Augmented Reality has seen significant popularity since it became easy to build on mobile with Apple and Android SDKs, in 2017. It is now a common technology that is well known in the public, but years are passing, and it is taking longer than expected to replace all of our electronics and be in a wearable glasses format that we can wear all day, every day. Frustration grows in the developer community, tech industry but also in the VCs who are reducing investments to prioritize AI efforts as they are struggling to see short term revenue from it.
What happened since 2017? How are companies changing their strategy? What should we focus on now? I will try to provide an history and insights on what matters today.
For context, I used to work in AR Developer Relations at Apple and Snap, both building mobile and wearable AR technology, while also working closely with companies and individuals building their businesses on these platforms, but I will only leverage publicly available information. In the following paragraphs, I will try to summarize what have been the main stages of this development and I will finish by what is the current state of the field, what are the opportunities.
Note: I will not focus on AR headsets such as Magic Leap, Hololens which are amazing devices but have not been as widespread as mobile based AR or the most recent MR headsets, because of their costs, size and weight.
1. History of consumer AR since 2017
a. Golden age of Mobile AR: 2017-2020
Apple released their AR SDK ARKit in 2017, bringing large enthusiasm from Apple platform developers, who saw the potential of future AR glasses and an opportunity to shape the future of a product with the potential to change people’s lives in a similar way to the iPhone. AR apps and concepts were viral on a weekly basis for a while. Google then released a similar SDK called ARCore in 2018.
Over the next couple years, thousands of apps were released or updated with AR. Considering that looking around arm up with your device only had retention of 60 seconds or so, it quickly became clear that users were not ready to pay for this type of experience, but were happy to use it. It also became apparent that AR is an incredible addition for contextual and location based experiences, but that it would not generate high retention or high revenue without a headset.
After 3-4 years of large AR updates, Apple started to be more quiet about them, preparing in silence for its upcoming headset. A sign of this has been the AR page of the Apple AppStore featuring pretty much the same apps for years, even when they were almost deprecated. I believe that since covid, we have seen significantly less innovation on mobile AR, and AR became a regular add-on but not a reason to download an app, which is where social AR took an interesting position…
b. Golden age of Social AR: 2020-2023
Seeing the potential to drive social posts, social media companies such as Snapchat, Instagram and TikTok released AR tools to build face filter experiences that could reach hundreds of millions of impressions on iOS and Android. Not having to download apps and just having to download 4 to 8 Mb experiences that a user can try instantly has been a major advantage compared to regular native mobile AR. Apple seems to have tried to reproduce that as well with AR AppClips, small apps that could be loaded on the go, but it never really took off as it is for Apple devices only.
Snapchat launched Lens Studio in 2017, Instagram followed in 2018 with Meta Spark Studio, and TikTok introduced Effect House in 2022. After initially focusing on front-camera AR, they shifted towards world AR with features similar to native mobile AR. By 2022, they became just as advanced, if not more, offering 3D hand tracking and real-time ray tracing. They also began offering mobile SDKs for developers to integrate these features directly into their apps, like Snap’s CameraKit SDK and TikTok’s BytePlus SDK. Additionally, they expanded support for AR headsets and glasses, including Snap Spectacles and Meta Quest.
Used for mobile advertising on these social media apps, the platforms’ sales have leveraged AR efficiently and grown their revenue from it. Over the years, it became clear to advertisers that high end AR experiences were expensive and taking a while to produce compared to more regular 2D photo or video advertisement. CGI AR-like videos, commonly called FOOH (Fake Out Of Home) became for example cheaper ways of having impressive immersive experiences while creating a similar lasting memory to end users.
While Generative AI came to the main stage in 2023, these companies kept their interest for AR but reprioritised a lot of resources for AI, to focus mainly on engaging AR, bringing revenue rather than advanced experiences with poor retention on social platforms. As I write this article, Meta even just announced shutting off completely their mobile AR tool Spark AR in August 2024, previously allowing to build AR for instagram. In the meantime, new hardware is coming out and many AR related employees of these companies have been laid off, switched teams or left for other opportunities.
c. Golden age of Mixed Reality & Smart AI Glasses 2023-2026(?)
With Apple showcasing their Mixed Reality headset in 2023 and releasing it in 2024, Mixed Reality has become the new innovative platform for AR. For context, the industry-famous AR glasses hardware limitations with heat, battery life and a large field of view, have moved the current interest from optical see through AR glasses like Magic Leap One and Two, HoloLens 1 & 2 and Snap Spectacles to passthrough headsets (with cameras and great screens in a headset) like the Meta Quest 3, Apple Vision Pro or Pico 4 Ultra.
By making impressive real-time AR accessible on $500 headsets instead of the $3000-$5000 ones, these devices have become the new leading way to experience AR and are currently where we’re seeing the most momentum.
In the meantime, AI smart glasses like the Meta Ray-Bans are seeing great interest and are selling like hot cakes (1M+ forecasted in 2024), but do not yet have screens allowing to experience real time AR, and no SDK for third party companies or individuals to build products on them.
I would guess that consumer-friendly AR+AI glasses that we can wear all day, every day for an affordable price ($500-1000) would happen starting 2026 at the earliest with the current stage of the tech, and the time it takes to develop hardware products.
2. What can we do now in 2024?
a. Devices sales and use cases that actually work for AR
Industry rumors are saying Apple Vision Pro has sold globally 200-300K units, Meta has announced having sold 30M+ Quest headsets, Xreal Glasses have sold 300K+ Units, the market is growing quickly but not as fast as we all would like. If you ask someone on the street, the odds of them having a headset are still low.
Interestingly aside from gaming being huge on the Meta Quest, the main public use cases for consumers of these headsets or glasses are mainly viewing giant displays comfortably such as for viewing 2D or 3D content (Vision Pro, Xreal) or working on a large display (Vision Pro, Xreal, Quest). People are not downloading as many apps as we would have thought, the killer use case is the comfort of huge displays.
All the Vision Pro owners I know are mainly using it for work, to view their screen or sometimes watch content. The issue is just the weight of the device on the neck, which makes it hard to use for more than 2 hours at once.
b. What are companies doing about it?
With the Quest 3, Meta has focused on gaming in VR for as a main selling point, but for businesses with B2B use cases like trainings, remote assistance, virtual collaboration, etc. With the Meta Ray-Bans, they released a product that fades in everyone’s life with a practical usage of photo/video as well as AI. With Meta Connect next month, they are expected to showcase a prototype which blends these two worlds; with large ray-bans glasses that include a display, leverage hand tracking, and use AI to tie it all cleverly together.
With the quite low sales of the Vision Pro and related apps, Apple has reduced significantly advertisement for the headset, making almost no noise when releasing it in other markets such as Asia and Europe. Releasing many B2B related job positions, it seems that, like Magic Leap and Microsoft in the past, especially at this expensive price tag, they are shifting to B2B even though their products has been tailor made for individuals and not to be shared among individuals. For context, it requires a 15 minutes process to adjust to every individual, notably for eye tracking additionally to the custom made lenses when buying them. Rumors are saying they are hurrying to build a cheaper version as well as building glasses that would integrate AI in a Meta Ray-Bans fashion way.
The chinese competitor to Meta, ByteDance, building the Pico headset, also refocused in 2023 from B2C and B2B, to B2B only and just released a mixed reality version of their Pico 4 headset, competing with the Quest 3.
Xreal with their glasses – focusing on viewing content everywhere on a giant screen – has been an incredible sell compared to all the other consumer AR glasses of the industry and is probably the only company making great AR glasses at an affordable price for practical use cases.
c. What can we do about it?
For indie developers: AR / MR with AI: building the future of experiences will indeed involve AI. An interesting lead to take right now is building futuristic in AR or MR experiences leveraging AI, which can shape this future. If you want to make revenue from it, make sure to take part in challenges or paid programs from these companies, such as this program from Meta: the Meta Quest Lifestyle Accelerator. I also recommend checking out the glasses from Brillant Labs allowing to leverage AI on smartglasses with a third party SDK.
For companies and job positions: B2B / enterprise is what brings revenue today the most to XR related companies: Magic Leap, Apple seem to prioritize it, ByteDance Pico is focusing on it. Taking part of how these technologies can be useful in these companies is one way to make sure to contribute to how these technologies generate revenue, and where resources are the most allocated internally, while more consumer friendly devices are being built.
For AR production / game studios: building AR ads experiences on the social platforms is definitely a great path, but it’s quite a crowded market at the moment, so best is to have a strong network of clients already.
Conclusion
In conclusion, the state of Augmented Reality in 2024 reflects a maturing but still evolving technology landscape. While AR has definitely become more accessible and well-known, its integration into everyday life remains a work in progress, particularly in the wearable format that has long been anticipated. The golden age of mobile AR has passed, giving way to a period dominated by social AR and, more recently, mixed reality and AI-integrated smart glasses.
Consumer-friendly, all-day wearable merging AR and IA together in a practical device is clearly the next step and still a few years away, with current efforts focusing on mixed reality headsets and AI-enhanced smart glasses. Companies are increasingly shifting towards B2B applications, where revenue potential is more immediately realized. For developers and businesses, the emphasis now lies in leveraging AI to create innovative AR/MR experiences and focusing on enterprise solutions where the technology is most commercially viable.
As AR continues to develop, the key to success will be to keep aligning technological advancements with practical, consumer-friendly applications, and preparing for the day when AR+AI glasses are as ubiquitous as smartphones.
Quelle:
Foto: Igor Omilaev on Unsplash
