The number itself is worth pausing on: past roughly eighteen distinct flavor options, customers stop being able to name what they just vaped. That figure comes from internal menu audits and retail floor feedback rather than a peer-reviewed trial, so treat it as a working threshold, not a law of nature. But it aligns with a well-documented pattern in choice research, and it raises a question worth taking seriously: if flavor names are how a shop communicates identity, memory, and craft, what happens to that communication when the menu grows past the point where the human mind can hold it?
The Shape of the Problem: Choice Overload in Flavor Menus
The foundational work here is Iyengar and Lepper's 2000 jam study, in which shoppers encountering a display of twenty-four varieties were far less likely to purchase than those shown six — and, notably, less satisfied with what they eventually chose. The mechanism isn't mysterious. Each additional option imposes a comparison cost, and once the comparison set exceeds working memory, the mind shifts strategy. It stops evaluating and starts satisficing: grabbing the first recognizable thing and moving on.
Flavor menus hit this wall faster than most product categories because flavor names carry no inherent meaning. "Blue Raspberry Ice" and "Frosted Blueberry" don't tell a customer anything they can verify before purchase. They're labels pointing at sensory experiences the customer hasn't had yet. That means every name has to do double duty: distinguish itself from neighbors and communicate a taste profile. At six SKUs, that's manageable. At thirty, the names blur into a wall of adjectives.
Why Names Fail Before Products Do
There's a useful distinction between recognition memory and recall memory, and flavor menus stress both. Recognition — "I've seen that name before" — degrades slowly. Recall — "the one I liked was called something like..." — degrades much faster under interference. Interference is the key term. Similar stimuli compete for the same memory trace, and a menu full of fruit-forward names with overlapping adjectives ("Berry," "Blast," "Ice," "Frost") creates near-maximal interference.
A customer who vaped a specific mango blend three weeks ago and wants to reorder it is running a recall task against a menu that has since added four more mango-adjacent SKUs. The 24% drop isn't really about the new flavors. It's about the old ones becoming harder to retrieve.
Variable Reinforcement and the Flavor Hunt
Here's where behavioral psychology gets genuinely interesting, and where shops should pay close attention to their own incentives. Variable-ratio reinforcement — the schedule where a reward arrives after an unpredictable number of responses — produces the most persistent behavior in the operant conditioning literature, going back to Skinner's work in the 1950s. It's the same schedule that makes searching feel compelling. You don't know which option will hit, so you keep sampling.
A sprawling menu inadvertently builds this schedule into the shopping experience. The customer tries something, it's good, they try to find it again, and the search itself becomes mildly rewarding — which sounds positive until you notice the cost. The reward is now attached to hunting rather than to a specific product. That's a fragile relationship for a shop to depend on. Customers who enjoy the hunt have no particular loyalty to any SKU, and customers who don't enjoy it experience the same menu as friction.
Loss Aversion in the Reorder Decision
Kahneman and Tversky's loss aversion work applies directly here. When a customer can't confidently identify their previous favorite, they face a choice between two risks: reorder something close and risk disappointment, or try something new and risk wasting money. Framed that way, the perceived loss of a wrong pick looms larger than the potential gain of a good one. The rational move for a risk-averse customer is often to disengage — to buy less, or to buy from a smaller, more legible menu elsewhere.
This is the quiet cost of menu bloat. It doesn't show up as complaints. It shows up as smaller baskets and shorter visits.
What Eighteen Actually Represents
The 18-SKU threshold shouldn't be treated as magic. It's better understood as the point where a typical customer's working memory, already loaded with price comparisons, nicotine strength options, and device compatibility questions, runs out of room for name-level detail. Working memory capacity estimates cluster around four to seven chunks for novel information, and flavor names are novel by definition — they don't chunk together the way familiar categories do.
In practice, the threshold shifts with menu design. A shop that groups flavors into clearly distinct families — say, four tobaccos, four menthols, five fruits, five desserts — can carry more SKUs while preserving recall, because the customer only has to remember a category and a position within it. A shop that lists thirty flavors alphabetically has created a recall nightmare regardless of total count.
A Concrete Illustration
Consider two shops with identical inventory. Shop A lists every flavor in a single scroll, ordered by when it was added. Shop B organizes the same inventory into five named collections with four to six items each, and repeats the collection name in the product title. A customer who wants "the vanilla custard one" has a much easier retrieval path at Shop B: they remember the collection, scan four names, and find it. At Shop A, they scan thirty. Same products, same prices, different memory load. The recall difference between these two layouts is likely larger than the difference between eighteen and twenty-four SKUs.
Designing for Recall, Not for Display
The forward-looking move for flavor retailers is to treat the menu as a memory architecture rather than a catalog. That means a few concrete shifts.
First, cap the named tier. Not the inventory — the named tier. Rotating or limited flavors can live in a separate, clearly marked section that customers understand is temporary. This preserves the searchable core while still offering novelty.
Second, use collection names as retrieval anchors. If a customer can remember "the Reserve line" or "the Menthol series," they've offloaded most of the recall burden onto a single chunk.
Third, test recall directly. Ask regulars to name their last three purchases without looking. If they can't, the menu is failing at its most basic job, and no amount of visual design will fix a memory problem.
Fourth, watch basket composition, not just basket size. A customer buying one familiar item repeatedly is a different signal than a customer buying three unfamiliar items once. The first is a recall success. The second may be a recall failure dressed up as exploration.
The interesting research question going forward isn't whether eighteen is the right number — it's whether flavor naming conventions themselves can be redesigned to survive larger menus. Descriptive, non-overlapping names that map to distinct sensory profiles may raise the ceiling considerably. "Toasted almond" and "almond cream" interfere with each other. "Toasted almond" and "black licorice" don't. The vocabulary a shop chooses may matter more than how many items it stocks.