It's time to say goodbye to the Oculus Go
At the end of June it was announced that Oculus would be discontinuing the Go. Developers were informed that any Go apps in development must be submitted and approved for Release by end of this year. In early August, Oculus pushed out a new version of the Oculus Software Development Kit and it was quickly discovered that Go support has been completely removed. For developers who have spent many months or even years working on a Go title, reality has now sunk in. Oculus's first, affordable stand alone device; the darling product of 2018 that was in such high demand two years ago that resellers were commanding almost double the retail price, that product is now at the end of its life cycle.
In the communications Oculus sent to users they indicated that Go has been a successful product and that the retirement of the Go is in response to the community's reaction to Quest.
"... The community response has been overwhelmingly positive, and you’ve told us loud and clear that 6DOF feels like the future of VR. That’s why we’re going all-in, and we won’t be shipping any more 3DOF VR products. We’ll end sales of Oculus Go headsets this year "
On the surface the reasoning behind Go's retirement seems straightforward. What's being implied without being stated directly is that the Quest's success has been so great that they don't see a market for a media consumption VR device at half the price of Quest. What follows will be my perspective on what I believe to be a wider view into the story behind the Go's retirement. To be clear, I do believe Oculus's stated reason is honest. I just think, as is often the case with vendor to consumer communications, that there's a lot more to the story under the surface level communications.
An abridged history of the Oculus Go
For the remainder of this blog post I'm going to discuss the history of the Go, then I'm going to dive a bit deeper and discuss some of the non stated reasons I think the Go has been retired.
While Oculus's first consumer product was the DK1 (Developer Kit1, released 2012). The company's first mass market, refined product offering came in the form of a partnership with Samsung. The brilliant minds at Oculus knew the capabilities of mobile processors and realized the potential of co-engineering and co-marketing a phone VR based headset. Let it be noted that Oculus wasn't the first to enter the phone based VR game. In 2014 at their annual developers conference, Google introduced Google Cardboard and went so so far as to give all conference attendees a free cardboard viewer. The concept of the Cardboard viewer was to bring VR capabilities to all moderate to high-end Android devices and to provide an open spec for the viewer so that any vendor could implement their own Cardboard viewer. Although Google was first to deliver something to developers, the GearVR story predates the release of Cardboard.
According to Samsung's blog. GearVR was in prototype as far back as 2013. This means that it's likely the collaboration with Oculus existed prior to Facebook's acquisition of Oculus.
The think tank at Oculus knew from research the importance of setting the bar high for mobile VR. When certain FPS and refresh thresholds aren't met, it can have a very negative impact on the user experience and if someone gets sick from using VR then chances are you're going to have a very difficult time getting them to give VR another chance. Google's Cardboard didn't take this into account. It seems Google had no problem dishing up horrible quality VR experiences in order to say "me first" and to utilize the novelty of VR as a cheap conference giveaway.
It's my opinion that Cardboard devices and Cardboard apps have done more to damage the overall public perception of VR that anything else. In 2020 it's very difficult to explain to a relative that got a $20 VR viewer from Walmart for Christmas three years ago the difference between that and where we are today with stand alone devices. Again, bad VR experiences are very hard to forget. Oculus and Samsung were on the same page with desiring a quality experience. By working with a single vendor Oculus was able to write low-level, native code specific to the SoC chipsets Samsung was using. By restricting GearVR to only specific phone models capable of hitting the desired benchmarks, Oculus and Samsung were able to deliver the highest quality mobile VR experience leagues above what Cardboard could offer.
In Nov of 2015 with much promotion through mobile service providers, GearVR launched. Promotion packages would go on to provide the headset for free or discounted to compatible Samsung phone buyers through select carriers. This enabled GearVR to gain much market share in a short amount of time. Chasing the same market (and realizing the limitations of Cardboard), Google aimed for a higher standard and launched the Daydream viewer the following year. While the first GearVR was a gaze based device, Daydream upped the ante and provided a 3DoF remote control. Samsung and Oculus followed with a revised version of GearVR that included their take on the 3DoF remote controller the following year.
Outside the benefits of rendering more powerful SoC's bring to the equation, from a hardware interface standpoint, phone based VR had reached its threshold by 2017. Oculus had already begun working on their own stand alone devices and the 2017 version of the GearVR would end up being its last incarnation.
A game changer in the works
The development of the Oculus Quest was made official in Oct 2016 when at Oculus Connect, Mark Zuckerberg announced the development of a stand alone HMD that would feature full Roomscale inside out tracking. The project was code named Santa Cruz and it would go on to become what we know today as the Oculus Quest. At the same time, Oculus was also working on a "stand-alone device similar to the GearVR". By leveraging Samsung's existing market placement and co-branding GearVR as a Samsung/Oculus product, Oculus had made great headway establishing their brand. Thanks to promotion offers through major US cell carriers the GearVR had become the best selling VR headset on the market. At that time the partnership with Samsung had really fulfilled its potential.
Thanks to the open nature of the Android OS, the transition to an all-in-one device built with a new hardware partner would be relativity easy for Oculus and a standalone product would give Oculus complete control over the hardware and software. Oculus found a new partner in one of the largest mobile phone vendors in the world, Beijing based Xiaomi
The Xiaomi partnership made sense on a few levels, Oculus was in need of a hardware partner and Xiaomi wanted to develop an all-in-one VR device, but needed great software. Xiaomi had no market share in the United States and Oculus had no market share in China. So the pairing provided a great opportunity for both companies to help one another without stepping into the others space. While the details of the deal negotiated with Xiaomi remain confidential, we can speculate based on the outcomes that the deal gave each partner control and primary branding in their respective territory. Xiaomi's hardware would power the Go, but Xiaomi's visible contribution reduced to a small Mi logo on the side of the device. In China Xiaomi would sell the Go with a slightly different external aesthetic, under the Mi VR Standalone product name, powered by Oculus software.
The Go provided a means for Oculus to sunset the Samsung partnership and at the same time maintain Oculus's 3DoF VR platform. It's reasonable to believe Samsung was taking a cut from all Samsung Gear VR app sells. With Go, Oculus would own the hardware and software and all of the vendors cut of app store revenue would go to Oculus.
Another issue with GearVR was the variation in spec. Mobile SoC (system on a chip) processors were seeing exponential performance gains in a matter of years. For example, the Samsung Note 4 was the first phone to support GearVR and it had a Qualcomm Snapdragon 805 processor which used the on-board Adreno 420 for GPU. By 2017 the Samsung Note 8 phone had a Snapdragon 835 with Adreno 540. This is the same SoC Oculus would go to to use for the Quest and many GearVR owners had that power in device in 2017! Looking at 3DMark scores the difference between what the Note 4 and Note 8 is significant. Depending on the specific benchmark looked at it's anywhere from 50% to 170% increase in capabilities. But judging the quality of available GearVR experiences the users wouldn't notice. On a PC game developers bake in graphics settings to allow the player to adjust for their hardware, but on a mobile platform that's a lot more difficult. It's more challenging because there's limited storage capacity. There's not enough room to provide different texture sizes and other device spec specific assets. So that put app developers in a position to target the lowest common denominator to gain market coverage. The mobile VR experiences for GearVR weren't evolving.
By being able to lock down the hardware spec with Go, Oculus could provide developers a standardized spec to target for performance. It also enabled Oculus engineers to do some tinkering and we ended up with some features on Go that never made it to GearVR like Fixed Foveated Rendering and Subpixel rendering. Some of these features continue to be the best option. For example the LCD screen on the Go enables it to get higher sub-pixel resolution than the Quest and LCD in general provides lower-persistence than LED based displays. making it easier on the eyes for watching video play black.
The Go arrives
On May 1 2018 the Go was released. Retailers quickly sold through their initial stock and the Go became a scarce commodity for the first four months after its release. During this time and up through at least Dec of 2018, Oculus ran a series of TV and Web spots. The ads featured various celebrities and overall the ads were considered by many to be a pretty solid ad campaign.
No doubt, Oculus had a successful launch and a hit with customers in the Go. Five months later at Oculus's Connect developer conference Mark Zuckerberg announced that the Santa Cruz R&D project had concluded. The product coming out of this effort would be known as the Oculus Quest. It would be the first all-in-one VR headset with full 6DoF head tracking and 6DoF controller support. It would retail starting at $399 and be available in Spring 2019. Go had a successful 2018 Holiday season and judging by Amazon's best seller list in Video Games category had an even better 2019 Holiday season consistently ranking in the top 5 spots for Amazon's Video Games category, right below the Oculus Quest.
The Sunset begins
I believe Oculus had plans to sunset Go long before the announcement. I believe the end for Go was probably being discussed at Oculus as far back as July 2019. In August 2019 right before OC6 (Sept 2019) Venture Beat published an article discussing how Xiaomi had discontinued the Mi VR Standalone and disbanded the VR team behind the product. Did Xiaomi stop producing Go for Oculus at that time? No. But the news of Mi VR Standalone being discontinued did mean that if Go was to have a successor other than the Quest it would not likely be coming from Xiaomi. It also meant that Xiaomi had abandoned their investment in the composite Oculus-Xiaomi product. For whatever reason, Xiaomi was done the partnership in terms of leveraging it in the Chinese market. Just how much impact this had on Oculus's decision to discontinue Go for the US and other markets may never be known, but more than likely it added some weight to the decision. When I read that Xiaomi was discontinuing the Mi VR Standalone, I assumed the Go had already been discontinued.
To add to speculation about the mid 2019 Go retirement theory, John Carmack delivered a rather emotional eulogy for GearVR at OC6 in which along other things he talked about preserving early VR titles for future generations and gaming historians to explore. If you go back and listen to his keynote discussion regarding the retirement of GearVR he seems to be having a eulogy for 3DoF in general. I will speculate at that time Oculus already knew they wanted to get out of 3DoF VR and probably were having some internal debate on how much of the GearVR/Go catalog they should preserve.
It has also been noted on Reddit and other on-line forums that all Oculus Go marketing/advertising efforts seems to have ceased a few months prior to Oculus Quest being released. I recall the last Oculus Go ad spots appearing around March 2019. When a company is behind a product they are going to advertise that product. For whatever reason Oculus had moved on from pumping money into further advertising the Oculus Go.
In retrospect, I believe Carmack and a few other insiders at Oculus already knew the fate of the Go at OC6. There are other tiny clues that I think this was probably the case. For example, at OC6 John Carmack held a app review session where he only accepted Go apps to review. From all he has said over the past year, my intuition tells me Carmack was to some degree frustrated with the overall handling of the Go. At the very least he felt it never really saw its full potential.
In the months following OC6, updates were delivered to the Quest to enable more Go compatibility and a small curated set of Go experiences were allowed to live in the Quest store, available to owners of the Go license for the apps.
Go's mark on Enterprise VR and educuation- the roots of Oculus for Business
In 2017 Oculus began quietly working direct and indirectly with handful of Fortune 500 companies to developer Enterprise VR experiences. Walmart was among the companies involved the pilot program, purchasing 17,000 Oculus Go stand-alone headsets and developing a custom kiosk training simulation for the Go. The project was a big success for Walmart and its story documented on Walmarts blog
The day prior to OC6 Oculus held their first annual Oculus Education Submit. The event featured many budget sensitive projects made possible by Go. The price point and ease of use made Go a huge win overall for media consumption and for many training scenarios.
What went wrong?
Great products sold at a great price may seem to consumers like an obvious success, but the true mark of a product's success is how much money it makes for the companies and investors who took the huge risk on bringing it to market. If a product can't make money, regardless of how well it's designed and how much it's loved, it's going to be a failure.
The Oculus Business model for the Go pivoted around a primary revenue stream coming from app store commissions. As with the Apple and Google Play app stores, Oculus takes a 30% cut of gross profits from app sales and in-app purchases. The actual build cost of the Oculus Go remains unknown, but at the time of it's launch it was speculated to be around $155-$195 USD for the entry level $199 model. The cost break-down on wholesale/OEM parts went something like this
|Snapdragon 835 (SoC)||$50-$60|
|Other electronic components
|Manufacturing/import cost per unit||$15-$25|
Big box retailers like Walmart and Best Buy who have agreements to sell Oculus hardware typically make about 10% margin on video game console sales. Which means to hit the $199 price point on Go, Best Buy would have to buy them from Oculus at around $180 per unit. When you factor in Oculus's marketing and distribution cost on Go, it's impossible that the company was making much, if anything on Go hardware sales. It's quite possible Oculus was even taking a loss on every unit sold through a big box retailer and recovering that loss with their direct sells. So what about software? It's pretty clear that Oculus followed the traditional game console model of having break-even strategy for hardware revenue with accessories and app royalties being the revenue streams.
Another Chicken-and-egg problem
Oculus decided to make Go compatible with GearVR catalog. From a hardware standpoint this compatibility was inherent, but from a business stand point Oculus felt it would make sense to help GearVR developers bring over their apps and to grandfather in the majority of the GearVR's 1000+ app catalog for the Go launch. One of the problems with this approach would be the price point expectations for Go apps being heavily influenced by GearVR app pricing. On GearVR the average price of a VR game app was $5-$10. That's the market value for a mobile based game, but when you consider the 30% commission that the store takes that's $8 a game. Account for taxes and the net revenue on a $10 app sales is about $5. Keep in mind on a game console like XBox One or PS4 the average cost for a full featured AAA title with 30+ hours of game play is about $14 million. A successful novelty VR title like what we saw with Rush or Ultrawings on Go, has an estimated development cost of $2.5-$4 million.
So to even do an high-quality Indie title like an Ultrawings or Thumper, the game studio is taking a big risk. They're going to have to sell through 500k to 1.5 million copies to break even on development cost then probably 250k-500k in unit purchases to cover marketing, continued development, support, etc...
The Go needed AAA quality experiences and the money wasn't there to lure in AAA studios. So it's a chicken-and-egg problem. To continue to gain market AAA quality titles are needed, but because of market value of games on the device no AAA studio wanted to develop for it and loose money on it. To be fair, it wasn't just the pricing but also the limitations of 3DoF. I think there's a great variety of games that can be enjoyed in a VR headset with just a game controller, but those type of experience don't have much novelty to them. They don't show off all that VR is capable of and so they don't do a good job selling VR to new customers.
Also Oculus was hoping media consumption partnerships would play out and lead to revenue streams. We can infer this by the app catalog. The launch catalog for Go was heavy on media consumption- Apps for the The New York Times, FOX Sports, CNN, Netflix, etc... There never appeared to be a business plan tied to making revenue from media partnerships and its hard to know what Oculus was hoping for, but a vision for media consumption revenue had to be in place and this hasn't played out very well. It's possible some deals were made, but if they were the user base and interest just hasn't been there and I'm sure those deals have since dissolved.
That leaves the area of Enterprise VR as a potential revenue stream and when Oculus For Business launched in Sept 2019 the Go and Quest were initially stated to be supported. For whatever reason, Oculus decided to officially drop Go support for the Oculus for Business program in Feb 2020 (again, I believe Oculus's plans to retire Go have been in the works for over a year). While there continues to exist a market for 3DoF VR in training and classroom settings, it probably just didn't make sense to the powers that be at Oculus to keep a 3DoF product around just for this market segment. By exiting the 3DoF market altogether other, products like the new Lenovo Mirage VR S3 will have a better shot and will fill the gap.
Some final thoughts
I think the Go will be remembered fondly. It has been Oculus's first composite hardware/software all-in-one product to begin building an ecosystem. To that end, it paved the way for Quest by allowing Oculus to establish better channels for developer and consumer support. Lessons were learned about partnering to outsource hardware engineering and manufacturing. I see the Go as having provided a huge step in helping Oculus get to where they needed to be to successfully launch the Quest. As much as I love my Go and will continue to use it while I can, I think I understand Oculus's reasoning for retiring the Go.
On the subject of speculation. Many of us believe that Oculus will fill the void left by the Go. Leaked photos of yet-to-be-announced Oculus hardware have leaked reveal a light grey all-in-one headset that bears a lot of the aesthetic characteristics of the Go, except with full insight-out tracking cameras and 6DoF controllers. I'm sure in the coming weeks we'll learn the official details of what this device is, but marketing folks can tell you color selection, material choices and overall device aesthetics are typically used to imply certain product categories. For example, the budget iPhone S lineup isn't available in the same color options as the higher end models. There's a great deal of research that goes into the psychology of all of this and I'll boil it down by saying the light grey color along with some other value aesthetic choices on this device tell me that it's probably intended to occupy the value/entry level category left void by the Go.
Only time will tell.