Hello Sense Sleep Tracker: A Retrospective Review
Hello Sense was a sleep tracking system that did several things differently from competing products. The main unit — the Sense — was a translucent sphere that sat on the bedside table monitoring room conditions rather than the sleeper's body. The Sleep Pill was a small clip attached to the pillow that tracked movement and used that data in combination with the room sensor data to generate sleep quality scores. The company, Hello Inc., shut down in June 2017. The app went offline. Users lost access to their historical sleep data. This retrospective covers what Sense actually measured and how well it worked, the hardware quality, the company's arc from Kickstarter to shutdown, and what the story illustrates about the risks of connected devices where the value is tied to cloud services. This is part of the reviews section.
What the hardware did
The Sense bedside sphere contained sensors for ambient light level, noise level, temperature, humidity, and air particulate matter. It tracked room conditions throughout the night and correlated them with movement data from the Sleep Pill. The system used this to generate a nightly sleep score and a breakdown of sleep stages (light, deep, REM) that was, by the company's own acknowledgement, an estimate rather than a clinical measurement. The Sleep Pill used accelerometer data to detect movement — a proxy for sleep stage, not a direct measurement of sleep physiology.
The environmental monitoring was the Sense's genuinely differentiated feature. Most sleep trackers of the era monitored only movement (via wristband). Sense added room condition data — whether the room was too bright, too warm, or unusually noisy at the times when sleep quality scores dropped — which created more actionable data. "Your sleep was disrupted around 3 AM" is less useful than "your sleep was disrupted around 3 AM when the noise level spiked to 72dB." The correlation between room conditions and sleep stages was presented clearly in the app.
The hardware quality
The Sense hardware was well-made by Kickstarter standards of 2015 — far better than the typical crowd-funded hardware that ships late and fails early. The sphere was solid, the material quality was good, and the ambient light sensor functioned accurately. The Sleep Pill's pillow clip was the weakest point: the clip mechanism loosened over months of use, and some users reported the pill detaching from the pillow cover during the night, causing gaps in tracking data. This was a known issue that the company acknowledged but didn't resolve before shutdown.
The LED indicators on the Sense provided a useful passive ambient display — the orb glowed different colours to indicate sleep environment quality and could be set to function as an alarm with gradually brightening light. The alarm feature worked through the app and required an active internet connection to the Hello servers, a dependency that became relevant when the servers went offline.
The app and sleep data quality
The companion app was clean and well-designed by 2015–2016 standards. Sleep scores were presented with breakdown into stages and room condition annotations. The app had a social layer (comparing sleep scores with friends) that few users actually used.
The sleep stage classification accuracy was modest. Accelerometer-based sleep staging — inferring sleep stage from body movement — is known to overestimate REM sleep and struggle to differentiate light from deep NREM sleep, particularly for side sleepers who move less. Sense's estimates correlated roughly with user-perceived sleep quality but were not reliable enough for clinical use or detailed self-experimentation. The environmental data was more reliable than the sleep staging.
Sense in operation (2015–2017): The app functioned. Sleep data accumulated. Historical charts showed trends over weeks and months. The room condition annotations were genuinely useful for identifying environmental factors affecting sleep. The system worked as described, within the limitations of the underlying measurement approach.
After the shutdown (June 2017 onwards): The Hello app went offline. Historical sleep data became inaccessible. The Sense hardware became a non-functional sphere — the local functionality (alarm, ambient display) required the app, which required the servers. Hardware that cost backers $129+ became decorative within weeks of the company's announcement. Some users exported data before shutdown; most did not, unaware of the closure timeline.
The company's closure
Hello Inc. raised $40 million in venture capital after a successful Kickstarter campaign. The Sense launched in late 2015 and received broadly positive reviews. A second generation product (Sense With Voice) was announced. The company then went quiet. In June 2017, with no public deterioration narrative, Hello announced it was shutting down operations and that the servers would go offline.
The closure announcement gave users approximately one month's notice. There was no mechanism to export all historical data — the app had limited data export functionality, and what existed required users to know it was available and take action before the deadline. Many users had months or years of sleep data that became permanently inaccessible when the servers went down.
The data loss outcome was not unique to Hello — it has since occurred with numerous other consumer IoT products whose cloud services were discontinued on or shortly after company closure. The pattern is now well-established: hardware value is contingent on server availability, server availability is contingent on company health, and company health in consumer IoT is unpredictable. The response from the hardware industry has been incomplete — some products have added local data export functionality, but many remain fully cloud-dependent.
Lessons and what they mean for connected devices
The principal mistake users make with cloud-dependent devices is treating the hardware purchase as the primary transaction. The hardware is functional because of the service layer, and the service layer is a separate ongoing commitment from the company. Evaluating a connected device's value should always include the question: what does this do if the servers go offline tomorrow? For Sense: nothing. The alarm didn't work. The ambient display didn't work. The historical data was gone.
Sleep tracking as a category has matured since Sense. The viable approaches are now: tracking via a wearable that stores data locally and syncs to cloud secondarily, or dedicated sleep tracking hardware with local storage and open data export. Pure cloud-dependency for a device that collects daily personal data about sleep and home environment is a risk profile that the Hello story made concrete.
The Sense hardware design was genuinely good. The environmental sensor combination was ahead of most competitors. The company failed, not the concept. The lesson is about infrastructure architecture, not about whether environmental sleep monitoring is valuable — it is, when the data is yours to keep.