Google is an advertising company.
In spite of its dominance in mobile operating systems, productivity tools like Gmail, and forays into media with subscription services such as Google Play Music and YouTube TV, the company still makes nearly 90% of its money from advertisements and its advertising platform. Advertising is also the primary driver of revenue for its holding company, Alphabet, which has a large portfolio of other businesses including AI research lab DeepMind, autonomous driving company Waymo, life sciences company Verily, and investment groups GV and CapitalG.
Many of those other companies are long-term bets that aren’t expected to pay off in the near term—or maybe ever. There are grandiose objectives like DeepMind’s quest to “solve intelligence,” or Calico’s mission to cure death. But there are slightly more grounded ideas that are finally poised to be real contributors to Alphabet’s bottom line: Waymo and Google Cloud, its server space and AI services that can be rented based on usage.
Google doesn’t make it easy for investors to see exactly how much its cloud business actually makes, but earlier this year Cloud CEO Diane Greene mentioned in an interview that the service had crossed $1 billion in revenue per quarter.
Another barometer of the cloud’s success is Google’s “other revenue” line item, which the company uses to report revenue from everything that’s not advertising, such as services and hardware. Revenue from that segment is the company’s fastest growing, bringing in $4.4 billion in latest quarter, up 37% year over year, the company reported last week. Google Cloud also announced it added large new customers including Target, Carrefour, and PwC.
This revenue is dangerously close to surpassing Google’s ad network, its second-largest revenue stream, which brought in $4.8 billion last quarter on 14% year over year growth.
But Google also faces huge competition in the cloud space. Microsoft and Amazon have both reported substantially bigger revenue from the market, which is estimated to be more than $180 billion this year, according to Gartner. When asked about this during a conference call with investors earlier in the week, Google CEO Sundar Pichai shrugged off the implication that there’s not enough business to go around.
“It feels far from a zero-sum game,” Pichai said. “Businesses are going to embrace multiple clouds over time too. So, I think not only is this early, but I think it is going to transform. And there is a lot of opportunity here.”
One advantage Google holds over its competition is its deep bench of AI researchers, whose work is constantly being translated into cloud products. At Google Cloud Next, its annual developer conference held last week, the company announced that it was readying a product for call centers that would allow human-sounding “virtual agents” to handle simple calls, handing off more complex questions to humans. It also announced a suite of self-improving AI algorithms called AutoML; Quartz reported on the company’s research into this field before it was turned into a product.
Buried in the conference’s announcements was a new hardware play by Google—a smaller, mobile version of its Tensor Processing Unit used in Google datacenters. Google hasn’t released specifications on the mobile chip yet, but the company claims that the datacenter version offers up to 30x speedups for machine learning while reducing power consumption.
Getting a chip like this into Google’s hardware products, such as its Pixel phones or Google Home devices, could mean the devices would be able to use AI to understand and answer questions without contacting Google servers. That capability would be a boon for privacy—every query wouldn’t be stored on Google servers—but also improve the speed of the devices. Elsewhere, the chip could sustain Google’s dominance as the go-to framework for building artificial intelligence algorithms, extending AI into smart sensors for enterprise use.
Last week also gave us a clearer look at how Alphabet’s autonomous driving company, Waymo, could start generating revenue. The company announced an expansion of its pilot program in Arizona, and told Quartz that five partners will pay Waymo to integrate autonomous vehicles into their businesses. These partners aren’t local businesses, either: Walmart, Marriott-owned Element Hotels, Avis, AutoNation, and DDR Corp. It’s easy to see the value of Waymo’s cars to these businesses, as they’d provide a cheap ride to a store, hotel or rental car office. On the other side, Waymo gets to lock up some long term, potentially big dollar partnerships to support its business. Neither Waymo or the partners disclosed the amount of money changing hands in the deal, though the partnerships are expected to scale as Waymo enters more cities.
Alphabet is also making sizable investments into Waymo, including adding up to 62,500 autonomous minivans to its fleet from Fiat Chrysler. Back of the napkin math suggests that the minivans, which retail for about $40,000 each, could easily be a billion-dollar deal as they’re delivered over the next few years.
As a part of its “early rider” program, Waymo isn’t currently charging customers for the rides—but once the ride-hailing service launches in earnest, each of those 62,000 vehicles would be making money for the company alongside corporate partnerships. Waymo’s ride-hailing goals can be seen in its permit application; Quartz previously reported that the company had filed the paperwork to start a commercial venture in Arizona this year.
But besides “later this year” we still don’t have a concrete timeline for Waymo’s commercial operation at scale. And even though Waymo and Google Cloud are going to add to Google’s bottom line, it’s a long road to the more than $20 billion the company generated from ads last quarter.