Adrian Watkins

Alphabet is using its dominance in the search and advertising spaces — and its massive size — to find its next billion-dollar business. From healthcare to smart cities to banking, here are 10 industries the tech giant is targeting.

With growing threats from its big tech peers Microsoft, Apple, and Amazon, Alphabet’s drive to disrupt has become more urgent than ever before. The conglomerate is leveraging the power of its first moats — search and advertising — and its massive scale to find its next billion-dollar businesses.

To protect its current profits and grow more broadly, Alphabet is edging its way into industries adjacent to the ones where it has already found success and entering new spaces entirely to find opportunities for disruption. Evidence of Alphabet’s efforts is showing up in several major industries. For example, the company is using artificial intelligence to understand the causes of diseases like diabetes and cancer and how to treat them. Those learnings feed into community health projects that serve the public, and also help Alphabet’s effort to build smart cities.

Elsewhere, Alphabet is using its scale to build a better virtual assistant and own the consumer electronics software layer. It’s also leveraging that scale to build a new kind of Google Pay-operated checking account. In this report, we examine how Alphabet and its subsidiaries are currently working to disrupt 10 major industries — from electronics to healthcare to transportation to banking — and what else might be on the horizon.

1. Consumer electronics
Artificial intelligence could be the key to Alphabet owning the space.

Within the world of consumer electronics, Alphabet has already found dominance with one product: Android. Mobile operating system market share globally is controlled by the Linux-based OS that Google acquired in 2005 to fend off Microsoft and Windows Mobile.

Today, however, Alphabet’s consumer electronics strategy is being driven by its work in artificial intelligence. Google is building some of its own hardware under the Made by Google line — including the Pixel smartphone, the Chromebook, and the Google Home — but the company is doing more important work on hardware-agnostic software products like Google Assistant (which is even available on iOS).

Google hasn’t demonstrated a strong ability to compete with Apple on hardware — but in Alphabet’s vision of the next generation of consumer electronics, AI will be the critical differentiator, not hardware. Google’s reach through consumer apps like Maps and Assistant, plus the extensive adoption of Android, means that Alphabet can ultimately have a disruptive impact across software and hardware, from TV to voice assistants to watches to automakers.

Though the company has seen several of its products struggle or even fail due to tepid public reception or internal management difficulties — Google Glass, for example — its AI product launches, like Google Assistant, have shown more promise.

Android is the foundation of Google’s work in electronics. It has been the best-selling mobile OS every year since 2011, and today has about 85% market share in the smartphone market, with iOS accounting for most of the remaining 15%.

In 2014, Google launched its plan to expand the Android operating system to a range of other devices, which included a wearables project called Android Wear and Android TV.

In 2017, the tech giant released a selection of new “smart” home devices that are not reliant on the Android ecosystem: the Google Home, new Chromecast devices, a new VR headset, and a new Chromebook laptop with built-in Google Assistant.

Since its launch, Google Assistant has been rated far more capable and useful than its main competitors, Microsoft’s Cortana and Apple’s Siri. Driven by Google’s internal deep learning-focused Tensor Processing Unit chips (discussed later in Next-gen computing), Google Assistant is available on all Android and Google Home devices and runs on a reported 1B devices. Android itself is running on more than 2.5B devices today, while Android Wear is available on smartwatches from companies like Michael Kors, LG, and more.

Now automakers are coming to Google for help building out software for their in-car infotainment systems. Volvo, General Motors, Fiat Chrysler, and Nissan will all be working with the company to integrate Google Assistant, Google Maps, and Google Play Store into upcoming makes and models.

This approach is Google’s trademark play in consumer electronics: not building its own branded products, but building software that powers compelling consumer electronics products. Consumer demand puts pressure on watchmakers, automakers, and other companies to integrate tools like Google Assistant and Google Maps into their products. Even Microsoft is now releasing an Android phone.

With Android’s flexibility and reach, it can extend its software to a wide range of devices from different manufacturers.

In this sense, Google’s disruptive potential may have less to do with knocking out Huawei, Samsung, and other traditional electronics manufacturers and more to do with owning the software stack that makes those electronics work.

2. Healthcare
Alphabet is focusing on partnerships and leveraging machine learning to tackle a wide range of healthcare issues.

One of the industries that Alphabet is most engaged in today — and where it has the highest likelihood of having substantive effects on the world — is healthcare.

Alphabet is learning from the mistakes it made trying to go it alone in the healthcare space. Today, Alphabet is partnering with major institutions, focusing more on realistic applications rather than moonshots, and working on a wide range of solutions to different health problems. As shown below, work on these disease spans organizations under the Alphabet structure.

It might seem like a distraction for a company like Alphabet to invest so much in a field like healthcare — at least compared to autonomous cars, where it has a clear technological and engineering head start. But healthcare presents numerous $100B+ market opportunities, including pharmaceuticals, cancer research, and diabetes treatment — and Alphabet is betting that its superior machine learning capabilities and culture of innovation can allow it to find at least a few winners.

Google’s earliest attempt to disrupt healthcare was hamstrung by process issues and difficulties integrating with legacy tech. Google Health, founded in 2006, was a personal health record (PHR) service, which aimed to connect doctors, patients, and pharmacies. The main problem it faced was that it couldn’t pull in external data from a wide variety of third-party providers, which made it hard for most patients to use, and the service was discontinued in 2012.

Today, Alphabet’s focus is on forming valuable partnerships and applying its strengths — machine learning, artificial intelligence, robotics, and data — to a range of problems.

For example, Alphabet has run several AI-focused healthcare projects through its DeepMind subsidiary since 2016, including tests to better diagnose breast cancer and eye disease. Meanwhile, Calico is focusing on how AI can help extend the human lifespan and slow down the aging process.

Through Verily, Alphabet is working on applying technology to a broad array of life sciences issues, with separate projects dedicated to studying diabetes, cancer, wearables, robotic surgery, population health, and pharmaceuticals. Alphabet is working in partnership with existing institutions in these fields around the world, resulting in numerous clinical trials, patents, scientific papers, product launches, and community health initiatives. This suggests that Alphabet takes healthcare seriously as a pillar of its future portfolio.

Diabetes
So far, diabetes has been one of the most productive fields of inquiry for Verily, generating a series of pilots and relationships that have resulted in several clinical trials of its technology.

Verily partnered with Nikon subsidiary Optos to work on better detection methods for early diabetic eye disease in 2016. In 2019, the companies announced that the first clinical trials of their machine learning algorithm for detecting diabetic retinopathy were live in Aravind Eye Hospital in Madurai, India.

With this technology, a technician takes a photo of the patient’s eyes and uploads it to the Verily algorithm, which immediately screens for diabetic retinopathy and diabetic macular edema. The technician can then refer the patient to an eye care physician if necessary. On December 18, Verily announced a similar partnership with Thailand’s Rajavithi Hospital.

In 2016, Verily formed a new company, Onduo, with the drug manufacturer Sanofi. Though Sanofi restructured its role in the joint venture in 2019 amid a change in corporate strategy, it is staying on as a financial backer. Onduo’s offerings include a mobile app that analyzes glucose data from a cellular-connected blood glucose meter and photos of patients’ meals to provide nutritional guidance and insights.

The Onduo for Diabetes app includes access to past glucose readings and data, as well as the ability to reach out to a personalized care team for more information. In 2018, the company announced a plan to develop a new type of “all-in-one” insulin patch pump that’s simpler to use for patients.

In 2019, Onduo partnered with Orpyx Medical Technologies to make the SurroSense Rx system available to certain members of the Onduo diabetes management program. The system consists of a thin wearable that goes inside a patient’s shoes and alerts them when hazardous levels of pressure have been reached, helping to protect them from foot ulcers and issues that can lead to limb loss.

All of these efforts could be extremely valuable for Verily and for Alphabet as a whole. The global market for diabetes devices is projected to be worth $38B by the year 2026, propelled mostly by increasing incidence of the disease. That said, Alphabet isn’t the only company working on diabetes. Amazon, for example, is selling blood sugar monitoring devices directly to consumers and using Alexa to make it easier for diabetic patients to understand their current health. Meanwhile, Apple is more focused on creating integrations between existing diabetes devices and the Apple Watch and iPhone.

Robotic Surgery
Verily has identified another big opportunity in robotic surgery, a discipline similar to telemedicine designed to make surgery more accessible to parts of the world lacking in physicians and doctors.

Working with Johnson & Johnson, Verily launched Verb Surgical, a company with the aim of “changing the future of surgery to enable better patient care,” in 2015.

Verily’s robotic surgery patents show that the company has made significant progress in developing tools for doctors to conduct tests remotely. One patent describes an abdominojugular reflex test, which Verily claims can be performed more accurately using a pressure cuff and camera than by an in-person physician.

At the end of 2019, Johnson & Johnson announced that it would be acquiring the remaining stake in Verb Surgical, bringing the team and technology onboard as it continues working on the mass rollout of surgical robotic techniques.

Community Health
While much of Alphabet’s work in healthcare has focused on using AI and other advanced technologies to investigate specific diseases and improve treatment delivery, Verily and Sidewalk Labs have run projects exploring more local initiatives aimed at improving the health of entire communities.

In February 2019, Verily launched the OneFifteen project, a not-for-profit treatment center dedicated to studying opioid addiction and helping treat addicts in Dayton, Ohio.

The project also aims to generate data on which environments are best for treating addiction, what kinds of additional care are useful (including “recovery housing and vocational training”), and what types of in-person treatment are the most effective at helping addicts stay clean.

The Sidewalk Labs-incubated Cityblock Health started as a research project focused on the effects of urban living environments on health, then pivoted to setting up in-person “health hubs” in areas with high percentages of Medicaid and Medicare customers.

Today, Cityblock Health assists low-income patients with accessing care, and helps care providers better serve them. The company focuses on high-risk beneficiaries because this relatively small group of benefits recipients drives “the majority of the healthcare costs for Medicare and Medicaid” in the United States. In April 2019, Cityblock raised a $65M Series B funding round from Redpoint Ventures and 8VC, among others.

But Alphabet’s efforts to help broader populations of patients have not been uniformly well-received. For example, in November 2019, the Wall Street Journal reported that the medical records of millions of American patients had been made available to Google without their knowledge or the knowledge of their doctors. The records were released as part of a collaboration — code-named Project Nightingale — between Google Cloud and Ascension, a healthcare system made up of 2,600 medical facilities across 21 states.

Revelations about the depth and detail of the data that was given to Google — including patient names, dates of birth, hospitalization records, immunizations, and more — triggered the beginning of a federal Health and Human Services inquiry. It also prompted a Google whistleblower to write an editorial on their concerns about the project’s HIPAA compliance, which was published in The Guardian.

Cancer
Alphabet’s DeepMind and Verily have both run projects over the last several years dedicated to identifying and treating various types of cancers.

DeepMind announced in 2017 that it had successfully trained a set of algorithms to analyze images of breast tissue and identify tumors with about 92% accuracy.

In 2019, the Google Health team published a paper in Nature demonstrating the technology’s ability to spot breast cancer on a larger, more conclusive dataset — namely, mammograms from 91,000 women in the UK and the US.

During this study, DeepMind’s software reduced the incidence of false positives — an endemic problem in traditional mammogram screenings — by 5.7% among US participants. It also reduced the incidence of false negatives by 9.4% among US participants.

In July 2019, both Verily and GV took part in the $160M Series B round for Freenome, a San Francisco-based startup working on a early cancer detection screening system that uses machine learning to find markers of cancer in the blood before conventional tests can detect them.

Wearables
Verily is also exploring applications of wearables in the healthcare space. Its Study Watch, a biometric data-gathering wearable, received its first round of FDA approval in January 2019.

The device — specifically its electrocardiogram (EKG) feature — has been used in studies run by Verily and its partners since April 2017, and is now available more widely on a prescription basis.

About 10 months after the Study Watch was approved by the FDA, Google annouced its $2.1B acquisition of Fitbit. This gives Alphabet a consumer launchpad for its health monitoring technology and a potential vantage point from which to take on its main competitor in the wearables space: the Apple Watch.

Verily has made progress, but still lags behind Apple, which has been focused on incorporating healthcare features in the Apple Watch for years. In early 2020, Verily received additional FDA clearance for its “Study Watch With Irregular Pulse Monitor,” a version of its previous Study Watch that gives users the ability to monitor for arrhythmia. This feature made waves when it was first introduced into the Apple Watch in late 2018. Patents suggest Apple is working on even more advanced health features for the Watch, including noninvasive glucose monitoring, but it remains to be seen what Alphabet will work on next.

Pharmaceuticals
With its legacy systems, high margins, and middle men, the pharmaceutical industry presents numerous $100B+ opportunities for Alphabet.

Verily’s Project Baseline started in 2017 as an attempt to change the way that the $50B clinical trial space works.

Traditionally, a clinical trial requires a control group and an experimental group. The idea behind Project Baseline is to use the Verily Study Watch to crowdsource health data from 10,000 volunteer patients on a daily basis for five years, thus creating a kind of “universal control” that would eliminate the need for future control group testing.

Creating a universal control and monitoring people’s health on a daily basis, rather than just during a clinical study, means running trials can be more convenient for patients and analyzing the results can be easier for companies.

Four of the largest pharmaceutical companies in the world — Novartis, Otsuka, Sanofi, and Pfizer — joined the Project Baseline initiative in May 2019, with the aim of using its data to run more efficient clinical studies in areas like cancer and mental health.

While studies have been run on similar or larger scales in the past, no study has ever looked into this volume of people at this level of detail — including sleep, emotional health, heart rate, the genome, blood, tears, urine, and more — according to Stanford cancer researcher Sam Gambhir.

AI
While most of Alphabet’s public healthcare work has come through its two main healthcare-focused subsidiaries Calico and Verily, Alphabet AI subsidiary DeepMind also has healthcare as one of its primary fields of interest.

In 2019, DeepMind announced that its machine vision tool for diagnosing eye disease, during trials at Moorefield’s Eye Hospital in the UK, was able to correctly diagnose ailments like diabetic retinopathy as well as trained medical experts.

Alphabet’s advantage over other AI startups and corporates working on disrupting healthcare is not just rooted in technology — it is also rooted in scale and resources. For the trials at Moorefield’s Eye Hospital, a team of trained ophthalmologists and optometrists had to review more than 14,000 eye scans, with a senior medical expert on staff to mediate discrepancies.

In July 2019, the DeepMind Health team released a paper about the performance of a deep learning model developed to help predict the likelihood of a patient being diagnosed with an acute kidney injury.

In addition to Alphabet’s other work with Sanofi on diabetes and pharmaceuticals, the two companies are also working on a new AI-focused partnership as of 2019. The main goal of the partnership is to use AI to better understand specific “key” diseases, reveal the best treatment methods for patients, increase the personalization of healthcare treatment, and “better forecast sales and inform marketing and supply chain efforts.”

Life Extension
Calico, which is run by ex-Genentech CEO Arthur Levinson and focuses mostly on age-related diseases, has been using AI to analyze health data with the goal of tracking and extending healthy human lifespans since 2013.

A benefit of operating inside the Alphabet umbrella is that companies like Calico gain access to more sophisticated technology, making more effective laboratory techniques possible.

In one experimental example, Calico was able to use machine vision software to vastly expedite its study of yeast cell aging — important in understanding what causes yeast cells to degrade over time, and how that aging process works.

Having access to cutting-edge AI tools built by companies like DeepMind represents one of Alphabet’s biggest advantages over competitors like Amazon when it comes to building radical new solutions to healthcare problems.

Calico has worked closely on the biology of aging and anti-age-related therapies with the Broad Institute of MIT & Harvard since 2015, and with the pharmaceutical company AbbVie since 2014.

3. Next-gen computing
Alphabet is prioritizing quantum computing technology.

Google was one of the first massively successful tech companies built entirely on software, but that hasn’t stopped Alphabet from pursuing hardware interests aligned with its long-term ambitions.

Alphabet’s primary interest in this space revolves around the unique hardware demands of emerging technologies like artificial intelligence and quantum computing. Alphabet isn’t interested in competing with today’s OEMs to own the hardware market — it’s interested in building the hardware platform for these future technologies.

The company is looking to put its machine learning and engineering expertise to work building the next generation of high-performance computer chips and revolutionizing computing itself with quantum computing developments.

The main issue Alphabet is looking to address is that efficiently solving a certain class of problems with machine learning requires processing power that isn’t available through modern chips.

Since dedicated ML chips don’t need to run traditional software programs, designing one means starting more or less from a blank slate — so the playing field is relatively level.

While there are some physical limitations to contend with in developing chips for AI, Alphabet’s focus on quantum computing could help overcome them.

AI
When it comes to artificial intelligence, Alphabet is working on solving a two-pronged problem. AI applications require hardware with extremely high computation capacity, but they also need to optimize for efficient energy consumption.

In 2012, the image recognition system AlexNet, used by OpenAI, performed about one thousand trillion separate calculations every second. In 2016, Google’s DeepMind built a game-playing neural net called AlphaZero, which was able to become a grandmaster at the game Go within hours of simulated training. AlphaZero consumed about 300,000x the computing power of AlexNet.

Power cycle consumption with deep learning computations has, on average, doubled every 3.5 months for the last 7 years — and that trend has shown no signs of slowing.

New chips, on the other hand, are increasing at a pace of only about 3% every year, and even top-of-the-line hardware is inadequate for the needs of the most complex AI work today.

To address its own AI hardware needs, Google built a custom ASIC chip that it released in 2016 called the Tensor Processing Unit (TPU). The chip is specially designed to operate using neural net workloads, with its parallel computing capabilities today powering several of Google’s most critical applications: Translate, Assistant, AlphaGo, and Search.

Google’s Tensor Processing Units naturally integrate with Google’s widely-used (the fifth most popular open-source project overall on GitHub) machine learning framework TensorFlow. Engineers using TensorFlow within their own organizations can use TPUs to achieve much faster processing times on their own machine learning projects.

TensorFlow Lite, a new mobile version of TensorFlow, makes some of that same computing power available to developers working on mobile and other leaner projects, making it possible for smartphone-focused developers and others to bring technologies like rapid image classification and smart reply options to their applications.

Quantum Computing
While Google began working on TPUs largely to get around the diminishing returns of computing power associated with traditional chips, TPUs are running into issues scaling their processing power for the most complex deep learning applications.

One of Alphabet’s bets for the long-term future of AI involves developing an entirely new form of computing: quantum computing.

Google’s work on the research behind quantum computing dates back more than a decade, but today, Alphabet is focused on using the technology to break through the computing bottlenecks that are holding AI research back today.

Quantum computing is another space where Alphabet’s head start has allowed it to build a competitive advantage. Partnering with some of the most advanced organizations in the space, including D-Wave and NASA, Google has now claimed quantum supremacy, meaning it has reportedly achieved a crucial milestone in quantum computing ahead of all of its competition.

Google first started working on quantum computing back in 2006. In 2013, Google became the second firm to purchase a quantum computer from D-Wave, a Canada-based company that has been a major player in quantum hardware for decades.

A year later, Google hired a team of quantum computing experts to start their own lab at Google and focus on quantum computing research full-time.

This team built Google’s Quantum Artificial Intelligence Lab, which grew to become a collaboration between Google, NASA, and the Universities Space Research Association.

In September 2019, Martinis and his team announced the development of a new microchip called Sycamore. They estimated the 3 minute and 20 second long sample calculation run using this quantum chip would take somewhere around 10,000 years if you had 100,000 conventional computers.

The main area where Google’s work in quantum computing promises to be useful is artificial intelligence, where it’s technology promises to help address the need for increasingly high computation capacity. It will take time before Google is using quantum computing to better screen for disease or make advancements in materials science, but the advancements the company has made so far suggest this kind of breakthrough could occur in the future.

4. Transportation
Machine learning and AI are fueling Waymo’s dominance in self-driving tech, though commercialization is likely still far off.

Today, more than 40 automakers and tech heavyweights are working on building autonomous car technology. Progress towards working prototypes on the road has been slow, due to disasters like the fatal crash involving a self-driving Uber vehicle in 2018, and the inherent technical difficulty of building a self-driving car that can navigate complex environments safely.

Alphabet’s Waymo has advanced ahead of autonomous driving projects at Apple, Amazon, and Microsoft by a significant margin, and is broadly considered to be the industry leader in autonomous driving technology.

Alphabet has leveraged its expertise in machine learning and AI hardware, as well as its massive scale and war chest, to fuel Waymo’s progress.

The company’s cars have logged the most miles, both in the physical world and in simulations, and Waymo is currently operating the world’s only operational self-driving ride-hailing business. In 2019, the company started selling one of its 3D lidar models to third parties. In a statement, Waymo explained the move by citing its desire to scale up its production of sensors.

Waymo’s dominance on the technological front, however, has been undercut by the company’s more recent sluggish progress and failure to meet self-imposed deadlines on the commercialization of that technology.

Alphabet, however, in addition to building a ride-hailing business around Waymo, is also looking into applying the technology to the freight industry. The company is also expanding into other parts of the mobility space, including mapping, and through an investment in Lime’s electric scooter business, micromobility.

Waymo first showed off its self-driving hardware in February 2017, and began public demonstrations of its software installed in Chrysler Pacifica minivans a few months later. Not long afterward, Waymo partnered with ride-hailing company Lyft.

In October 2018, the company announced that Waymo cars had surpassed 10M real-life miles driven on the road. In the summer of 2019, the company announced that Waymo had surpassed 10B simulated miles — a critical component of the way Waymo cars are trained to understand obstacles and navigate the world correctly.

Amid speculation that Waymo’s partnership with Lyft would lead to a Google-supported ride-hailing service, Waymo launched its own commercial self-driving ride-hailing service in 2018. Today, that program — Waymo One — has more than 1,000 users in the Phoenix area. Among the volunteers using the service today, prices are comparable to regular rides from Uber or Lyft — about 60 cents a mile.

In July 2019, Waymo received permission to begin transporting human passengers in self-driving cars in California as well.

Waymo is unique among the technology companies and corporates working on autonomous driving technology mainly due to this public-facing orientation: its goal is not to produce its own autonomous vehicles, but to partner with an existing automaker or automakers to sell a safer, more efficient, and potentially cheaper ride-hailing service to the general public.

The potential prize here is massive. Despite the difficulties involved in building and bringing autonomous vehicles to the road, Goldman Sachs estimates that the popularization of autonomous car technology will take the ride-hailing industry from $5B to $235B by the year 2030.

However, a truly autonomous car experience for all is still on the distant horizon for Waymo. California still requires the company to place a safety driver in the driver’s seat when testing Waymo vehicles. No driver is now necessary for tests in Arizona, but as Waymo itself has acknowledged, Arizona represents many of the ideal testing conditions for the company’s autonomous vehicles which are unlikely to be found elsewhere around the world.

Reports from early trials of the technology have not been uniformly positive, with complaints emerging about how Waymo cars dealt with certain kinds of complex road situations, leading ultimately to Waymo electing to put safety drivers back into vehicles that had gone driverless. Waymo vehicles have also been involved in “dozens” of crashes according to the company’s own reporting, though no significant injuries have been reported.

Despite these issues, Waymo’s cars have the lowest instances of disengagement — when the algorithm running the car gives control back to the human monitor — per miles driven among all the different autonomous vehicle companies testing in California. In 2018, Waymo’s human safety drivers in California trials were only required to take back control of a vehicle once in every 11,000 miles. As far as accidents, Waymo was also superior, with just three collisions over more than 350,000 miles driven (GM had 22 over 132,000 miles).

But while the company’s progress has eclipsed that of its main competitors in the autonomous driving space, its own stated timelines have proven impossible to meet. In 2012, Google co-founder Sergey Brin predicted consumers would be riding in self-driving vehicles within five years.

In 2018, Waymo CEO John Krafcik admitted to the attendants of a National Governors Association fireside chat that the “time period will be longer than you think” for autonomous vehicles to be on US roads.

Autonomous driving, however, isn’t the only mobility solution that Alphabet has placed bets on. Both Google Ventures and Alphabet took part in scooter-sharing unicorn Lime’s $335M Series C round in July 2018, broadly anticipating a world where cars are less important as a form of personal transportation.

In August 2019, Alphabet announced a new integration with Lime that would enable users of Google Maps to see nearby available Lime scooters in more than 100 cities where the on-demand scooters are available. For more than 50% of all smartphone users in the US, Google Maps is the default navigation tool.

Alphabet’s integration with Lime is a useful example of how that popularity could give Alphabet a viable means towards owning the entire distribution layer of the mobility space. Google already dominates navigation in general. Today, Google Maps helps identify locations, give turn-by-turn directions, and give context to different destinations (like restaurant reviews and open hours).

If Maps can extend to actually helping move people from A to B, Alphabet can be the mediator between users, ride-hailing, and other mobility services.

5. Energy
Alphabet subsidiaries are looking to renewables to make the conglomerate more energy efficient and extend solutions to the public.

Alphabet is currently working on solving its intractable energy consumption challenges — largely driven by Google’s data centers — and then taking what it learns to offer household solutions that could disrupt the renewable energy industry in multiple ways.

Alphabet includes a few energy-focused companies under its umbrella, including Dandelion which offers geothermal energy services. For a while, Alphabet was also developing a new type of kite-based wind energy under its subsidiary Makani, but it pulled support for the business earlier this year.

Elsewhere, DeepMind is applying machine learning to the problem, looking for ways to use AI to bring energy usage and costs down.

Alphabet’s preoccupation with energy stems partly from its own energy consumption. Google said that its worldwide operations consumed a total of 5.7 terawatt hours of electricity in 2015: about as much as all of San Francisco county in the same year, according to the California Energy Commission. By 2018, Google’s energy consumption had nearly doubled to more than 10 terawatt hours, according to Statista.

For years, Google has also been one of the biggest consumers of renewable energy in the world. In 2016, no company in North America purchased more renewable energy, according to Bloomberg New Energy Finance. Finding cheaper, more efficient sources of renewable energy represents one of the best ways for Google to lock in energy prices for the foreseeable future and protect its business from shocks, like rises in the price of oil. That’s why the company is also focused on building out its own intelligent, renewable energy infrastructure. Using machine learning, Alphabet can consume electricity more efficiently. In 2016, the DeepMind team announced they had reduced the electricity bill for cooling Google’s data centers by 40%. Today, the program it built to monitor energy usage and make recommendations — which humans at Google were ultimately responsible for implementing — runs autonomously, with data center operators only supervising the recommendations that the AI-driven system implements.

Alphabet has also invested in offering energy from renewable sources like geothermal and wind. With companies Dandelion Energy, Alphabet is bringing some of that research to the public.

Dandelion Energy, for example, promises homeowners a cheaper home energy solution through the use of geothermal energy. The company raised a $16M Series A led by GV and Comcast Ventures in 2019, and raised a further $12M earlier this year to bring its total disclosed funding to $35M.

Geothermal energy, which extracts heat from the ground, can be both cheaper and cleaner than oil and gas. Additionally, Dandelion’s system can cost about half as much as traditional geothermal installations (which can cost up to $40,000) — driven partly by using more standardized units and a less expensive drilling technique for laying the associated infrastructure in the ground.

6. Smart cities
Sidewalk Labs aims to achieve dominance in the space by offering a comprehensive smart city package.

Alphabet’s smart city startup, Sidewalk Labs, began in 2015 with the mandate to come up with new kinds of technologies to improve urban life. In the 5 years since, Sidewalk Labs has become a kind of one-stop shop vendor for smart city technology.

Sidewalk Labs is positioning itself to be the dominant smart city vendor in two big ways: by leveraging Alphabet’s resources to make it financially feasible for towns and cities to work with the company, and by using other Alphabet companies to make its smart city offering more comprehensive.

Other Alphabet companies contributing to Sidewalk Labs’ work include Waymo (autonomous cars) and DeepMind (machine learning), as well as smaller internal startups and spin-offs including Cityblock (personalized health), Coord (mobility), Intersection (Wi-Fi), and Replica (urban planning).

While various startups offer some of these services, Sidewalk Labs can offer cities the full range. In June 2019, the company released a plan showing how it would put those services to work on its first full-scale project along a stretch of abandoned industrial space on Toronto’s Lake Ontario shoreline.

Sidewalk Labs plans to invest in a series of new mixed-use buildings, a renewable energy grid, and underground delivery tunnels, along with an extension to help bring the area onto the city’s light rail infrastructure.

A planned renewable grid will use energy recycled from buildings, sewers, lakes, and the earth to provide heating and cooling to the rest of the neighborhood.

The Quayside development will have public Wi-Fi, as well as a variety of sensors collecting information about traffic patterns in the area, energy consumption, and building activity.

On the business side, Sidewalk has announced that it will bear a “disproportionate share of the cost of upfront innovation” on the Quayside development, in exchange for later compensation if and when the system meets certain performance requirements.

With Alphabet’s ability to share in the risk of getting the smart city up and running, as well as its ability to be a “one-stop shop” for smart city technology, Alphabet and Sidewalk Labs have built a powerful set of competitive advantages in the smart city space. What that dominance means depends largely on how the Quayside experiment goes and how willing other cities and towns prove to be when it comes to giving a company like Alphabet a key role in urban development.

7. Travel
Google is leveraging its search capabilities in an attempt to disrupt online travel agencies.

Some of the most disruptive technology companies of the early 2000s were the online travel agencies (OTAs). Companies like Expedia and Priceline allowed people to bypass traditional travel agents, searching for and booking their own flights, hotels, cruises and rental cars more conveniently and more cheaply than before.

The supremacy of OTAs is largely built on the ability to search for your own travel accommodations; today, that’s being challenged by the company that pioneered internet search.

Alphabet is angling to disrupt OTAs by building a better search experience, asserting its dominance of the search space in general, and getting its results in front of consumers before companies like Expedia and Priceline (now Bookings Holdings) can.

In 2011, Google completed its acquisition of ITA, the developer of the Matrix search engine and used the acquisition to power its Google Flight Search. The Matrix search engine, built by scientists at MIT in the 1990s, was the first commercial software that allowed travel agencies and others to find and compare low airline fares for a given origin and destination.

Over the years, Google’s flight search competed with traditional OTAs’ search engines with various new features, including the ability to automatically identify a consumer’s home airport and customizable search options like looking for flights within a certain duration range. More recently, Google launched a feature that aims to better integrate browsing for flights and hotels. By saving your destination and date information from your flight search, Google can autofill that information into your hotel search, saving time and creating a more personalized experience for the user.

Google doesn’t yet sell its own flights and hotels — it aggregates listings from all the major OTAs and presents them on its search engine results pages and on its Google Travel tool pages.

The major OTAs spend significant sums every year marketing and running search ads through Google, which means that Google must toe a cautious line to avoid upsetting its partners.

But through its dominance in search, Google has a powerful vantage point from which to attack the highly valuable online travel bookings market — with its data superiority, it is aiming to deliver a product that is more personalized and relevant than that of the major online travel agencies (OTAs).

Google can use its massive amount of contextual data and infrastructure to deliver more personalized search results to potential travelers, but also to deliver lower and more predictable prices.

In 2019, in what may have been Google’s most aggressive move against the OTAs yet, the company launched a program that offered certain locked-in prices on holiday travel to Google Flights users. Using historical data on what specific flight routes cost, Google offered consumers the ability to get a refund if the prices on the pre-booked flights decreased after they booked them.

This move was also an indication of another big advantage that Google might have in the travel market: the ability to more accurately predict demand. Being able to project which airline routes and what destinations will be busy, and to what degree, could be a hugely valuable service for Google to offer to hotels, airlines, rental car services, cruise lines, and the rest of the travel industry.

In 2019, Google launched another feature that could divert consumers from OTAs like Expedia and Booking: smart hotel search.

Users can input their travel dates and see a list of all the available hotels that have capacity for that period, including highlighted “great deals” (rooms that are normally more expensive for that historical date). Each hotel’s page shows the different classes of room available, reviews sourced from Google users, other things to do in the area, and booking options.

Lastly, when booking a hotel room, the transaction takes place on Google itself, rather than forcing customers out to other booking sites to check out.

Google’s search platform also gives it an opportunity to push its airline and hotel results in front of users even sooner in the funnel. Currently, when it detects a query regarding a specific flight route, Google places a personalized Flights search box inside the search results page.

Google could one day sell flights, hotels, and other kinds of trips and accommodations to the millions of people who use Google every day right through its search results — and there’s evidence to suggest that many consumers would be on board.

Around 60% of US travelers use OTAs to plan trips, according to a 2018 MMGY study on travel habits. However, the study also found that Google ranked No.2 next to Expedia for “one-stop shops” that prospective travelers consider online. Google also ranked as the most popular method overall for researching flights and prices before traveling.

8. Gaming
Alphabet wants to change the way people game with Stadia, but the platform hasn’t lived up to its hype thus far.

From bets on VR and streaming to building its own video game console, few tech companies are as intent on bending the future of entertainment towards its own vision as Alphabet.

With Stadia, the company is planning one of its biggest entertainment projects yet: leveraging Google’s infrastructural might and its acquisition of YouTube to reinvent the way people game.

Alphabet has been seriously investing in gaming for a while. In 2015, it launched YouTube Gaming, a version of YouTube (and a Twitch competitor) specifically catered to video game content creators and their viewers.

YouTube currently has hundreds of thousands quarterly active gaming streamers. Over the 12 months leading up to September 2018, more than 50B hours of content in total were filmed. Those numbers make YouTube the second-most popular game streaming service behind Twitch.

Alphabet has invested in or acquired a host of gaming companies, including:

• Owlchemy Labs
• Agawi
• Niantic
• Green Throttle Games
• Beyond Games
• Bionic Panda

Of the FAMGA companies (Facebook, Amazon, Microsoft, Google, and Apple), only Microsoft has invested in or acquired a comparable number of gaming startups, with Amazon, Facebook, and Apple investing in far fewer. With its newest and biggest gaming project, Stadia, Google aims to leverage its scale, machine learning expertise, and technological infrastructure to build a new kind of console-less gaming experience.

Users pay for a subscription to Stadia’s online gaming platform and can play using a special Stadia controller that can link up with devices like smartphones or computers. Stadia’s model doesn’t require customers to buy a dedicated games console or physical games for their home.

Perhaps most importantly, all of Stadia’s processing takes place on Google’s cloud infrastructure — meaning, in theory, that powerful remotely-situated computers can beam high-quality graphics and CPU-intensive gaming experiences to users in their homes. However, cloud-based gaming has historically suffered from issues with delays in responding to gamers’ actions. Stadia aims to leverage Google’s extensive infrastructure to eliminate detectable latency and make cloud-based gaming as seamless as gaming at home. Google is also planning a free version of its Stadia platform, which it hopes will attract a wider audience.

One of Stadia’s features that could be compelling for many in the gaming community is the touted ability to play live with popular video game streamers. This type of feature could help drive growth for Stadia as the game streaming and e-sports spaces continue to grow.

However, Stadia’s efforts thus far have generally not lived up to the hype. The platform has experienced issues with pre-ordering, lags in gameplay, and lackluster reviews for its games library. Whether Stadia succeeds in changing the business model behind gaming will ultimately rely on whether it can deliver on its game-changing features before the platform expires from lack of interest.

9. Media
YouTube TV is battling with Amazon Prime and Hulu to win over cord-cutters.

Since Google acquired YouTube in 2006, the streaming site has become the most popular destination for online video. Around 5B videos are reportedly watched on the platform every day.

With its live TV and DVR service YouTube TV, Alphabet is trying to leverage YouTube’s dominance in general streaming media and the power of its brand to compete with traditional broadcast TV.

YouTube TV offers live TV from 70+ channels for $50 a month. The average cost of an extended cable TV service is over $70 per month, according to the FCC. YouTube TV’s model also offers unlimited DVR with no fees, as well as not requiring a fee for renting a box or for canceling.

Another key part of YouTube TV’s value proposition is the platform’s ability to use past viewing data to offer personalized recommendations about what to watch next.

For advertisers, YouTube TV could potentially offer a place to reach consumers in a more targeted way than traditional broadcast television. For example, the personalization data that YouTube collects could be used to give advertisers a more complex audience model and help them better target their ads.

One big problem YouTube has faced is concern over its reluctance to censor allegedly damaging or harmful content uploaded to the website. Alphabet’s Jigsaw, an internal unit working on identifying emerging societal threats on the internet, is trying to solve that problem.

One output of the company’s efforts is the Perspective API, which was developed by Jigsaw and Google’s Counter Abuse Technology team. The software uses machine learning to identify instances of abuse online, gives those instances a “score” depending on their severity, and makes that information accessible to human content moderators to help them make a decision about what action to take.

Since being made available in 2017, organizations like The New York Times, Wikipedia, and Reddit have tested Jigsaw’s ability to combat toxic comments sections.

Despite these issues, YouTube does have one big advantage over other television streaming services like Amazon Prime Video and Hulu: it has better penetration into regular consumers’ everyday lives.

Over 20% of video streaming during Q4 2019 took place on YouTube, making YouTube second only to Netflix (31%) according to Nielsen data. The report found YouTube to be far ahead of other streaming rivals like Hulu (12%) and Amazon (8%).

As the competition heats up between these four streaming services and emerging rival options from companies like Comcast, YouTube’s ability to stay top-of-mind could be vital as the company looks to expand its media ambitions.

10. Banking
Google is working to become the gatekeeper between its users and bank services.

Despite Google’s recent announcement that it would be launching checking accounts in 2020, its goal in banking appears to be less about becoming a bank and more about acting as the mediator between its millions of users and the services offered by traditional banks.

By becoming the conduit through which banks can offer services, Google stands ready to collect a vast amount of valuable data on what services are useful and why. It could then use that data to aggregate demand, similar to the way it approaches hotels and travel searching.

Google’s biggest fintech product so far has been Google Pay — formerly Android Pay — which merged with Google Wallet in January 2018. Google Pay was active in 28 countries by the end of 2018, and processed 1B transactions in its first year. In India, the product has 45M+ monthly active users.

India has been a particularly rich region for Google’s fintech ambitions, with the company launching the mobile payments platform Tez (later rebranded under Google Pay) in the country in September 2017.

Now, in 2020, the company aims to make its biggest foray into banking yet. Google says it plans to begin offering consumer checking accounts in 2020 in partnership with Citi and Stanford Federal Credit Union.

The checking accounts — under a project codenamed “Cache” — will be run by Citi and branded with the bank’s imagery. In keeping with Google’s desire to mediate the relationship with the end-user, users will ultimately access their accounts using Google Pay.

Offering checking accounts could be a way for Google to help its Pay service compete with mobile payment services from companies like Samsung and Apple, which have seen higher usage so far. It could also help Google collect valuable data on payment patterns and other types of consumer financial information, possibly helping it market its advertising services.

The move is not surprising from Citi’s point of view either, considering the partnerships that it has formed with companies like Paytm (India) and Grab (Southeast Asia).

For Citi, Google presents an opportunity to increase customer acquisition by working with one of the main platforms that consumers use to access the internet.

What’s next for Alphabet
Google’s search and advertising businesses have laid the foundation for Alphabet to pursue a broad range of projects, experiment with moonshots, and strategize how it might disrupt staid industries like healthcare, banking, and transportation.

However, this report is by no means representative of everything going on inside Alphabet today. The company has plenty of other projects at various stages that could one day prove disruptive.

For example, Project Loon is Alphabet’s play at expanding high-speed internet to remote and rural areas around the world. At the same time, the company’s umbrella includes a cybersecurity spinoff called Chronicle and a renewables project called Malta, among many others.

One of the most promising avenues of future disruption for Alphabet may be retail. The company partnered with a collection of major retailers in 2017, allowing consumers to use a Google Express front-end to shop from stores like Costco, Walmart, and Target.

Shopping Actions, which launched in 2018, bridged the gap between the mobile, desktop, and voice versions of the company’s shopping platform, which the company says led to as much as 30% higher cart totals in early tests.

Today, Google’s shopping service still can’t beat Amazon on logistics. But the success of one of Alphabet’s other in-progress experiments — the autonomous delivery drone service Wing — could one day be the last-mile solution that connects Google’s online storefront directly to customers and reconfigures the entire retail landscape in the process.

This article first appeared here.