Flavour retail has quietly become a study in timing. Anyone who has watched a customer reorder the same blue-raspberry concentrate three weeks running knows the decision isn't purely about taste—it's about when the nudge arrives and how much slack the buyer has left in their supply. The specific question worth pressing on: why does a restock alert fired roughly two hours before a customer's typical reorder window convert into repeat flavour orders at a higher rate than the same alert sent twelve hours out? The answer sits at the intersection of behavioural psychology, decision-making under uncertainty, and the reward structures that keep small-batch buyers coming back.
The Two-Hour Window and the Architecture of Attention
The two-hour alert works because it lands inside what behavioural economists call the "hot" phase of a decision—the period when the relevant need is salient and the cost of acting is low. Twelve hours out, the same message arrives during what is effectively a cold phase: the customer is at work, driving, or asleep, and the alert becomes a piece of ambient noise that gets swiped away. By the time the need actually surfaces, the message has already been forgotten or mentally filed as "later."
This is not a new observation in consumer research, but it has been under-applied to small-batch flavour retail. The classic illustration is the 2006 study by behavioral economists at a large U.S. university who tracked deadline-driven purchasing among student savers. Participants who received reminders close to a self-imposed deadline saved at significantly higher rates than those reminded days in advance, even when the total number of reminders was identical. The variable wasn't frequency—it was proximity. The closer the prompt sat to the moment of action, the more it functioned as a trigger rather than an interruption.
For flavour shops, the implication is concrete: a customer who orders a 30ml concentrate every 18 days does not need a reminder on day 17 at 8 p.m. They need it on day 18 at the hour they typically sit down to mix. The two-hour lead captures that hour. The twelve-hour lead misses it.
Variable-Ratio Reinforcement and the Small-Batch Restock
There is a second mechanism at work, and it is where the flavour niche gets interesting. Small-batch restocks are inherently variable. A popular custard or tobacco blend may sell out in a day; a niche floral may sit for a month. Customers who order repeatedly are, whether they articulate it or not, participating in a schedule of intermittent availability. Behavioral psychology has a well-worn term for this: variable-ratio reinforcement, the schedule in which a reward arrives after an unpredictable number of responses. It is the same schedule that makes scratch cards and social-media feeds sticky, and it produces notably persistent behaviour.
The ethical distinction matters here. A flavour shop is not engineering a compulsion; it is managing genuine scarcity that comes from small production runs. But the psychological texture is similar enough that the alert design should be handled with care. A two-hour restock alert on a limited batch tells the customer something true and useful: this is available now, and it may not be in three hours. That is information, not manipulation. A twelve-hour alert on the same batch, by contrast, often arrives after the batch has already sold through—which trains the customer to distrust the alert and, over time, to discount the shop's messaging entirely.
The practical upshot is that alert timing and alert honesty are the same problem. If the two-hour window is used to signal real availability, it builds a reliable cue. If it is used to manufacture false urgency on stock that will clearly be there tomorrow, customers learn the cue is noise. The reinforcement schedule only works as long as the reward is real.
Loss Aversion in the Reorder Decision
Kahneman and Tversky's work on loss aversion—the finding that losses loom roughly twice as large as equivalent gains—has an obvious application to restock alerts that is easy to overlook. A customer deciding whether to reorder is not weighing "gain a bottle of concentrate" against "save the money." They are weighing "keep my mixing routine intact" against "run out mid-week and have to substitute a flavour I don't like as much." The second frame is a loss, and losses are stickier.
A twelve-hour alert gives the customer time to rationalize: I'll order tomorrow, I still have some left. The loss feels distant and therefore small. A two-hour alert compresses the decision. The customer is closer to the empty bottle, closer to the moment of substitution, and the loss is concrete rather than hypothetical. The alert doesn't create the loss aversion—it just stops the customer from talking themselves out of acting on it.
There is a competitive-play dimension here worth naming. Flavour buyers, especially in enthusiast communities, compare notes. They know which shops restock reliably, which ship fast, and which run out of the good stuff. A shop that consistently fires a two-hour alert on a restock that genuinely sells out is building a reputation for being on top of it—a form of competitive credibility that a twelve-hour alert, arriving after the fact, cannot match. The alert becomes part of the shop's identity, not just a transactional ping.
Decision-Making Under Uncertainty and the Cost of the Wrong Alert
The twelve-hour alert is not merely less effective; it is actively costly in a way that is easy to miss. Every alert a customer receives and does not act on is a small negative signal. Behavioral research on notification fatigue—most visibly the work on "alert fatigue" in clinical settings, where clinicians begin ignoring warnings that fire too often or too early—suggests that the cost is cumulative. A customer who receives twelve-hour alerts that never land at the right moment learns to ignore the channel. By the time a genuinely urgent two-hour alert arrives, it is filtered out.
This is a decision-making-under-uncertainty problem for the shop as much as the customer. The shop cannot know with certainty when any given customer will next need to reorder. But it can estimate the distribution: most customers reorder within a predictable band, and the modal reorder hour is learnable from order history. The two-hour alert is a bet that the modal hour is the right target. The twelve-hour alert is a hedge that spreads the message across a wider window—and, in doing so, dilutes it.
The evidence from small-batch retail operations that have tested both windows is consistent: the tighter window produces higher open rates, higher click-through, and, most importantly, higher repeat-order rates on the same flavour. The mechanism is not mysterious. It is proximity plus salience plus a real, time-bounded reward.
What to Build Next
The forward-looking move for flavour shops is to treat the two-hour alert not as a marketing tactic but as a scheduling problem with a behavioural tail. That means building the order-history model that identifies each customer's modal reorder hour, then firing the restock alert inside that window rather than at a fixed daily send time. It means tying alerts to real inventory states—batch confirmed, batch packed, batch live—so the two-hour window always carries true information. And it means resisting the temptation to stretch the window when a batch is slow to move; a two-hour alert on stock that will clearly last the week is a false cue, and false cues erode the reinforcement schedule that makes the whole system work.
The shops that get this right will not be the ones with the loudest alerts. They will be the ones whose alerts arrive at the moment the customer is already half-deciding—bottle nearly empty, routine about to break, substitution looming—and who have earned enough trust that the message is read as help rather than noise. The two-hour window is not a trick. It is a recognition that flavour buying, like most repeat purchasing, is a timing problem dressed up as a taste problem.