https://www.matthewball.vc/all/bigtechbiggestbets

By my estimates, Meta has spent roughly $52.7B on its Reality Labs division since 2012, with roughly $7.2B in cumulative revenue over the same period, resulting in roughly $49B net loss. The exact figures aren’t available, but we know the division was effectively formed in 2012 following the $2B acquisition of Oculus VR, and Meta has disclosed its spending from Q1 2019 through Q1 2023. Given these data points, the early years of the division are relatively easy to model and can be cross-referenced against sales records and various statements by Facebook throughout the decades. Revenues and expenses have been comparatively modest during that time, and thus assumption errors are less consequential. But even if we just use Meta’s disclosures and guidance for Q2 2023, the company is out at least $42B net. And this sum continues to grow.

2023 and 2024 should see Meta’s Reality Labs achieve new revenue highs. The Meta Quest 3, set to debut in October, boasts substantial improvements over 2020’s Quest 2 (+30% resolution, standardized 120 FPS, a mixed reality mode, a lighter and thinner form factor). It will also coincide with the launch of Roblox on the Meta Store, which is likely to become the platform’s most popular application (potentially orders of magnitude more popular than Meta’s VR-only Roblox competitor, Horizon Worlds). Although Apple’s forthcoming headset will create competitive pressures for Meta, which has heretofore enjoyed 90% market share of new device sales, but it will also produce a halo effect.

Apple is one of the most beloved companies in history, and its debut in the XR headset category will have the effect of legitimizing it while also helping consumers better understand its use cases and benefits—something Apple has done more frequently and consistently than any tech company. Reports also suggest that Apple’s device will cost $3,000—six times the cost of the Quest 3, which will probably drive many households that want the Cupertino company’s headset to instead go with the one from Menlo Park. Amazon saw a similar benefit with its Fire Tablet, which launched a year after the iPad, but at 40% of price.

Over its first 12 or so months, the Quest 2 outsold the Quest 1 (May 2019) by a factor of 10, or 10MM to 1M. Such a steep change is unlikely for the Quest 3, but some multiple seems well within reach. Furthermore, these sales may sustain better than the Quest 2, which saw sales slow considerably in its second and now third years in market—during which Meta’s costs surged, leading to a doubling of annual net losses. Meta is also suggesting that the device will no longer be sold quite so far below cost (Meta likely lost about $150 per Quest 2, i.e. $1.5B for every 10MM units). This will harm sales but also reduce losses. With this in mind, it’s likely that Meta’s losses peak in 2023—perhaps to the tune of $16B or $17B—and then improve every year thereafter. Still, annual net profits seem far off. And if so, then Meta’s cumulative losses might cross $80B first.

These figures are stunning. They partly, though not exclusively reflect how brutally difficult XR devices have proven to be since Meta, Microsoft, Google, and others began investing in the form factor a decade and a half ago. At the time, many believed that there would be a few hundred million “wearable” supercomputers in use by 2023, if not more. But the number of technical challenges, as well as their difficulty, has gotten in the way of these sales, even as it amped up the investment required to get to a “minimum viable product” (I wrote about these challenges in detail here). To this end, Meta reports that 90% of its investments in the category are hardware-related innovations in optics, microLEDs, batteries, cameras, sensors, and the like.

Related Essay: Why VR/AR Gets Farther Away as It Comes Into Focus

The difficulty of producing an “MVP” XR headset has prompted most of Meta’s competitors, with the exception of Apple, to effectively abandon their XR ambitions—or at least to place them into hibernation pending new (and external) advances in the field. For Meta, these changes are both good and bad news. On the one hand, the competitive field has thinned. On the other hand, the company’s bet is increasingly contrarian, increasingly behind schedule, and substantially (for this very reason) more costly. For these investments to generate a positive return, they’ll need to generate at least $100B in cash flow—preferably in the next 15 years. Such a sum is theoretically achievable. In fact, the very product that Meta’s XR devices are designed to supplant—smartphones—have generated several times as much just in the past decade alone. But this is no ordinary watermark—after all, the iPhone is the most successful product in the history of capitalism.

Meta’s metaverse investments are high-profile due to the company’s name change, purported focus, and most importantly, decision to disclose these investments separately. But these investments are far from unique. In AR/VR, for example, Snap, which has a $15B market cap, has spent at least $1.5B acquiring AR startups in the last five years, plus billions more in R&D (20% of their 5,000+ employees focused on AR hardware in 2022) as well as its investments to publicly launch four Spectacles AR models. These proved unsustainable for Snap, which effectively shut down its efforts to build a device and laid off all dedicated employees in the back half of 2022 and early 2023. Yet AR/VR is not the only computing category that has seen such extraordinary investment with, thus far, limited financial return.

Beyond AR/VR

In his April 2015 letter to shareholders, Amazon founder/CEO Jeff Bezos listed the four requirements for a “dreamy business”: “Customers love it, it can grow to very large size, it has strong returns on capital, and it’s durable in time—with the potential to endure for decades.” Bezos added that Amazon (and thus its shareholders) had three such businesses—Marketplace, Prime, and Amazon Web Services—and that he hoped to build more.

Everyone wants a “dreamy business,” but part of what makes Amazon so unique is that the creation of new businesses is one of its core competencies, as Benedict Evans and Ben Thompson have varyingly described. Amazon has created or pioneered dozens of other businesses—often in areas where it had no prior experience, such as its dreamy cloud services, as well as the Kindle business, which is not dreamy but helped establish the ebook business and aided the company’s later investments in hardware. Creating new businesses is an extraordinary talent and arguably the most important a company can have over the long run.

But 2014, the year that Bezos’s letter addressed, saw the largest new business failure in Amazon’s 20-year history. In July, the company released its first (and only) smartphone, the Fire Phone. Unlike Amazon’s other hardware plays, it was fairly late to market. The Kindle was not the first e-reader, but it effectively pioneered the category. The Kindle Fire tablet debuted barely after the iPad. The Fire Phone, meanwhile, came seven years after the iPhone, six after Android, and at a comparable price. Amazon hoped to differentiate the Fire Phone through the integration of Amazon’s e-commerce platform as well as its use of advanced “3D cameras and sensors” that could scan and understand a user’s local environment (ideally to help the user purchase something that caught their eye). The device was an outright flop, terminated barely three months after its debut, with an Amazon-record $170MM writedown in the process (nearly half of which was for unsold inventory). A decade has passed since the Fire Phone, and it is still the company’s largest writedown and flop.

Normally, a dud like the Fire Phone would squash a company’s ambitions—at least when it came to building consumer computing devices. Making matters worse, the Kindle, while still a success, was in its third consecutive year of decline as tablets and smartphones robbed it of share and the annual updates to various Kindle models became more marginal, leading many current Kindle owners to replace their old devices only when they had broken. And while the Fire Tablet was still doing well, it was mostly a low-cost TV player, rather than the computing platform Amazon had originally envisioned. But the company was not deterred; dreamy businesses aren’t easy and they’re rarely cheap. AWS, after all, had cost tens of billions to get started.

In the years that followed, it was increasingly important to Bezos that shareholders understand the difficulty of big bets as well as their rationale. In 2016, he told The Washington Post’s Marty Baron, “If you think that’s a big failure, we’re working on much bigger failures right now—and I am not kidding,” he said. “Some of them are going to make the Fire Phone look like a tiny little blip.” In his note to shareholders the following year, he reiterated the point, writing, “As a company grows, everything needs to scale, including the size of your failed experiments. If the size of your failures isn’t growing, you’re not going to be inventing at a size that can actually move the needle.” Bezos also argued that the typical analogies for success actually limit the appetite for risk: “We all know that if you swing for the fences, you’re going to strike out a lot, but you’re also going to hit some home runs. The difference between baseball and business, however, is that baseball has a truncated outcome distribution. When you swing, no matter how well you connect with the ball, the most runs you can get is four. In business, every once in a while, when you step up to the plate, you can score 1,000 runs.”

By this time, we also knew which specific businesses Bezos thought might be the next home run. A few months earlier, he told Recode’s Walt Mossberg that Prime Video and Alexa were the top prospects for a “fourth pillar” at Amazon.

At the time, Prime Video was spending an estimated $3B on content annually, a figure that has since grown to nearly $7B. We don’t know the revenue, nor the broader cost basis, but cumulative content investments are probably about $35 billion. Even if Prime Video is, on a direct basis, unprofitable, the upside is obvious because the market already exists and the opportunities in digital delivery is clear. Video is the most valuable leisure category globally, generating more than $700B in revenue, but less than $150B is currently streamed. Among subscription video services, Prime Video is now the third-largest globally. It also helps Amazon participate in the rest of the digital video space, such as TVOD (where Amazon has a 40% share of a $20B+ market, collecting 15–30% gross margins), subscription reselling (25% share in a $25B market, collecting 15–30% gross margins), and connected TVs (40% share). Prime Video also provides Amazon’s enormous advertising business with more ad inventory and expands its ad product into premium video and, most importantly, live premium video.

Alexa was Amazon’s more speculative bet. Rather than disrupt a century-old market while it transitioned to a new medium, Alexa sought to pioneer a new market and medium altogether. Instead of trying to participate in a multi-player ecosystem, it wanted to serve as the foundation for one. Beyond building out a new business line for Amazon, Alexa would drive Amazon’s entire business—from e-commerce to AWS, Audible to Kindle. And while Prime Video, which launched in 2011, was launched out of Amazon’s five-year-old digital video store, Alexa was brand-new. What’s more, it debuted only four months after the Fire Phone was launched and one month after it was canned. Alexa was also slow to start, selling 2.8MM units over its first two calendar years. But sales took off in 2016, with 8.5MM units shipped, while 2017 saw sales grow more than 300% to 25MM. In 2021, over 65MM units were shipped.