We Ran HomePod Through the Smart Speaker Gauntlet

Conclusion. Yesterday, we put HomePod through the smart speaker gauntlet which included 782 queries along with comprehensive tests of sound quality and ease of use. Our methodology was comprised of 3 HomePods being tested throughout the day by the Loup Ventures team. Here are four key takeaways.

  • HomePod is hands down the best sounding smart speaker.
  • As a smart speaker, it answered 52.3% of queries correctly compared to recent tests of Google Home at 81%, Alexa at 64%, and Cortana at 57%.
  • The user experience of HomePod is measurably better than its competition (setup, communication style, listening ability).
  • We have added HomePod to our Apple financial model. For CY18, we’re expecting HomePod unit sales of 7 million (12% global smart speaker market share) with a $349 ASP, which adds approximately 1% to revenue and earnings. In CY19, we expect unit sales of 10.9 million, which adds just over 1% to our model. While only factional revenue contribution, we expect HomePod units to grow between 40% and 45% per year over the next 3 years. Link to model here.

Query results. Siri understood 99.4% of queries and answered 52.3% of them correctly. This places HomePod at the bottom of the totem pole in terms of AI assistant performance. Siri is particularly strong in Local (where can I find a good cup of coffee around here?) and Commerce (help me buy some new shoes.) queries, handily beating Alexa and Cortana, but still falling short of Google Home in those areas. Overall, Siri performed above our expectations given the limited scope of HomePod’s music focus.

Adding domains will quickly improve Siri’s score. Some domains like navigation, calendar, email, and calling are simply not supported. These questions were met with, “I can’t ___ on HomePod.” Also, in any case that iPhone-based Siri would bring up Google search results, HomePod would reply, “I can’t get the answer to that on HomePod,” which forces you to use your phone or give up on the question altogether. Removing navigation, calling, email, and calendar-related queries from our question set yields a 67% correct response, a jump from overall of 52.3% correct. This means added support for these domains would bring HomePod performance above that of Alexa (64%) and Cortana (57%), though still shy of Google Home (81%). We know Siri has the ability to correctly answer a whole range of queries that HomePod cannot, evidenced by our note here. Apple’s limiting of HomePod’s domains should change over time, at which point we expect the speaker to be vastly more useful and integrated with your other Apple devices.

Other observations

  • HomePod has superior listening skills to other smart speakers. This is partly due to a noise cancellation feature which allows you, even at a volume where you would have to raise your voice to talk to others in the same room, to use your regular speaking voice with Siri. This was HomePod’s most stellar feature.
  • Wireless setup was super easy. In fact, it was the easiest of any of digital assistant we have used (Alexa, Google Home, Cortana, etc.). You already have a companion app and don’t need to wrestle with wifi networks.
  • Siri’s voice sounds smoother and more human than it does on your iPhone.
  • Her communication is also more human-like. Specifically, after asking a question, she does not repeat the whole thing back to you as is the case with Google Assistant and Alexa, which makes for a subtly smoother process.
  • The tap UI on the touch-sensitive display requires a small amount of instruction, as it is not immediately obvious that you tap to play.pause, double tap for next track, triple tap for previous track, and touch and hold to bring up Siri.
  • HomePod’s packaging is a new level of perfection for Apple. The perfectly fitted box requires you to open it at “reveal” pace, and even the external plastic wrap is pleasing to remove.
  • As a speaker, it sounds incredible – mission accomplished. As a digital assistant, whether it is a direct competitor or not, it is better than what we expected for version 1, but still lags behind Alexa and Google Home.

Survey suggests demand for HomePod similar to Apple Watch. Last week we surveyed 500 people in the U.S. and found 3.3% said they would purchase a HomePod in the next year. Among those surveyed who already own an Apple product, 5.2% planned on buying a HomePod. This is similar to 7% of Apple product owners planned to buy an Apple Watch ahead of its launch in Spring of 2015 (survey was conducted in December of 2014, four months ahead of the Apple Watch launch). While Apple does not disclose Watch unit sales, we estimate in its first 12 months Apple sold 10.2m Watches at an ASP of $475. This compares to our first 12 months of HomePod sales estimate of 7m units with an ASP of $349.

Smart speaker market share. We see Google Home as the long-term smart speaker unit share winner, but Alexa and Apple as the two other key players. In 2018 we expect HomePod will capture 12% of the global smart speaker units, compared to our estimate of Alexa at 52% share, Google Home at 32% and others at 4%. In 2022, we expect HomePod will hold a similar 12% market share (HomePod ASP estimate declines to $149 from $349 today), Google Home at 48% and Alexa at 37%.

Apple’s grander vision around HomePod.  Don’t be fooled by HomePod’s sound quality-focused first step into smart speakers; Apple has a grander vision than delivering a better sounding Echo. While not present in the first version of HomePod (i.e. you can’t even make a phone call with HomePod), we believe Apple’s goal is to make Siri a ubiquitous, ambient presence that connects and controls all your connected devices and services – and to make a leap forward in the transition to voice-first computing.

The way humans interact with computers is changing. Today, we use our keyboards, mice, and touchscreens to interact with computers, but in the future, we’ll simply rely on our voice, gestures, or even our thoughts. Voice is quickly becoming a preferred interface. Apple’s device ecosystem delivers a frictionless experience, which will only get better with the adoption of voice and with the addition of HomePod supported domains (9 supported domains today). Interestingly, Apple has included an A8 chip in its HomePod, the same chip included in an iPhone 6. The A8 chip is much more powerful than the chips competing home assistants run on, which poses the question: what else is Apple planning with the HomePod?

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

Nvidia Results: Intoxicating Combination of Faster, More Profitable Growth

What’s new. Shares of NVDA are up 7% after hours as the company reported an intoxicating combination of faster, more profitable growth. What’s new?  Tonight’s results are further evidence of the magnitude of the opportunity ahead of Nvidia. Its products are a foundational part of the future of technology, based on their use in data centers, autonomous vehicles, virtual and augmented reality platforms, cryptocurrency mining, and eSports. We remain believers in the long-term Nvidia story. While shares of NVDA are up 100% over the past year (market cap of $129 billion), we think there’s further upside given Nvidia’s foundational exposure to frontier technologies.

Earnings and model recap. Nvidia reported Dec-17 revenues of $2.9B vs. Street at $2.7B (up 34% y/y), and EPS of $1.72 vs. Street at $1.16. Updated model here.

Pole position in three green field growth markets: As we’ve written about previously, gaming, datacenter, and automotive, all have open-ended growth opportunities.

1. Gaming

Nvidia’s gaming business revenue was $1.74B, up 29% y/y. This was driven by high-quality, hit games in the market, notably PlayerUnknown’s Battlegrounds (PubG), Destiny 2, Call of Duty WWII, and Star Wars Battlefront II. PubG has reached 30 million players in 9 months, two months faster than Activision’s Overwatch. It’s worth noting the new Overwatch league launched last month and recorded 10 million unique users in its first week. The success of these games is leading to both an increase in the number of GPUs sold, as well as ASPs for Nvidia.

Additionally, demand for GPUs for cryptocurrency mining has boosted Nvidia’s gaming segment results. The rise in cryptocurrency demand has contributed to historically low channel inventory levels of GPUs being reported.

Long-term, the outlook for gaming remains promising.

I’ve always believed that the video game market is going to be literally everyone. In 10-years’ time, 15-years’ time there’s going to be another billion people on earth. And those people are going to be gamers. Not to mention that, almost every single sport could be a virtual-reality sport. – Jensen Huang

Part of the upbeat guidance for the April quarter is because gaming channel inventory is at historically low levels and the company expects to fill channel inventory during the quarter.

2. Datacenter

Nvidia’s datacenter revenues were $606M, up 105% y/y. For the 7th consecutive quarter, Nvidia’s datacenter revenues saw three-digit growth.

Driving datacenter growth are investments in artificial intelligence from a broad range of companies. Nvidia commented on the earnings call that every major cloud provider, including Alibaba, Amazon, Baidu, Google, IBM, Microsoft, Oracle, and Tencent have adopted Nvidia’s Tesla V100 GPUs for training deep learning networks.

We’re in the early innings artificial intelligence and just as all companies evolved to be internet companies in the early 2000s and mobile companies in the late 2000s, they will soon evolve to be AI companies.

3. Automotive

Nvidia’s automotive revenue was $132M, up 3% y/y. Despite little growth, remains Nvidia’s biggest opportunity. On tonight’s call, Jensen Huang outlined three near-term opportunities in Automotive:

  1. Nvidia’s DGX system is used to train neural networks for autonomous driving at data centers.
  2. Nvidia’s DRIVE PX platform provides cars with the necessary on-board computing strength for autonomous capabilities.
  3. Nvidia is reaching development agreements with major OEMs, ride-hailing companies, startups, tech companies, and many others to support their efforts in autonomy. Each project is engineering intensive, and companies rely on Nvidia for help.

Longer-term, the market for autonomous vehicles will be larger than most people think. Today, automotive accounts for 6% of revenue, and we expect it to rise to 13% by 2023. As we outlined in our Auto Outlook 2040, we expect 90% of vehicles on the road in 2040 to have level 4 or 5 automation, which would require a platform such as Nvidia’s DRIVE PX. 

Transportation is a $10 trillion industry. Between cars and shuttles and buses, delivery vehicles, I mean, it’s just an extraordinary, extraordinary market. Everything that’s going to move in the future will be autonomous. – Jensen Huang

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

Google Stays Focused on AI Theme

AI-First. We start with artificial intelligence, as Sundar has done the last five consecutive quarters. He continues to make it more than apparent that AI is essential to Google’s vision and will be a key component of their products and services. AI topics like Google Assistant, Waymo, AutoML, and AI-first hardware were highlighted on the call, but financial metrics like TAC overshadowed. Here is a look at Sundar’s opening remarks of the last five earnings calls.

  • Dec-17 (today) – “Thanks, Ruth. Our teams are off to a great start in 2018 (omitting 6 lines) Technology is an incredibly dynamic industry. We have been laying a foundation for the next decade as we pivot to an AI first company powering the next generation of Google products like the Google Assistant.”
  • Sept-17 – “Thank you, Ruth. We had another great quarter. (omitting 1 line) Even though we are in the early days of AI, we are already rethinking how to build products around machine learning. It’s a new paradigm compared to mobile first software, and I’m thrilled how Google is leading the way.”
  • Jun-17 – “Thanks, Ruth. We had a phenomenal quarter. Google continues to lead the shift to AI-driven computing.”
  • Mar-17 – “Thanks, Ruth. It’s been a terrific start to the year. (omitting 10 lines) Now, turning first to machine learning and access to information. I’m really happy with how we are transitioning to an AI-first company.”
  • Dec-16 – “Thanks, Ruth. 2016 was a great year for Google and 2017 is shaping up to be even more exciting. (omitting 11 lines) First, machine learning and access to information. As I’ve shared before, computing is moving from a mobile­first to AI­-first with more universal, ambient and intelligent computing that you can interact with naturally, all made smarter by the progress we are making with machine learning.”

Cloud. Google’s cloud business is gaining momentum as the world’s fastest-growing major public cloud provider. It is now a $1B per quarter business, but it still dwarfed by AWS revenue of over $5B per quarter. Sundar says the momentum is coming from their efforts to be enterprise scale ready. “Now we can handle any type of enterprise,” he says. We expect Google’s Cloud business to continue this growth in market share.

Hardware. As the largest contributor to “other revenue,” and with 2000 new HTC engineers, hardware is becoming increasingly important to Google. Hardware shipments “doubled” this quarter which means sales of Google Home and Pixel 2 have been strong. Further, when asked about monetizing the voice platform around Google Home, Sundar said they will remain focused on the user experience rather than monetization for “some time.” This means we can expect an ad-free experience for the foreseeable future. We think Google Home’s superior AI will lead it to steal market share from Alexa in the coming years.

Waymo. Waymo was brought up several times, but details about the recent order for vehicles from Fiat Chrysler and timing on fleet deployment were danced around. Nonetheless, the excitement from Ruth, Sundar, and analysts was palpable. We expect Waymo to be the first to bring a widespread fleet online and will be tuning in to see how early deployments play out in Phoenix.

Buzzword Bingo. Over the same five quarters, we tracked how many times presenters and analysts made comments artificial intelligence by tallying instances of AI jargon (AI, artificial intelligence, machine learning, TensorFlow, natural language processing, etc).  This is evidence of the intensity level at which Google is pursuing AI.

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

Apple is Flush With Cash, But Don’t Expect Big M&A

Don’t expect any big M&A. Everyone is asking who Apple will acquire. The answer is: nobody. At least nobody big ($5B+).

  • Apple announced they’re paying $38B in repatriation taxes, which implies they’re bringing $215B back to the U.S.
  • If there was no tax holiday Apple would have paid about $80B in repatriation taxes, compared to the $38B they actually are paying.
  • Apple also announced $30B in capex investment over the next 5 years, including a new Apple campus for technical support.
  • We don’t expect this infusion of U.S. cash to change Apple’s capital allocation strategy and believe share buybacks will be the primary use of funds.
  • We don’t expect any big M&A deals above $5B, but see the company continuing to do tuck-in deals and invest in their supply chain.
  • We believe 80% of Apple’s motivation related to today’s news is for economic reasons, 20% for political reasons, and both are good for the company long-term.

Potential uses of cash, by order of significance.

1. Returning capital back to AAPL investors. AAPL will increase its buyback over the next 3 years by about $70B. We also believe Apple will announce a one time cash dividend of $12B. Lastly, we anticipate a 15% annual dividend increase that will cost Apple about $10B over a 4 year period. We expect these three to be announced when Apple reports their Mar-18 quarter sometime in April. We believe most of this is already priced into AAPL’s stock price.

2. U.S. capital expenditures. As mentioned, Apple also announced $30B in capex investment over the next 5 years, including a new Apple campus for technical support. Separately, Apple increased its commitment to its Advanced Manufacturing Fund to $5B from $1B.

3. Invest in component suppliers. As they’ve done in the recent past (Corning, Finisar), Apple will continue to invest in key component suppliers, likely optics, displays, or processors, to lock out competitors and expand their lead.

4. Tuck-in acquisitions. This includes hardware, apps, and services. Hardware: Magic Leap’s AR hardware is a logical acquisition on the heels of Apple’s recent (summer of 2017) purchase of SensoMotoric. Magic Leap’s last private round was valued at $5.5B, suggesting this deal is slightly more than Apple typically would pay. Services: Apple Watch and its ecosystem would be bolstered by Peloton. This reminds us of Beats – hardware that adds to the lineup and, most importantly, a services play that would also expand the Watch/fitness data platform. Apps: We could see similar deals to the Lattice Data acquisition related to apps and AI.

5. Longshot: Auto. For starters, they should buy Tesla. But it will never happen because Musk won’t sell. Clearly, Apple believes in autonomy, through what appears to be a smart-shuttle approach. Since the shuttle initiative is early, it’s a high-probability M&A area for Apple.

6. Employee bonuses. Bloomberg is reporting that Apple will be giving to $2500 bonuses in restricted stock to most employees. We estimate about 100,000 employees will benefit, which implies a $250m liability that will vest likely in 2 years.

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio.  Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.

Investing in Enjoy

We’re investing in Enjoy as a counter-automation play on the future of retail. Read our thesis on retail’s future here. In short: retailers must either embrace full automation or compete on experience by focusing on uniquely human capabilities: creativity, community, and experience. We call it “empathic retail.” Enjoy delivers the future of retail by focusing squarely on empathic retail. Enjoy hand delivers products bought online from the world’s premier companies and delivers them with an experience. The service comes at no additional cost to consumers and it’s fast, with nearly 50% being delivered the same day.

Amazon is changing consumer expectations related to the price, availability, and delivery of products and services. But the in-person retail experience is outside of Amazon’s core competencies. Enjoy offers its premier companies (including AT&T, Sonos, DJI, and others) a high-touch, personalized delivery and setup service. Enjoy optimizes the customer experience, reduces returns, and increases customer satisfaction.

At the same time, automation technologies are already replacing retail jobs. Enjoy offers its team of Experts (delivery and setup employees) flexible work, salaried, with benefits – a transformative employment model for the new retail workforce. In our view, Enjoy is creating the optimal go-to-market channel for premium brands in the automation age.

Enjoy’s CEO, Ron Johnson, has spent his career innovating in retail. His experience as VP of Merchandising at Target, SVP and head of retail at Apple, and as CEO of JCPenney, along with his network of leaders at consumer electronics and luxury goods brands, uniquely positions Enjoy for success in these markets and beyond.

We’re excited to be a part of delivering retail’s future with the team at Enjoy.

Disclaimer: We actively write about the themes in which we invest: artificial intelligence, robotics, virtual reality, and augmented reality. From time to time, we will write about companies that are in our portfolio. Content on this site including opinions on specific themes in technology, market estimates, and estimates and commentary regarding publicly traded or private companies is not intended for use in making investment decisions. We hold no obligation to update any of our projections. We express no warranties about any estimates or opinions we make.