Headspace vs On-Skin: Why Panel Results Disagree
A fragrance can show strong volatile release in a headspace vial and still be rated quiet on skin, or measure flat in the lab and perform well at home. Teams then argue about which result is right, when the methods answer different question
A fragrance can show strong volatile release in a headspace vial and still be rated quiet on skin, or measure flat in the lab and perform well at home. Teams then argue about which result is right, when the methods answer different questions.
This article explains what headspace analysis and skin panels each measure, why their rankings diverge, and how to build a fragrance performance testing programme that uses both without pretending they are interchangeable.
Table of Contents
- What headspace analysis measures
- What skin does to a fragrance
- Where the two results diverge
- Designing a panel that can be compared
- Reconciling the two datasets
- Practical guide: a two-method performance protocol
- Frequently asked questions
- Conclusion
What headspace analysis measures
The vapour above the sample, not the smell in the room
Headspace analysis samples the volatiles that partition from a liquid or solid into the gas above it, reported as a concentration in air after GC separation. In fragrance performance testing it describes release from a matrix under controlled conditions, not perception, which depends on odour thresholds, mixture effects and adaptation.
The partition coefficient between sample and air is temperature-dependent, so a vial at 25 °C and a forearm at 33 °C are not comparable. Matrix matters too: ethanol, water and oils change activity coefficients, and a 20% alcoholic base releases differently from the same concentrate in a lotion.
Static, dynamic and extracted headspace
Static headspace waits for equilibrium and samples a fixed volume. Dynamic purge-and-trap sweeps gas through the sample and concentrates trace analytes. Solid-phase microextraction adsorbs analytes onto a fibre; it is sensitive but selective and can saturate, so the profile is not a neutral census of the vial.
Why the instrument signal is not an odour intensity
Detector response and odour threshold both differ by compound, by orders of magnitude. A tall peak may contribute little if its threshold is high, while a small peak can dominate. The useful derived value is the odour activity value, concentration divided by threshold, though it ignores suppression and synergy.
What skin does to a fragrance
Heat, oil and a reactive surface
Skin is not an inert blotter. It sits around 32 to 35 °C, carries sebum and sweat, and presents a slightly acidic film. Materials partition into sebum, bind to proteins and evaporate at a rate that changes as the film changes; some are absorbed and released slowly, which is why a musk can be faint at first and present the next morning.
Skin flora and surface chemistry can also transform materials. Aldehydes may react with amines, esters may hydrolyse, and oxidation products can appear on skin that were not in the bottle. Two wearers with different skin can therefore produce different drydowns from the same dose.
Dose and application are part of the measurement
A spray delivers a mass that depends on the valve and actuator and the distance from the skin, so two sprays at 10 cm and at 25 cm are different experiments. For fragrance performance testing, fix dose, site and technique before the first panel, or results cannot be compared across sessions.
Where the two results diverge
Headspace over-weights the volatile fraction
Headspace is biased toward materials that enter the gas phase readily: citrus, green, aromatic and light floral notes. Those materials create the opening, so a headspace profile looks impressive for a top-heavy fragrance. On skin, they evaporate within minutes and the panel reports a short life.
Skin under-reports the substantive fraction
The opposite bias applies to heavy materials. Musks, resins and woody ambers have low vapour pressure, so their headspace concentration is small even when they carry the drydown. A headspace-only evaluation can rank a long-wearing fragrance as weak because its persistent compounds barely reach the gas phase.
Panel effects that look like product effects
Adaptation is the largest. A panellist who smells the same fragrance for two minutes stops registering it and may rate a later sample higher simply because it is different. Order, room carryover, humidity and other fragrances on clothing all move scores, so a panel without randomisation and a blank control can produce a difference that is purely methodological.
| Method | What it measures | Blind spot | Best used for |
|---|---|---|---|
| Static headspace GC | Equilibrium vapour concentration | Ignores thresholds and skin | Release profile, batch comparison |
| SPME headspace | Adsorbed volatile profile | Fibre selectivity and saturation | Screening volatile differences |
| Blotter panel | Perceived intensity on paper | No heat, oil or absorption | Early structure screening |
| Skin panel | Perceived intensity on wearers | High variance, adaptation | Claim support and final selection |
Designing a panel that can be compared
Blind, randomised and replicated
Samples should be coded and presented in a randomised order, with each panellist assessing each sample more than once across sessions. Include a hidden duplicate to estimate within-panellist noise, and a blank or reference to anchor the scale. Without these, a difference between two samples cannot be separated from the difference between two days.
Control the room and the nose
Run the panel in a ventilated, fragrance-free room at stable temperature. Limit samples per session, allow minutes between them, and use clean air rather than coffee beans to reset the nose. Record what panellists ate, drank or wore, because those variables move scores more than formulators expect.
Measure time, not only intensity
A single intensity score collapses the whole wear into one number. Rate at fixed intervals, such as 0, 1, 2, 4 and 8 hours, and record detection separately from intensity: a fragrance can be faint but detectable, or strong but gone. The curve is the claim, not the peak.
Reconciling the two datasets
Use headspace to explain, panels to decide
When a fragrance performance testing panel result surprises you, go back to the headspace data and ask which compounds changed. If the top notes increased and the base decreased, the "shorter" verdict has a chemical explanation, and that explanation is what lets you reformulate rather than guess.
Report distributions, not just averages
Ten panellists who split five to five have two groups, not an average opinion. In fragrance performance testing, report the spread and check whether it tracks age, skin type or gender: a consistent split is a segmentation finding, while random scatter means repeating the session. Our note on triaging weak-spray complaints applies the same logic to customer emails.
| Observation | Likely cause | Next test |
|---|---|---|
| Strong headspace, weak skin panel | Top-heavy formula with little substantive base | Extend the panel to 8 hours; compare base materials |
| Weak headspace, strong skin panel | Low-volatility musks or resins carrying the wear | Solvent-extract skin after 8 hours |
| Scores fall across a session | Adaptation or room carryover | Randomise order and add clean-air breaks |
| One panellist always differs | Skin chemistry or anosmia to a material | Recruit a larger panel; test that material alone |
| Result changes between batches | Raw material or alcohol variation | Headspace comparison against retained reference |
Practical guide: a two-method performance protocol
Step 1 — Write the claim before choosing the method
Decide whether the claim is about opening impact, eight-hour persistence or room presence. Each maps to a different method, and a protocol built before the claim will usually measure the wrong thing.
Step 2 — Fix dose, site and application technique
Specify spray count, distance, skin site and time of day. These are experimental parameters, not housekeeping, and changing them invalidates comparison with earlier data.
Step 3 — Capture a headspace baseline for every batch
Run the same method on the approved reference and the new batch, and compare the profiles rather than a single total. Store the chromatograms with the batch record.
Step 4 — Run the skin panel blind and replicated
Code samples, randomise order, include a duplicate and a reference, and rate at fixed intervals. Record individual scores, not only the mean.
Step 5 — Reconcile the two results before reporting
Ask whether the headspace change explains the panel change. If it does not, suspect the method, the panel or the substrate before reformulating the juice.
Step 6 — Keep the protocol with the approved version
File the method, the panel composition and the raw scores with the formula version, so a later substitution can be tested the same way. Our retail and sample pack planning note shows where these records sit in a launch schedule.
Frequently asked questions
What is private label perfume manufacturing?
The product carries your brand while a manufacturer handles the formula and filling. You approve the scent and claims; they hold the batch documentation.
What is a typical minimum order?
Minimums are a packaging question. Once the bottle and pump are chosen, the supplier can quote the smallest run that covers their setup.
How long does development take?
Development takes weeks, dominated by sampling and stability rather than perfumery. Bottle lead time is the usual bottleneck.
Why does my fragrance smell stronger in the lab than at home?
Usually because the lab tests on paper or at a fixed dose in a still room. Skin temperature, movement, humidity and spray technique all change release.
Can headspace analysis replace a panel?
No. It measures vapour concentration, not perception, and it under-represents low-volatility materials that carry the drydown. It is a diagnostic tool, not a substitute for wear testing.
How many panellists do I need?
Enough to see the effect against the noise, which depends on the question. Small panels can guide development; a measured difference claim needs more people and replication.
Should I test on paper or on skin?
Paper for structure and early screening, skin for the decision. A skin product approved only on paper has not been tested in the condition it will be used.
Why do two people disagree about the same fragrance?
Skin chemistry, body temperature, anosmia and application habits all differ. A consistent split is information; random disagreement means the test needs more control.
Conclusion
Headspace and skin panels disagree because they measure different things: the vapour above a matrix, and a person's perception over time. Neither is wrong; treating either as the truth produces a chemically elegant fragrance nobody can smell, or a consumer favourite nobody can explain.
The practical answer is to state the claim first, fix the dose and the method, and use the instrument to explain what the panel found. That combination is slower than a single test, but it is the only version that survives a second batch, a new supplier or a customer complaint.
Discuss your fragrance programme with our team:
WhatsApp: +86 17324762943
Email: adon@customperfumeoem.com
Website: customperfumeoem.com
