Your App Buzzes on Every Tap. That's Not Haptic Design — That's Noise You Can Feel.
Pull to refresh. Buzz. Toggle a setting. Buzz. Delete a message. Buzz. Same buzz, every time, regardless of whether you just confirmed something, broke something, or merely scrolled past it. Your phone has a dedicated linear actuator engineered to communicate through touch, and most apps use it the way a smoke alarm uses sound: one signal, no distinctions, fired at everything.
That's the part of haptic design nobody talks about with the seriousness it gets applied to visual design. Nobody would accept a product where every button, error, and success state used the identical shade of gray. Yet that's exactly the standard most teams apply to touch — one generic vibration standing in for confirmation, warning, selection, and impact, because vibration is the one channel product teams still treat as a boolean instead of a language.
There's Already a Grammar. Almost Nobody Is Using It.
Apple's Human Interface Guidelines don't leave this to intuition — they specify three distinct generators, each mapped to a different category of meaning. UIImpactFeedbackGenerator is for the sensation of one object hitting another — a dropped card snapping into place, a slider hitting its limit. UINotificationFeedbackGenerator is reserved for outcomes — success, warning, and error each get a structurally different pulse pattern, not just a different sound layered on top of the same buzz. UISelectionFeedbackGenerator is for the lighter, more frequent tick of moving through a set of options — a picker wheel, a segmented control. Three purposes, three distinct physical vocabularies, built into the platform a decade ago.
Almost no third-party app uses more than one of them. The overwhelming majority default to a single impact-style buzz wired to nearly every interactive event, which means the platform shipped a language and the industry collectively decided to speak in grunts. Dave Birnbaum, a design strategist who has spent years specifically on haptic craft, made the comparison that should reframe how product teams think about this: "Good haptic design is like typography — it contributes significantly to perceived quality, often unnoticed until poorly executed. It's a hallmark of craftsmanship." Nobody notices good typography either, until you put it next to bad typography. The same is true here, except almost nobody has seen the good version to notice the gap.
The Window Where a Buzz Still Feels Like It Belongs to You
Here's the part that turns this from an aesthetic complaint into an engineering constraint: haptic feedback isn't just about which pattern you fire. It's about when. Human factors research places the threshold for perceived realism at under 10 milliseconds of latency between a touch event and its haptic response — below that line, the vibration reads as part of the same physical event as the tap. Cross a slightly wider window, and the perception starts to change: patent research on touchscreen-haptic coupling has found that once feedback misses roughly a 30-millisecond mark, users will physically re-attempt the tap, as if the first one hadn't registered — a subtle sign that the connection between action and response has already broken, even though the interface eventually did the right thing.
Push further out and the failure gets more forgiving but doesn't disappear: CHI research on tactile latency puts the boundary for feedback still reading as roughly instantaneous around 150 milliseconds, with forearm-level stimulation tolerating up to 500 milliseconds before users register a lag. What all three numbers describe is a shrinking window in which a haptic response still feels causally connected to the thing that triggered it. Miss the window, and you haven't just delayed feedback — you've disconnected it. The buzz still arrives. It just arrives as a separate event instead of as the confirmation of the one that came before it, and a user's hands know the difference even when their conscious attention doesn't.
This is where haptic design quietly becomes an engineering problem wearing a design costume. The same discipline that governs whether an animation explains state or just decorates it applies here with a stricter clock — a visual micro-interaction that's 200 milliseconds late still reads as slow. A haptic response that's 200 milliseconds late reads as broken, disconnected, or arbitrary, because touch has no visual continuity to paper over the gap. There's nothing on screen bridging the two moments. There's just the hand, and a pulse that no longer feels like it belongs to the tap that caused it.
The Products That Got This Right Shipped a Decade Ago and Nobody Copied the Homework
Apple didn't introduce this as a UI flourish. The Taptic Engine launched hardware-first — in the Apple Watch in 2014, then in the iPhone 6S and MacBook trackpad in 2015 — specifically because a trackpad that doesn't physically move needed to simulate a click precisely enough that engineers could stop pretending it wasn't real. That's the origin story worth remembering: haptics on these devices exists to solve a specific deception — a surface with no moving parts had to convince a hand that something had moved — and the entire discipline was built around getting that illusion right at a latency users' nervous systems couldn't outpace.
Ten years on, most software teams inherited the hardware and skipped the discipline. They got a Taptic Engine capable of distinct, purposeful, sub-10-millisecond responses, and used it to replicate a single doorbell buzz across every interaction in the app. The capability scaled to every device. The design language that was supposed to ride along with it mostly didn't.
So Actually — the Fix Isn't More Haptics. It's Fewer, Used on Purpose.
The instinct, once you see the gap, is to reach for more — richer patterns, more distinct buzzes, haptic feedback on every interactive element to finally use the full vocabulary Apple built. That's the wrong lesson. A grammar used indiscriminately is still noise; it's just more elaborate noise. The actual craft isn't deploying three generators instead of one everywhere. It's deciding, deliberately, which interactions deserve a physical response at all — reserving impact feedback for genuine physical metaphors, notification feedback for outcomes that matter enough to interrupt, and silence for everything else, the same way a good typographic system reserves bold for genuine emphasis instead of using it on every line.
Most users can't articulate why one app feels considered and another feels cheap under their thumb. They can feel it instantly, every time, without knowing why — which is exactly what a design language is supposed to do when it's working: operate below the threshold of conscious notice and above the threshold of felt experience. The question worth asking before your next release isn't whether your interactions have haptic feedback. It's whether you could describe, out loud, why each one uses the pattern it uses — and if the honest answer is "we just added a buzz," the fix isn't a bigger vibration library. It's the discipline to leave most of it off.