similar-formulas ML model and pre-validates compatibility on a GNN, so the lab work compresses to focused validation.
Step 8
Odor Control
Deodorant effectiveness — does this mix actually neutralize body odor?
Time saved this step
19h
TTM days saved
7d
Purpose
Evaluate whether a deodorant / odor-control mix is effective: which body-odor pathways it covers (thiol, SCFA, steroid, lipase, urease), per-organism log-reduction, and where the formula is weak.
Time saved vs Traditional14h – 26h
Traditional
effective costHours
20h – 30h
| Sub-task | Time |
|---|---|
| Literature review on body-odor pathways | 4h |
| Map active candidates to pathways (alcohols, AMPs, enzyme inhibitors, fragrance maskers) | 4h |
| Design preservation + antimicrobial system for axilla microbiome | 4h |
| Lab challenge / panel / instrumental odor testing | 5d |
| Iterate to converge on an effective mix | 2d |
With COSMIC
✓ savingHours
4.0h – 6.0h
↧Bundled into: 1 COSMIC Odor Control assessment (5 pathways + 8 organisms + ISO 11930 pre-screen)
| Sub-task | Time |
|---|---|
| Literature review on body-odor pathways | incl. |
| Map active candidates to pathways | incl. |
| Design preservation + antimicrobial system for axilla microbiome | incl. |
| Lab challenge / panel testing — focused validation only | 1d |
| Iterate to converge — focused only | 4h |
What COSMIC delivers hereclick to expand
Features used
Per-pathway score 0-100 (thiol-bearing volatiles, SCFA oxidation, steroid byproducts, lipase, urease)
API (Antimicrobial Protection Index) against Staph hominis, Corynebacterium, C. acnes, M. furfur, S. aureus, E. coli, P. aeruginosa, C. albicans
Per-organism log-reduction with top contributors (which INCIs are doing the work, with their MIC)
ISO 11930 pre-screen against the 8 reference pathogens — runs the same logic as the Antimicrobial Preservation chapter
Booster suggestions when a pathway score is weak (e.g. add Triethyl Citrate for lipase, Zinc Ricinoleate for SCFA, Glyceryl Caprylate for thiol)
Data the formulator gets
- Pathway-by-pathway breakdown: which odor mechanism the formula covers and which it misses
- API 0-100 score + tier (Excellent / Good / Weak / Poor)
- Per-organism log-reduction estimates with the contributing INCIs
- Booster recommendations to close gaps
- ISO 11930 pre-screen result so the deodorant doubles as its own preservative system
- Formula-specific advice (natural / aluminum-free / alcohol-free deodorants)
Time saved this step
19h saved · 79% faster
Traditional
24h
With COSMIC
5.0h
Asking 'will this deodorant mix work?' traditionally means a literature scan on body-odor pathways, then panel testing or instrumental odor measurement, then iterating. COSMIC's Odor Control assessment scores the formula against each odor pathway deterministically and tells the formulator exactly where the mix is weak before any panel — so iteration converges in a fraction of the time.
What that looks like in practice
Time saved per tool (junior view)
Light purple = traditional cost · saturated purple = COSMIC saving
INCI Explorer
23h saved
24min → 23h
Botanical Explorer
15h saved
30min → 15h
Peptide Explorer
16h saved
20min → 16h
Emulsion Design
12h saved
3.5h → 15h
HLB / HLD Calculator
6.8h saved
15min → 7.0h
Coformulator
26h saved
8.0h → 34h
Cloning Formulator
37h saved
9.5h → 46h
Odor Control
19h saved
5.0h → 24h
COSMIC timeTime saved vs TraditionalTraditional total
Junior view
Scenario A — Full new formulation
Steps 1–5 + Coformulator (6a)
Hours saved
97
days faster TTM
19
Across the 6 tools used in this scenario.
Junior view
Scenario B — Cloning benchmark
Steps 1–2 + 5 + Cloning Formulator (6b)
Hours saved
80
days faster TTM
15
Across the 4 tools used in this scenario.
Junior view
Scenario C — Deodorant brief
Steps 1–2 + 4–5 + Coformulator + Odor Control
Hours saved
116
days faster TTM
23
Across the 6 tools used in this scenario.
Ready to see the same on your own briefs?
Request a demo - we'll run one of your real formulas through COSMIC.
Numbers are honest mid-range estimates for B2B cosmetic formulation work. Each side cites specific sub-tasks; the COSMIC values reflect features actually shipped in the product.