Sensory Evaluation Record

This record answers whether a product's sensory profile still holds up against a reference — nothing more, nothing less. Using it as a general quality check without a genuine reference and controlled conditions produces a comparison that can't be relied on later.

BRCGS cl.5.1 expects sensory acceptability to be evidenced at shelf life points and change events, but the clause is silent on panel methodology — that discipline has to come from the conditions built into this template.

Sensory evaluation sits under a certification standard, an international methodology framework, and an EU regulatory driver it quietly supports — a panel can satisfy one without touching the others.

The fields are simple to complete. The four calls below are what actually decide whether the finished comparison means anything.

The failure modes below recur across sensory programmes and each one is visible directly in the conditions the template already asks the panel to confirm.

A producer switched to a lower-cost flavouring supplier and ran a sensory evaluation to confirm the product still matched. The samples were served identified — the panel was simply told which cup was the new supplier's — and Sample Order Randomised was answered "No" because the same two cups were always poured in the same order. Product Acceptable came back "Yes" and Difference From Reference Detected came back "No".

Complaints about a flatter flavour began arriving within weeks, concentrated in the batches made with the new supplier. A follow-up evaluation run genuinely blind, with samples randomised, found a clear difference the first panel had missed entirely — not because the change was undetectable, but because a panel told what to expect will usually confirm it.

Running the panel is the technical team's job. Making sure a marginal or biased result doesn't quietly support a supplier or process decision it shouldn't is the part that depends on someone tracking conditions across records.

KnowQuality

Holds the sensory evaluation record against the product register and shelf life study, and flags when panel conditions — blind, randomised, controlled — weren't actually met before a conclusion is relied on.

KnowTrain

Tracks panel member training and monitoring status, so "trained panel" on the record reflects a real, current qualification rather than a label carried over from a previous evaluation.

KnowComply

Keeps evidence that sensory conclusions supporting a supplier or process change were reached under defensible conditions, ready for the next audit.

Ella

Coordinates the crew across the evaluation, the shelf life study and the complaint trail, and holds every write for your approval before it touches a record.

Reference sample

A genuine, representative, well-characterised sample of the product as it should be, used as the fixed comparison point for every attribute a panel scores — without it, there is nothing valid to score against.

Blind tasting

A serving method where panellists do not know which sample is which, used to remove expectation bias from the result. Fully blind is the strongest form; partly blind and identified are progressively weaker.

Attribute scoring

Rating a product on individual dimensions — appearance, colour, aroma, texture, flavour, aftertaste, overall — separately, rather than collapsing the assessment into one overall impression.

Expectation bias

The tendency for a panellist who knows what change to look for, or expects a product to fail or pass, to report a result consistent with that expectation regardless of what they actually perceive.

Palate cleanser

A neutral food or drink, typically water or plain crackers, given between samples to prevent the flavour of one sample carrying over and biasing the assessment of the next.