SESH FRAGRANCE ACADEMY | ADVANCED
How do you correct one flaw without destroying a successful facet?
Follow sample introduction, column separation and detection. This GC schematic is not a SESH perfume test and does not independently establish sensory experience.
Open full-size image ↗Use hypothesis-led revisions, matched sensory comparisons and documented trade-offs.
Iteration starts with a diagnosis
'Too sharp' is incomplete feedback. Establish where the harshness occurs, whether opening, heart or dry-down, and how observers describe it: citrus bite, metallic edge or alcohol-associated sting. Collect structured input at fixed checkpoints. Only then form a single testable revision hypothesis instead of arbitrarily changing the entire blend.
Preserve the reference version
Freeze F01 as the reference and assign F02 a new version identifier with the exact change and rationale. Record sample dose, preparation age and environmental conditions. Blind-coded A/B tests reduce the temptation to perceive every newer revision as better. Any physical formula change requires rechecking restrictions, solubility and other technical controls.
Follow the route from the nasal cavity to the olfactory bulb. Smell is perception; it does not independently identify a perfume’s ingredient list.
Change one conceptual variable first
For an oversweet berry accord, vary perceived sweetness in one candidate while keeping the other conceptual axes as stable as possible. A second candidate can test added tartness separately. Changing sweetness, tartness and dose simultaneously may yield a better result but not explain why. Later, structured multi-factor designs can explore interactions when the simpler experiments are understood.
Optimize multiple attributes, not one
F02 might reduce unwanted sweetness yet erase the berry identity in the heart. Score brief alignment, core recognition, comfort, temporal continuity and appropriate strength separately. Define evaluation criteria before examining results. Improvement in one attribute does not constitute overall success if a mandatory requirement fails; record partial success honestly.
Three candidate revisions
Imagine F01 sweetness 5/5 and fruit identity 4/5. F02 reaches sweetness 2/5 but fruit identity falls to 1/5. F03 scores sweetness 3/5 and fruit identity 4/5. Under an educational pass rule of sweetness ≤3 and identity ≥3, only F03 meets both, despite F02 being less sweet. These are fictional sensory scores, not approved laboratory results.
Replication is part of an improvement claim
A single favorable test is not robust proof of improvement. Repeat at another session using comparable samples, dosing and timing, ideally with multiple observers. When observations conflict, report variability and investigate conditions rather than cherry-picking the best score. A useful reformulation repeatedly meets defined expectations.
Define the next-stage gate
Passing a sensory brief is a gate to further technical evaluation, not permission to manufacture or sell. Real changes may require recalculating restricted materials against final loading, reviewing base and packaging compatibility and conducting suitable stability and quality tests. Close the lesson with 'advance for verification' or 'revise again', not 'ready to launch'.
Case: optimizing the whole brief
For a fictional brief requiring sweetness ≤3 and fruit identity ≥3, versions F01 (5,4), F02 (2,1) and F03 (3,4) show why F03 wins the two-criterion sensory screen. That does not waive technical testing. The objective is to preserve effective attributes while correcting a specific defect.
Practical case and exercise
Build a version comparison table for F01, F02 and F03, including each hypothesis, strengths, weaknesses and go/no-go decision. Add a 5/30/120-minute evaluation plan and at least two explicit acceptance criteria.
Concept maps and key ideas
Hypothesis-led revision
- Describe the exact defect
- Identify when it appears
- Test one targeted change
- Preserve F01 control
Illustrative version scores
- F01: sweet 5 / fruit 4
- F02: sweet 2 / fruit 1
- F03: sweet 3 / fruit 4
- Goal: sweet ≤3 / fruit ≥3
Decision gates
- Sensory brief fit
- Repeatability check
- Independent safety review
- Further stability and compatibility
Scientific glossary
- Iteration
- Controlled improvement through a documented test hypothesis.
- Go/no-go gate
- Explicit proceed-or-revise decision based on criteria.
Further scientific reference
Reference starting points for fragrance terminology and verification: IFRA Fragrance Standards · PubChem
