Setting Up a Honey Tasting Panel: Training, Protocols and the Sensory Wheel
How to recruit and train tasters, control the tasting environment, and use a standard sensory wheel and scoring sheet to turn subjective honey flavour into reliable data.
Why casual tasting is not the same as a tasting panel
Anyone can taste a spoonful of honey and say whether they like it, but that kind of impression does not travel well between people, and it does not repeat reliably over time even for the same person. A proper tasting panel exists to convert that subjective impression into something closer to data: reproducible scores, a shared vocabulary, and enough structure that a batch tasted in March and again in October can genuinely be compared.
This matters most for producers trying to characterise a product line consistently, catch quality drift between batches before customers do, or build a documented flavour profile to support marketing claims about a particular honey's character. None of that is possible from occasional, unstructured tasting, however experienced the taster.
Recruiting and training panelists
Panelists do not need professional sommelier-level palates, but they do need genuine interest, reasonable sensory acuity, and reliable availability, since a panel with inconsistent attendance produces gappy, hard-to-compare data. Basic screening, checking that candidates can identify the five basic tastes at moderate concentrations and can tell which of three honey samples differs from the other two, weeds out anyone with significantly impaired sensory ability before they join a panel that will rely on their judgement.
Training then focuses on building a shared vocabulary and calibrating panelists against each other rather than against some external absolute standard. A handful of practice sessions using reference honeys that clearly demonstrate specific attributes, floral, fruity, woody, herbal, caramel-like, gives panelists a common language and lets an organiser spot outliers whose scores consistently diverge from the group, who may need additional coaching or, occasionally, removal from the panel.
The sensory wheel: a shared vocabulary for flavour
A honey sensory wheel breaks flavour and aroma into broad categories, each with more specific descriptors underneath, giving panelists a structured way to name what they are noticing rather than reaching for vague terms like 'nice' or 'strong'. Typical top-level categories include floral (lavender, orange blossom), fruity (apple, citrus, dried fruit), spicy (cinnamon, clove), woody (resin, cedar), animal (beeswax, propolis), caramelised (toffee, molasses), herbal (mint, thyme), earthy (mushroom, wet soil), nutty (almond, hazelnut), and chemical or off-flavour notes (solvent-like, medicinal), which usually signal a fermentation or storage problem rather than a desirable character.
Printing this wheel and keeping it visible during tasting sessions, at least during training, gives panelists a reference point and noticeably speeds up the process of converging on consistent, comparable descriptions across a group of people who did not grow up using the same flavour vocabulary.
Controlling the environment and preparing samples
Sensory results are only as good as the conditions they were collected in. A neutral room, free of competing food or cleaning odours, held at a stable, comfortable temperature and humidity, with even, neutral lighting for accurate colour assessment, removes a surprising number of confounding variables that would otherwise bias results. Individual tasting booths, where feasible, stop panelists from unconsciously anchoring their scores to what a neighbour just said out loud.
Samples themselves should be brought to a consistent room temperature before tasting, since temperature noticeably affects both viscosity and aroma release, and each sample should be coded with a random number rather than a descriptive label, so panelists cannot unconsciously bias their scoring toward or against a sample based on what they expect it to be. Serving a consistent, small quantity, typically five to ten grams, on identical, odour-free spoons or cups keeps the physical presentation from becoming an extra variable.
Scoring, sequence and turning data into decisions
A structured scoring sheet, weighting appearance, aroma, flavour, texture and an overall hedonic rating, gives panel data enough structure to be averaged and compared meaningfully across sessions. A sensible tasting sequence, visual assessment first, then aroma, then a first taste, followed by a fuller mouth-coating swirl to assess body and texture, and finally an aftertaste check, mirrors how flavour is actually perceived and keeps panelists working through the same steps in the same order.
Once data accumulates across sessions, simple descriptive statistics, means and spread for each attribute, and a basic check of how closely panelists agree with each other, turn a stack of scoring sheets into an actual quality-control tool: a batch scoring noticeably outside the normal range on a specific attribute is a signal worth investigating before it reaches a customer, and a documented flavour profile built up over many sessions becomes genuinely useful supporting material for marketing claims about a honey's character.
Frequently Asked Questions
How many panelists do I need for reliable results?
Somewhere between six and twelve trained panelists is generally considered a workable size for a small-scale panel, enough to average out individual variation without making scheduling unmanageably difficult.
Do panelists need formal training before they can taste usefully?
Some structured practice helps enormously, even if it is informal. A handful of sessions using reference honeys that clearly show specific attributes, alongside a shared vocabulary like a sensory wheel, calibrates panelists against each other far more than raw natural palate ability does on its own.
What descriptors belong on a honey sensory wheel?
Common top-level categories include floral, fruity, spicy, woody, animal (beeswax, propolis), caramelised, herbal, earthy, nutty, and chemical or off-flavour notes. Each category typically has more specific descriptors underneath it that panelists can draw on as they taste.
Why does sample coding matter so much?
Random, blind coding stops panelists from unconsciously scoring a sample based on what they expect it to taste like rather than what they actually perceive, which is essential if the panel's results are meant to be trusted as objective quality data rather than confirmed expectations.