Source: Unsplash
Apple sold roughly 600,000 Vision Pros across two years. Meta sold over 7 million Ray-Ban smart glasses in 2025 alone. The hardware difference is real, but the deeper story is a UX story, and I think most of the industry is reading it wrong.
Apple Vision Pro did not fail because of weight, price, or chip generations. It failed because Apple tried to recreate the Mac inside your living room, and people did not want a Mac in their living room. Meta sold seven times more smart glasses than Apple sold Vision Pros in 2025, not because the Ray-Ban Meta is more powerful, but because it solved a real, narrow, glanceable problem without asking users to enter a new operating system. This piece breaks down what the divergence actually means for product teams building the next layer of spatial UX in 2026, and why most of us are about to make the same mistake again with AI glasses.
"He was similarly wary of the mixed-reality headset that became the Vision Pro. In both of those cases his skepticism was prescient. Apple eventually killed the car, and the Vision Pro has been a bust."
Mark Gurman, reporting on Apple CEO John Ternus, Bloomberg, April 29, 2026
The Numbers That Tell The Real Story
Let us start with the receipts. According to IDC and several analyst estimates, Apple shipped roughly 370,000 Vision Pro units in the first three quarters of 2024, and total shipments across 2024 and 2025 combined sit at around 600,000 units. Apple has never officially disclosed sales figures, but the production cuts at Luxshare in early 2025 told the story without a press release.
Now compare that with Meta. EssilorLuxottica, the parent company of Ray-Ban, reported selling over 7 million Ray-Ban Meta smart glasses in 2025 alone, with hardware revenue estimated at $2.15 billion. That is more revenue than Quest generated for Meta for the first time ever. Approximately 17 percent of U.S. online adults used smart glasses in 2025, up from 4 percent the year before, according to industry surveys. The global smart glasses market grew 110 percent year over year in the first half of 2025.
So when people say spatial computing is dead, they are reading the wrong scoreboard. Spatial computing is exploding. Spatial computing as Apple imagined it is what is dying.
The Real Reason Vision Pro Failed
I have been designing enterprise products and AI native interfaces for the past eight years, and one of the lessons that keeps repeating itself is this: users do not adopt new operating systems, they adopt new behaviors. The Vision Pro asked users to adopt both at the same time, and the math never worked.
Think about what the Vision Pro actually offered on day one. It offered the same apps you already have, floating in 3D space. Slack, Safari, Notes, FaceTime. Apple pitched it as spatial computing, but in practice it was a high resolution monitor strapped to your face. The interaction model was eye tracking plus a pinch gesture, which is impressive technology, but the question I kept asking my team was: what new behavior does this unlock that I could not already do on my MacBook?
The honest answer for most knowledge workers was almost nothing. The Vision Pro was a better movie theater and a slightly cooler way to look at a giant browser window, and that does not justify $3,499 and a sore neck after 45 minutes.
I wrote about this exact pattern in a Medium piece last month on AI native experiences. The same mindset shift applies here. Spatial computing is not desktop computing with depth added. It needs its own grammar.
What Ray-Ban Meta Got Right That Apple Missed
The Ray-Ban Meta is not a spatial computer. It is a pair of sunglasses with a camera, microphones, and a Meta AI voice assistant. That is the entire feature set, and that is exactly why it works.
- Glanceable, not immersive: Ray-Ban Meta does not pull you into a new world. It augments the world you are already in. You wear them while walking, biking, or cooking, and you ask Meta AI a question or capture a photo. The interaction lasts seconds, not sessions.
- Zero learning curve: If you have ever used Alexa or Siri, you already know how to use Ray-Ban Meta. There is no new gesture vocabulary, no eye tracking calibration, no app store to figure out. The product is so simple it is almost boring, and that is the point.
- Solves real moments, not theoretical ones: Hands free photos when you have a baby in your arms. Quick translation when you are traveling. Voice queries when your phone is buried in a bag. These are tiny, specific moments where reaching for a phone feels worse than asking your glasses.
- Social acceptability: You can wear Ray-Ban Meta to dinner. You cannot wear Vision Pro to dinner. This sounds shallow but it determines whether a wearable becomes part of someone's life.
The Vision Pro created what one Apple analyst called "a barrier between the wearer and everyone else in the room", and that barrier is fatal for a consumer product. Humans are social. We make eye contact. We read each other's faces. Any wearable that breaks that contract has to deliver enormous value to overcome the social tax, and floating Slack windows do not clear that bar.
What April 2026 Actually Revealed About Apple
On April 29, 2026, reports surfaced that Apple had dissolved the Vision Products Group, reassigned the software team to the Siri overhaul, and moved the hardware team toward smart glasses. Mark Gurman at Bloomberg confirmed parts of this with more nuance, while AppleInsider pushed back on the framing. Within two weeks, the story softened into something closer to a strategic restructuring than a full kill.
But the part that did not soften is this: Apple's new CEO John Ternus is reportedly bearish on the Vision Pro as a product category. Apple is hiring into the team, yes, but the people they are hiring are working on the next form factor, which is glasses, not a headset successor. That is a quiet admission that the original bet was wrong.
I find this honest and refreshing, even though Apple will never frame it that way publicly. The Vision Pro was a flagship of what Apple believed spatial computing would become, and the company is now reading the same market data the rest of us are reading. Ray-Ban Meta is what users actually want. Glanceable, glasses shaped, AI first.
The Design Lesson That Should Outlive Vision Pro
I have been doing product design at scale for eight years across fintech, healthcare, EdTech, and enterprise SaaS, and the lesson the Vision Pro teaches is the same one the Apple Watch taught us a decade ago. New form factors require new interaction grammars, not ported old ones.
When the Apple Watch shipped, it tried to be a tiny iPhone on your wrist. App grid, swipe gestures, tap targets. It almost died until the team figured out the watch was actually a notification triage device and a fitness tracker. The current Watch is a completely different product than the 2015 Watch, and the difference is that the team finally let the form factor define the experience instead of forcing the experience to define the form factor.
Vision Pro is at the same crossroads. The hardware can survive if the team accepts that floating macOS windows is not the product. The product is shared spatial experiences, location based AI, real world annotation, ambient awareness. None of that exists yet at a level that justifies the price tag, and that is the work that has to happen before the next headset ships.
What This Means For Product Teams Right Now
Google I/O 2026 opens on May 19, just two days from today as I write this. Android XR smart glasses will be previewed with hardware partners including Samsung, XREAL, Warby Parker, and Gentle Monster. Project Aura is expected to make its public debut with developer kits later this year. Meta is doubling down on AI glasses after cutting roughly 1,000 Reality Labs employees in January. The next 18 months will define the category for the rest of the decade.
If you are building anything for this category, the practitioner advice from my side is this. Stop designing for the headset and start designing for the moment. What does the user need in five seconds, hands busy, no time to read? That is the spatial UX brief. If your product needs the user to enter a special mode to use it, you have already lost.
This connects directly to what I wrote about at reloadux on AI readiness for product teams. The same framework applies. Pick the smallest unit of value, deliver it in the lowest friction way, then expand. The Vision Pro tried to launch with a full operating system. Ray-Ban Meta launched with three features. Guess which one the market rewarded.
The Counterpoint Worth Acknowledging
To be fair to Apple, there is a real argument that Vision Pro was always meant as a developer platform and a long arc bet. Daring Fireball made this case in April. Apple has done this before with the original iPhone, which was also expensive and limited at launch. The Watch was also slow to find product market fit. So writing off Vision Pro as a failure after 24 months might be premature, especially since Apple is clearly still investing in the underlying platform stack.
I think that argument has some weight, but it ignores the timing. Apple is no longer the company that can afford a five year arc on a flagship product without consumer traction. The market is moving faster, Meta is moving faster, Chinese OEMs are shipping AI glasses at sub $300 price points, and Apple's traditional moat of polished hardware does not save you when the form factor itself is wrong.
Where Spatial UX Actually Goes Next
The interesting work in spatial computing right now is not happening in headsets. It is happening at the intersection of AI, voice, and lightweight wearables. The Ray-Ban Meta is the rough draft. The next versions will add a small display, gesture input via a wristband, and context aware AI that knows what you are looking at without you asking.
For designers, the implications are concrete. You need to start thinking in terms of multimodal input pipelines rather than single screen flows. Gaze, voice, gesture, and ambient context all feeding the same interaction. You need to design for partial attention, because users will not be staring at your interface, they will be glancing at it while doing something else. And you need to design for social context, because a wearable that embarrasses the user is a wearable that gets left at home.
None of this is in your Figma plugins yet. Most design systems are still organized around screens. The teams that figure out spatial UX patterns in the next 18 months will be writing the playbook everyone else copies for the next decade.
What do you think? Is Vision Pro actually finished, or is Apple about to surprise us again? Drop your thoughts in the comments below. I would love to hear how you are seeing this play out in your own work.
Sources:
1. AppleInsider, "Potential Apple Vision Pro team dissolution isn't a death knell for the product": https://appleinsider.com/articles/26/04/29/rumored-apple-vision-pro-team-break-up-isnt-a-death-knell-for-the-product
2. Road to VR, "Meta Sold Over 7 Million Smart Glasses Last Year": https://www.roadtovr.com/meta-ray-ban-smart-glasses-sales-tripled-2025/
3. Tom's Hardware, "Apple has given up on Vision Pro, report claims": https://www.tomshardware.com/virtual-reality/apple-has-given-up-on-vision-pro-report-claims
4. CNBC, "Ray-Ban maker EssilorLuxottica says it more than tripled Meta AI glasses sales in 2025": https://www.cnbc.com/2026/02/11/ray-ban-maker-essilorluxottica-triples-sales-of-meta-ai-glasses.html
5. AppleInsider, "Analysts need Apple Vision Pro to be a flop": https://appleinsider.com/articles/26/01/01/analysts-need-apple-vision-pro-to-be-a-flop-whether-apple-considers-it-one-or-not
6. TechTimes, "Meta Cuts VR Staff to Double Down on Wearables": https://www.techtimes.com/articles/316639/20260514/meta-cuts-vr-staff-double-down-wearables-snaps-collapsed-perplexity-deal-reshapes-ar-industry.htm