The practical answer
Choose the sensory question before choosing the test. Descriptive profiling, difference testing and preference testing provide different information. Compare safe, approved prototypes under controlled conditions and record the full formula; tasting an unqualified raw powder is not an appropriate shortcut.
Decide what you need to learn
A formulator may need to identify an aftertaste, determine whether a supplier change creates a detectable difference, or understand which prototype consumers prefer. These are not interchangeable questions. ISO 13299 addresses establishing sensory profiles, while ISO 4120 addresses difference or similarity using a triangle test. [1] [2]
Write one primary question for each study. For example, an early development screen might describe bitterness, chalkiness, aroma and lingering mouthfeel using an agreed vocabulary. A later comparison might test a specific change against the approved reference. Avoid claiming consumer preference from a small internal descriptive panel.
Prepare the samples for a fair comparison
A blind code does not correct an uncontrolled formula. If one candidate uses more acid or sweetener, a perceived difference cannot be assigned solely to the magnesium ingredient. Hold the base constant for the initial comparison, then document any deliberate optimization as a separate development step.
- Use prototypes cleared for the intended sensory evaluation, with appropriate ingredient qualification and participant safeguards.
- Record formulation revision, ingredient lot, concentration, preparation, storage and serving conditions.
- Mask commercial names where practical and balance presentation order so expectations do not dominate the comparison.
- Use the same serving format and relevant elemental contribution, recording any difference in total ingredient mass.
- Plan how assessors report observations before the group discusses the samples.
Use the result for the right decision
Triangle testing can be unsuitable where strong lingering flavors make carryover difficult to manage; the ISO abstract specifically notes this limitation. [2] Ask a sensory specialist to select a design and analysis appropriate to the product instead of adopting a fixed panel size or decision threshold from an unrelated trial.
| Study question | Useful output | Do not infer |
|---|---|---|
| What attributes are present? | Defined descriptive profile and intensities | Automatic consumer preference |
| Can assessors distinguish a change? | Result interpreted under a suitable difference-test design | Which attribute caused the difference |
| Which product is preferred? | Preference evidence from the relevant target group | Clinical effectiveness or nutritional superiority |
| Does storage alter the experience? | Comparison with a defined reference and time point | Shelf life based on taste alone |
On smaller screens, scroll the table horizontally.
Keep formulation and commercial claims separate
A prototype can improve after flavor-system optimization without establishing that the raw material itself has a universally better taste. State the comparison conditions when sharing results with purchasing or a customer. “Preferred in this tested formula” is a narrower claim than “tasteless” or “best for every beverage.”
Retain individual observations, the reference formulation and the analysis plan. Combine the sensory result with composition, physical stability and relevant safety evidence before approving the ingredient. This workflow supports product development and does not ask customers to ingest unapproved samples, test therapeutic effects or use sensory impressions as proof of purity.
Sources & further reading
- ISO 13299:2016: Establishing a sensory profile, official abstract
- ISO 4120:2021: Triangle test, official abstract