Do you habituate to a bed shaker?
Short answer: yes, in general — the same way people habituate to a ticking clock, a partner's breathing, or a repeated alarm tone. A stimulus that arrives the same way, at the same time, night after night, is exactly the kind your nervous system learns to file as unimportant. There's no reason to expect vibration to be permanently exempt from that, even though it's a different channel from sound.
What habituation actually is
Habituation is a basic, well-documented property of nervous systems generally: repeated, predictable exposure to a stimulus that carries no consequence tends to produce a smaller and smaller response over time. It's why a ticking clock disappears from your awareness after a few days in a new bedroom, and why a smell you noticed on walking into a room is gone ten minutes later. It isn't a flaw in attention — it's closer to the opposite, a system correctly deciding that a stimulus with no new information doesn't need full attention every time.
Why loud alarms don't work for heavy sleepers covers this for sound specifically: a new alarm tone often works well for a week or two and then quietly stops, not because anything about the volume changed, but because the brain has learned the sound predicts nothing worth waking for. There's no obvious reason the same learning process would spare a vibration pattern that arrives with equal regularity every morning.
Why vibration doesn't get a permanent pass
Vibration has a genuine, real advantage the first time you try it: it's novel. A brain that has spent years learning to filter one particular alarm tone hasn't learned anything at all about a buzz against the mattress — it's an unfiled stimulus, which is why switching from sound to touch so often works immediately where turning the volume up further does not. Do vibrating alarms actually work? sets out that logic for the vibration category as a whole.
But novelty is not the same as permanent immunity. Once vibration itself becomes the predictable, repeated, every-morning signal, it's reasonable to expect the same general habituation process to start working on it too, just from a later starting point than a years-old alarm tone. The honest expectation isn't "vibration never fades," it's "vibration starts from a cleaner slate and may take longer to fade" — which matches what a fair number of long-term users report anecdotally, even without a dedicated study to point to.
What actually resets it
If habituation is about predictability, the fixes point at making the stimulus less predictable again, rather than simply making it stronger.
Change the intensity. A device with adjustable levels lets you start at the lowest setting that reliably wakes you and hold higher settings in reserve. If the low setting starts to fade, moving up isn't just "more of the same" — it's a different intensity your brain hasn't filed yet. A fixed, single-intensity shaker has nowhere to go once this happens; you're stuck buying a new device to get the same effect a built-in escalation would have given you.
Change the placement. Moving the puck from under the mattress to under the pillow, or shifting it to a different spot, changes exactly where and how the vibration is felt on your body, which functions as a fresh stimulus even though the underlying device hasn't changed. It's a low-effort thing to try before assuming the device itself has failed.
Alternate mechanisms. Switching between a bed shaker and a wrist-worn band, or occasionally reintroducing a sound alarm as a backup, keeps any single channel from becoming the years-long habituated one. This is really the same principle that makes switching from sound to vibration work in the first place, just applied a level further: don't let any one stimulus become the permanent, ignorable background.
Fix the schedule underneath it. A consistent wake time means your body starts preparing to wake before the alarm fires at all, so the alarm is doing less of the actual work and habituation to it matters less either way. This doesn't reverse habituation directly, but it lowers how much any single alarm needs to accomplish.