TREN
🧪 ESANS.COM.TR ACADEMY — Technical Formulation Portal
🧪
Esans.com.tr Academy
Technical Formulation Portal
← All Articles
Scent Training ↗

Blind Panel Design: How to Run Unbiased Fragrance Sensory Tests

A blind panel removes brand, packaging and price from the equation — leaving only the scent. Learn how to design, code and run rigorous blind sensory tests for fragrance evaluation.

Esans.com.tr Academy ·✍️ Esans Academy Technical Team ·~7 min read
01

What Is a Blind Panel and Why Is It Essential?

The nose can be fooled. Brand, bottle, price tag — even a comment from the person sitting nearby — can shape a decision before a single sniff is taken. A blind panel is a sensory test in which the evaluator is given no information whatsoever about the sample. The sole objective: eliminate bias and leave only the scent.

Trusting your own nose before bringing a formula to market is not enough. You are in love with what you have created; that love blinds not your eyes but your nose. A blind test reverses that blindness.

There are two fundamental blinding layers. If the evaluator does not know the identity of the sample, it is single-blind; if neither the evaluator nor the person administering the test knows the codes, it is double-blind. Double-blinding also eliminates the unconscious facial expressions and vocal cues of the presenter.

A blind panel does not answer "is this beautiful?" — it answers "is this noticeable, which is preferred, and why?" Clarify the question and the panel will clarify itself.
02

Setting Up a Panel: Venue, People, Conditions

A good panel starts with a good environment. Fragrance evaluation is a precise measurement; if the surroundings are compromised, the data is worthless.

The space must be odour-free. Coffee, cigarette smoke, food, scented cleaning products — even fresh paint — can distort judgement. Ventilation should be simple and constant. Lighting should be neutral; coloured lighting has an indirect effect on perception.

The number of panellists depends on the objective. For a quick preliminary screening, 5–8 people is workable; if you need statistical confidence, 20 or more is preferred. With trained noses (see Training Your Nose: Scent Recognition and Memory), even a small panel is powerful; for an untrained consumer panel, increase the numbers.

Panellists should be neither hungry nor overly full, must not have applied fragrance before the session, and should avoid coffee beforehand. Before each session, stepping outside into clean air or smelling the skin of one's own forearm is a small but effective reset for the nose.

Tip: Schedule the test session for early in the day. The nose tires as the day progresses; discriminatory ability drops from the third sample onwards in the afternoon.
03

Sample Coding and Blinding Technique

Coding is the heart of blinding. An incorrectly coded sample is a beam of light that pierces the blindfold.

Label samples with three-digit random codes (e.g. 482, 137, 905). Avoid sequential codes such as A/B/1/2; the human brain interprets a sequence as "the first one is the main product." Bottles must be the same colour, the same size and the same volume — even a difference in the glass can serve as a clue.

Randomise the presentation order for each panellist. If everyone smells samples in the same sequence, "first-sample bias" kicks in. The first sample assessed tends to have an advantage; distribute this effect by rotating the order.

  1. Preparation

    Present each sample diluted to the same concentration in the same carrier. If one is concentrated and another diluted, the test measures concentration, not fragrance design.

  2. Coding

    Assign a three-digit random code. Write the code–formula key on a separate sheet and keep it away from the panellists. In a double-blind setup, the presenter must not see this key either.

  3. Blotter preparation

    Mark test strips (blotters) with the sample code and dip the scent band from the end furthest from the mark (see How to Use a Test Strip Correctly).

  4. Presentation

    Hand out samples in a mixed order. Have each panellist smell each one and fill in the form; allow a neutralising break between samples.

  5. Reveal

    Once all forms have been collected, decode the samples. Data first, identity second. This sequence is sacred.

Classic mistake: writing "new formula" in your own handwriting while coding a sample. At that moment, blinding ends. Only three digits should appear on the label.
04

Test Types and Reading the Data

The question you ask determines which test you run. The wrong test produces the wrong answer dressed up as the right one.

Test TypeWhat It MeasuresUse CaseMin. Panellists
Triangle TestIs there a detectable difference between two samples?Checking whether a formula change is noticeable~12–24
Paired PreferenceWhich sample is preferred?A/B formula comparison~15–30
RankingPreference order among multiple samplesScreening more than two variants~10–20
DescriptiveScent profile, note intensity, perceived longevityProfiling with a trained panel~5–10 (trained)

The triangle test is the most robust difference test: you present three samples, two of which are identical, and the panellist identifies the odd one out. Because the probability of guessing correctly is 1/3, statistics filter out chance. It provides the most honest answer to the question "Is a new raw material changing the scent of the existing formula?"

When asking panellists for notes in a descriptive test, a shared vocabulary is essential. "Nice" is not data; "resinous, balsamic, lightly smoky" is data. The Guide to Describing a Scent in Words can help align your panel's vocabulary.

Tip: When measuring longevity and projection (sillage) with a panel, bear in mind that perceived longevity depends not only on the fragrance oil concentration but also on the volatility of the raw materials. A citrus-heavy sample at a high concentration may dissipate quickly; an ambermusk-heavy sample at a lower concentration may linger. For this reason, do not present time windows as absolutes; measure comparatively under identical conditions and record the panel's perception.
05

Common Mistakes and Frequently Asked Questions

The value of a blind panel comes from its discipline. A single weak link corrupts the entire chain. Here are the most common stumbling points in practice.

Presenting samples at different stages of maturity. Placing a freshly mixed formula side by side with one that has undergone two weeks of maceration is unfair. Maceration is not merely waiting; the scent settles through alcoholfragrance oil interaction and esterification, and the raw "alcohol bite" softens. Samples should be at the same stage of maturity after maceration (at room temperature, ~15–20 °C), protection from light, and cold filtration.

Failing to standardise dilution. Samples must be presented in the same carrier at the same concentration. Note: adding too much water to high-strength ethanol impairs solubility, causing cloudiness and phase separation. Keep the water ratio low, start with a small amount, test for clarity, and use an appropriate perfumer's alcohol/solvent if needed. A cloudy sample also provides a visual cue during the panel, piercing the blind.

Continuing with an exhausted nose. Discriminatory ability declines after 4–6 samples. Take a break and split the session.

How many samples can I blind-test in a single session?
With an untrained panel, 3–5 samples per session is the safe limit; beyond that, olfactory fatigue reduces reliability. Trained noses can handle a few more with breaks. If you have many variants, use an elimination-round structure: first use a ranking test to select the top three, then move to paired preference for the final round.
Does a blind test result guarantee IFRA compliance?
No. A blind panel measures the perception of a scent, not its safety. IFRA limits are based not on the total fragrance oil concentration but on the individual substances/allergens within the fragrance oil and the product category (leave-on/rinse-off). Do not make generalisations such as "any fragrance oil is safe up to this concentration." For the safety of your formula, read the IFRA certificate/compliance declaration for the fragrance oil and calculate accordingly.
I work alone — how do I set up a double-blind test?
A true double-blind is difficult with a single person, but you can get close. On a separate day, when you are calm and clear-headed, prepare the samples with three-digit random codes and seal the code–formula list in an envelope, then put it out of sight. On the test day, approach yourself as if you were a different person, fill in the form, and then open the envelope. To increase impartiality, ask an acquaintance to do the coding if at all possible, so that you only ever see the numbers. The rest comes down to your own discipline.

Continue

🛒 Related Product
Parfüm Esansları
Browse products →
🧪 Related Tool
Laboratory & Search
Open calculator →

esans.com.tr

Explore →