- Conversations are harder venues for ads than computer screens
- Answers from AI assistant blur line between free, paid results
The smartphone boom upended Google’s advertising profit engine and it took years for the Internet giant to adjust to the new mobile world. The next wave of computing will be an even bigger challenge.
At Google’s I/O developer conference this week near its Silicon Valley headquarters, the company unveiled new technology that will rely less and less on physical devices with screens to deliver information and services to consumers. Google hopes these advances will capture the human attention its business depends upon, and then it can figure out how to make money later, one executive said.
Google Home will sit in living rooms sucking in voice-based queries and delivering verbal answers from an artificially intelligent "Google assistant." A jacket due out in 2017 with computing built into the yarn will respond to touches and swipes and give voice directions through ear phones connected to people’s phones. Project Soli will let people control computers all around them through finger gestures in air. Even Google’s newest smartphone design -- Ara -- is less of a device and more a frame that phone components will slot in and out of.
Google Chief Executive Sundar Pichai describes this as "ambient" computing that will surround consumers whereever they go. He sees it as one of the most important changes that will envelope the technology industry over the next decade. What is less clear is how Google will make money from this.
Google’s main AdWords search advertising business generates more than $40 billion a year in highly profitable revenue from text-based ads that pop up alongside search results shown on personal computer and smartphone screens. When results are delivered by the company’s AI-powered assistant, the existing ad business won’t work.
"Where are the ‘good old’ AdWords ads if they succeed? They are nowhere. They are dead," said Carlos Kirjner, an analyst at Sanford C. Bernstein. "If you thought mobile was a big transition for Google, buckle up."
When searches started happening on phones, Google had less screen space to show ads alongside its free, or organic, listings. This triggered soul-searching, questions from investors, a period of stock price stagnation and finally a thorough re-organization of its search results and ads. Paid results now more often appear at the top of mobile search results and some ads have become more like transactions where Google takes a cut of a booking fee, rather than a payment from a marketer for a click.
Ambient computing will force Google to embrace more new ways of making money, Kirjner said.
"Compared to this fundamental change in business model, probably requiring that Google captures value from transactions and not from ads, mobile was just a minor, user interface tweak," he wrote in a note to investors.
Google will build its AI-based personal assistant first and focus on ways to make money from this later, the company’s new chief of search and AI, John Giannandrea, said during his first public remarks to a large developer audience at I/O Friday.
"We’re really focused right now on building something that users like to use," he said, when asked about monetization of the product. "Google has a long history of trying to build things that people find useful and if they find them useful and they use it at scale then we’ll figure out a way to kind of support that."
In a world of ambient computing, users could interact with fewer ads, but if Google has its way, those marketing messages will be better targeted and therefore more valuable to advertisers, according to a person familiar with the matter.
Despite the early stage of the work, AI systems will blur the lines between Google’s free and paid results. Even with the company’s new messaging service Allo, which will have the Google AI assistant delivering answers and making bookings in text streams, it’s unclear how ads or other paid-for information and services will be presented.
"What happens when Google’s assistant is asked to perform a specific transaction?" Kirjner wrote. If the system successfully picks the best counterpart for a consumer’s transaction, will it proceed even if that provider is not paying for the access? Or will the assistant offer a lesser counterpart that is paying for access without the user’s knowledge? "We are not smart enough to figure this one out," Kirjner said.
“We do as a general rule clearly identify what is sponsored and what is organic,” Giannandrea said. Products like Google Home haven’t launched yet and Google is still working on how these services will work, he added.
That decision is partially because of the early stage of Google’s AI efforts. Despite decades of research and a larger focus in the last four years, the company is still grappling with "big unsolved" problems of computer science, like understanding language and dialog, Giannandrea said. That’s a long way from designing new ad formats and alternative business models.
One way to make money now is to deliver some of Google’s AI technology as services to other companies over the Internet. Earlier this year, Google did just that, offering access to things like vision recognition through its cloud computing business. It’s a high-volume, low-cost business with the company charging cents to fractions of a cent per AI-based prediction.
"It’s very important to us that our natural language APIs become part of our cloud platform and people experiment with it and help," Giannandrea said. Google plans more of these AI cloud services, he added.
Still, Google will need more than cloud computing subscriptions to replace revenue and profits from text-based search ads. As the very concept of the "device" fades away, the company must find new ways to make money from a business that is becoming increasingly invisible.