Wearable Device Sales Decline: Where Did the Money Go?

Global wrist-worn device shipments dropped 4.3% year-over-year. Is nobody wearing them anymore, or are people no longer paying for specs?

IDC's data speaks for itself: in Q2 2026, global wrist-worn device shipments were 48.01 million units, with smartwatches at 37.08 million units, down 4.2% YoY. I subscribed to RSS feeds for 4 weeks, threw IDC and Omdia summaries into Kimi K3 for categorization, and used a three-week information extraction template to break down categories. I saw money changing pockets.

According to IDC metrics, global wrist-worn device shipments in Q2 2026 were 48.01 million units, down 4.3% YoY.

But Omdia said for the same quarter that global wearable wristbands fell 2% YoY, while smartwatches grew 6%, with high-end sports models performing strongly. Screenless basic bands are rising, while basic bands fell 9%. The numbers differ, but neither is necessarily wrong; the statistical scopes might just be different. Entrepreneurs shouldn't just watch the rise and fall; look at which SKUs can survive.

Metric	Total Market	Smartwatches	Bands	Judgment
IDC Q2	48.01M units, -4.3%	37.08M units, -4.2%	Incomplete	Replacement cycle dragging
Omdia Q2	-2%	+6%	Screenless up, Basic -9%	High-end minimalist surviving
IDC Q1	47.05M units, +2.2%	37.03M units, +4.8%	10.02M units, -6.1%	Mid-to-high end stable

The most glaring issue is the structural change. Low-end bands feel like giveaways, mid-range watches are stuck in the replacement cycle, and high-end sports watches and vertical health devices are starting to sell scenarios.

Huawei, Xiaomi, and Apple hold the top three market shares, which isn't surprising. In IDC's Q1 data, Huawei, Apple, and Xiaomi occupy the head, with CR3 around 54.1% and CR5 around 65.0%. General-purpose wearables are basically a game for giants. New brands should first ask about supply chain, channels, and after-sales networks before talking about ideals.

I've been testing HarmonyOS cockpits these past few days, and I'm just getting started with LiDAR. There's a common flaw in both the automotive and watch industries: when sensors multiply, launch events feel like whitepaper roadshows. Heart rate, blood oxygen, ECG, fall detection—the buzzwords are plentiful, but users only care about accuracy, annoyance, and who to call when it breaks.

Can this direction issue tokens? No. Is Tokenomics reasonable? Also no. Wearable data sounds like personal data assets, but if you actually issue tokens, running can be faked, heart rates can be spoofed, and sleep can be fabricated. Data without real cost constraints will definitely be arbitraged. I've worked on NFT platforms and DeFi protocols and seen beautiful narratives die at "who pays." I previously thought data ownership was the entry point; now my thinking has changed—the entry point is in the billing system.

A few days ago, I wrote an article about "The Cost of Pollination," saying breeding robots shouldn't rush to tell stories but should first check the math. Wearables are the same: don't start with health entry points; first see who pays for false alarms.

The viable direction is a service chain that is regulatable and billable. Post-surgical rehabilitation needs compliance, compensation boundaries, and anomaly alerts; sports teams need load monitoring, injury warnings, and training reports; factory inspections need positioning, fatigue status, and traceable records. Wearables can serve as data collection cards, connecting to services, insurance, healthcare, and enterprise security downstream.

Omdia mentioned that Apple holds a 46% global market share in smartwatches, Garmin had the largest growth among non-Apple manufacturers at 15%, and Samsung also occupies space in high-end sports watches. Professional sports and health monitoring still have money. Garmin survives because it doesn't pretend to be everyone's watch.

For startup teams, don't make the next smartwatch. Make a watch that charges by outcome. Battery life, calibration, OTA, battery replacement, privacy authorization, and anomaly handling—these dirty, tedious jobs are the moat. If you only have product managers and industrial design IDs, without algorithms, clinical validation, and after-sales support, you're basically doomed.

I've used AI for two months and read plenty of smart hardware reviews. Users verbally ask for more features, but their bodies are honest: less charging, fewer interruptions, fewer false alarms. I might not be right, but my testing suggests that hardware startups tend to overestimate flashy tech and underestimate maintenance costs.

The wearable market is declining, but the industry remains. White-label products can't compete anymore, average consumers realize $30-$40 bands are enough, and high-end brands are starting to sell to professional scenarios. Those likely to survive in the future are probably the companies best at connecting false alarms, after-sales, insurance, hospitals, and corporate compliance.

Ask one question first: Who pays monthly for this watch worn on the user's hand, and who is responsible when something goes wrong?