Flavour swapping in a liquid shop is a coordination problem disguised as a taste problem. The question worth asking is not whether customers can tell two tiers of a flavour apart, but how much resolution a customer needs before swapping stops feeling like a coin flip. If a shop displays five tiers, does the extra grain help anyone decide, or does it simply multiply the comparisons a customer has to make under uncertainty? The emerging answer from decision research and from the way competitive tasting cultures actually behave is that two well-anchored tiers outperform five — not because customers are simple, but because two tiers are the smallest display that still carries a usable signal.
The Case Against Five Tiers: Resolution Beyond Discrimination
Five-point scales feel scientific. They borrow the grammar of survey instruments, where a five-point Likert item is a familiar compromise between granularity and burden. But a rating display is not a survey item; it is a decision aid, and its value depends on whether adjacent points are discriminable — whether a customer can reliably tell tier 3 from tier 4 in a way that changes what they do next.
Most people cannot. Research on just-noticeable differences in taste and smell consistently shows that discrimination thresholds are wider than the fine gradations retail scales imply. A five-tier display asks customers to make four pairwise judgments to locate themselves, and three of those judgments are likely to be noise. The result is a display that looks informative while producing what behavioral scientists call false precision: the customer leaves with a confident position on a scale whose units they cannot actually perceive.
There is a second cost. Every additional tier raises the number of comparisons a customer must hold in working memory. George Miller's classic work on channel capacity put the practical span of absolute judgment at roughly seven categories, but only under ideal, well-separated conditions — and later work by Nelson Cowan and others pushed the realistic figure lower, closer to four. Five tiers sit right at the edge. Two tiers sit comfortably inside it, which means the customer can compare and act without the display quietly taxing the very judgment it was meant to support.
Two Tiers as Anchors, Not Compression
The intuitive objection is that two tiers throw away information. In practice, two tiers function as anchors, and anchoring is what makes a swap decision tractable. Daniel Kahneman and Amos Tversky's work on anchoring and adjustment showed that people evaluate options relative to a reference point rather than in absolute terms, and that a salient anchor shapes the judgment even when it is arbitrary. A two-tier display gives the customer a clean reference: this is the standard expression of the flavour, and this is the intensified or refined expression. The swap decision becomes a single comparison against a reference, not a search across a landscape.
This matters because flavour swapping is a risk decision under uncertainty. The customer cannot fully sample before committing; they are extrapolating from description, reputation, and prior experience. Prospect theory's central finding — that losses loom larger than equivalent gains — predicts that customers will overweight the chance of a bad swap. A five-tier display amplifies that asymmetry, because it invites the customer to imagine four ways of being wrong instead of one. A two-tier display narrows the imagined failure space to a single, legible alternative: "I tried the other tier and preferred this one." That is a loss a customer can actually price.
There is a competitive dimension too. In tasting communities — coffee, whisky, hot sauce, and the concentrate world that overlaps with liquid flavour retail — the most durable rating conventions tend to be coarse and ordinal. They survive because they are reproducible across raters. A five-tier scale fragments a community into people who disagree about the middle; a two-tier scale forces agreement about the poles and lets the middle stay private. Coarse scales are not a concession to laziness. They are a coordination technology.
Reward Loops and the Variable-Ratio Trap
Here the behavioural psychology gets genuinely interesting, and it is worth being precise because this is where flavour retail and reward-loop research overlap most directly.
B. F. Skinner's schedules of reinforcement established that variable-ratio schedules — where a reward arrives after an unpredictable number of responses — produce the highest and most persistent response rates. This is a robust finding, replicated across species and settings. The temptation for a flavour shop is obvious: if swaps are unpredictable in quality, the customer keeps swapping. Five tiers manufacture exactly that unpredictability, because the customer cannot reliably predict which tier will satisfy them. The display becomes an accidental slot machine of the palate.
The problem is that variable-ratio persistence is extinction-resistant in the wrong direction. It produces high response rates and low satisfaction, and it trains customers to treat swapping as a search for a jackpot rather than a refinement of preference. Over time, that erodes the thing a flavour shop actually needs: a customer who knows what they like and trusts the shop's descriptions.
Two tiers short-circuit this. When the display is coarse and honest, the swap outcome is predictable enough to be informative. The customer learns something from each swap — "I am a standard-tier person on this flavour, an intensified-tier person on that one" — and that learning is itself rewarding in the way that competence is rewarding. Research on intrinsic motivation, from Deci and Ryan's self-determination work onward, consistently finds that feelings of competence and autonomy sustain engagement better than unpredictable extrinsic payoff. A two-tier display is a competence-supporting structure. A five-tier display is a lottery with tasting notes.
A Concrete Illustration: The Triangle Test and the Middle Tier
The clearest empirical anchor here comes from sensory science rather than retail. The triangle test is a standard discrimination method: a panelist receives three samples, two identical and one different, and must identify the odd one out. It is used precisely because it isolates discrimination from preference. When panels are asked to distinguish adjacent points on a fine-grained intensity scale, performance frequently collapses toward chance for the middle categories — the differences are real in the chemistry but not in the perception.
The practical implication for a flavour shop is direct. If a trained panel struggles to separate tier 3 from tier 4 under controlled conditions, a customer browsing a display in a shop, under time pressure, with no reference samples, is not going to do better. The middle tiers of a five-tier display are, functionally, decoration. They occupy cognitive space without carrying decision weight — and worse, they invite the customer to believe they have made a fine discrimination when they have made a guess.
Designing the Two-Tier Swap
What does a defensible two-tier system look like in practice? It needs tiers that are perceptually separated (a real, testable difference, not a marketing one), named for what they do rather than for intensity alone (standard versus concentrated, or base versus amplified, so the customer knows what changes), and stable across the catalogue so that "tier two" means the same kind of thing on every flavour. Stability is what turns two points into an anchor. Instability turns two points into two more guesses.
The forward-looking move for shops is to treat the rating display as an instrument to be validated, not a label to be designed. Run your own triangle tests on your own tiers. If your panel cannot separate them, collapse them. Publish the discrimination result alongside the tier, the way tasting competitions publish their judging protocols. The shops that do this will find that two tiers do more work than five ever did — because a coarse scale that customers can actually use beats a fine scale they can only admire.
The next frontier is not more tiers. It is personalised anchors: a customer's own history as the reference point, so that "tier two" means "the tier you chose last time and did not regret." That is a decision aid built on loss aversion and competence rather than on manufactured variability — and it is the kind of display that gets more useful every time a customer swaps.