The reorder button on a flavour product page is a small piece of interface with an outsized job: it has to convert a pleasant memory into a repeated purchase before the memory fades. A recent internal test on a liquid flavour shop's reward-claim flow found that adding a two-second haptic pulse at the moment a loyalty reward was confirmed lifted flavour reorders by 11% over the control. That number is interesting less as a conversion metric than as a window into how touch, timing, and uncertainty interact in consumer decisions. What exactly is happening in those two seconds, and why would a vibration change what someone puts in their cart a week later?
The mechanics of a two-second pulse
Haptic feedback is not new, but its placement in a purchase sequence is. Most implementations treat vibration as confirmation—a buzz that says "received." The reward-claim flow is different because it sits at a decision boundary: the customer has just committed points to a flavour, and the system is about to tell them whether the exchange succeeded. That is a moment of mild uncertainty, and uncertainty is where attention concentrates.
The two-second duration matters. Short pulses (under 300 ms) read as notifications and are easy to ignore. Pulses that run longer than about three seconds start to feel like an error state or a lag. The two-second window lands in a range where the user is forced to hold still and wait, which is precisely the kind of brief, low-stakes suspense that makes an outcome feel earned rather than automatic. In the test, the pulse coincided with the reward confirmation rather than preceding it, so the physical sensation and the visual confirmation arrived together. The body registered the result at the same moment the screen did.
Reward prediction error and the flavour reorder
Behavioral psychology has a well-developed account of why this works. Wolfram Schultz's work on dopamine neurons in the 1990s established that the brain's reward signal tracks prediction error—the gap between what was expected and what occurred—rather than the reward itself. A confirmation that arrives exactly as expected produces a small signal. A confirmation that arrives with a slight, pleasant surprise produces a larger one.
The haptic pulse is a small surprise. Customers expect a screen update; they do not expect their phone to respond physically to a flavour reward. That mismatch is the point. It makes the confirmation feel more consequential than a routine UI event, and it attaches a physical memory to a specific flavour choice. When the customer later considers a reorder, the flavour is not just a label in an order history—it is tied to a moment that had texture.
This is also where variable-ratio reinforcement is often invoked, and it is worth being precise about the fit. Variable-ratio schedules—where a reward arrives after an unpredictable number of responses—produce notably persistent behavior, which is why they show up in everything from loyalty punch cards to notification design. The haptic reward claim is not a variable-ratio schedule in the strict sense, because the reward itself is deterministic. But the timing and feel of the confirmation can vary slightly across sessions, and that mild unpredictability in the sensory experience is enough to keep customers attending to the claim moment rather than clicking through it on autopilot.
Loss aversion at the claim screen
Daniel Kahneman and Amos Tversky's framing of loss aversion is the other half of the story. People weigh losses more heavily than equivalent gains, and loyalty points are a textbook case: once a customer has accumulated a balance, spending those points feels like a loss even when the exchange is favorable. The claim screen is therefore a friction point, not a celebration point. Many customers hesitate, abandon the claim, and let points sit idle.
A haptic confirmation does something useful here. It converts an abstract deduction ("your balance is now 340 points") into a concrete, completed event. The pulse signals that the exchange happened—that the loss is settled and the flavour is secured. That closure reduces the lingering sense of having given something up, which is likely part of why reorders rose. The customer who completes the claim without ambivalence is more likely to associate the flavour with a clean, finished transaction rather than a nagging trade-off.
There is a competitive dimension as well. Flavour shops compete on novelty, and novelty-seeking customers tend to churn. A small, well-placed sensory cue at the reward moment gives the shop a way to make a repeat purchase feel like a continuation of a relationship rather than a fresh decision made under uncertainty. The reorder is not just a reorder; it is a confirmation that the earlier choice was right.
What the 11% actually represents
It would be a mistake to read the 11% lift as a pure haptics effect. The test changed the sequencing of the reward claim—the pulse was synchronized with confirmation—and that synchronization is probably doing as much work as the vibration itself. A pulse that fired before the confirmation, or after a delay, would likely produce a different and smaller result. The finding is really about congruence: touch, timing, and outcome arriving as a single event.
The number also has a shelf life. Haptic novelty decays. Customers who experience the pulse on every claim will eventually stop registering it as a surprise, and the prediction-error signal will shrink. The practical implication is that the two-second pulse should be treated as a design element with a maintenance schedule, not a permanent fixture. Rotating the sensory profile—different pulse patterns for different reward tiers, occasional silence to preserve contrast—keeps the confirmation from becoming wallpaper.
Designing the next iteration
The forward-looking question is not whether to keep the haptic claim, but where else in the flavour shop the same logic applies. The mechanism is general: identify moments of mild uncertainty where a customer is committing to a flavour, and make the resolution physically legible. Cart confirmation, subscription renewal, and the first sip of a new flavour in a sampler set are all candidates.
Two cautions are worth carrying forward. First, the effect depends on restraint. A shop that adds haptic feedback to every interaction trains customers to ignore it, and the reward claim loses its distinctiveness. Second, the 11% figure should be replicated with a longer window and a larger sample before it is treated as a stable parameter. Behavioral effects at this scale are frequently real but smaller than the first test suggests, and the decay curve matters more than the peak.
The more durable lesson is about how reward claims are framed. A flavour reorder is a small act of trust—the customer is betting that the memory of the last purchase was accurate. A two-second pulse does not change the flavour, the price, or the points. It changes whether that bet feels settled. Designing for that feeling, carefully and sparingly, is a more interesting problem than optimizing the button.