Rick Osterloh has been on the job as the senior vice president of hardware at Google for just over 17 months now. In that time, he's had to repeatedly answer the same questions from reporters like me: just how serious is Google about making its own hardware? Is it a hobby or is it going to genuinely affect Google's financial bottom line? Is the company sure it won't repeat the same mistakes it made with its ill-fated Motorola acquisition and subsequent sale years ago?
He's heard it all before: Osterloh was actually president of Motorola for a time under Google. In an hour-long interview, his answers to those questions haven't changed since last April. They might not stop us from asking them over and over, but the consistency of the answers is important. And if there was any doubt about Google's ambitions in hardware, the company definitively put it to rest by acquiring 2,000 or so phone engineers from HTC last month, along with some IP and equipment.
Last October, Google wanted to show it was serious about hardware with a wave of hardware announcements. But Osterloh had just started a few months earlier, so he acted as more of a master of ceremonies for products than the original architect of them.
This year, everything Google is announcing was created under his watch. It’s our first real look at Osterloh's vision for what Google hardware should be. His vision includes no fewer than eight products, two of which are in completely new product categories for the company.
Last year was a coming-out party for Google hardware. This year is something different. It's a statement that Google is very serious about turning hardware into a real business on a massive scale — just maybe not this year.
Google's HTC deal wasn't a straight acquisition of the company. Instead it simply hired a ton of HTC engineers who will switch their badges to say Google but won't pack up and change offices. For Google, it wasn't about HTC's VR efforts, or even about its manufacturing capabilities. Osterloh says that Google will still choose the factories that are best for its future products.
"The deal was fundamentally to try to build our capabilities so that we could scale our business faster and strengthen our smartphone position," Osterloh says. Google is already very familiar with the team it's acquired. They built the original Android phone, the G1. They built the Nexus One. They were deeply involved in engineering last year's Pixel.
"The hardware business requires a lot of people to scale and to do big things," Osterloh says. "Especially smartphones, which are super complicated. ... And then the second big thing is that we needed a more expansive Asia operation."
Google has been down a similar road before: it acquired Motorola and kept it at arm's length before ultimately spinning it out to Lenovo (in part because it needed to placate other Android manufacturers like Samsung). But Osterloh doesn’t think history will repeat itself. "They are very, very different deals," he says. "I wasn't here when the decision was made to acquire Motorola. I came afterwards. But it was pretty clear a primary driver of that was the situation with [smartphone patents and intellectual property] at the time."
"This strategic deal is very different. We know exactly what we need."
Those patents didn't turn out to be anywhere near as valuable as either Google or Motorola thought, but to Osterloh the real issue is that Google simply didn't know what its goals were with Motorola, which had lots of different businesses besides phones. "This strategic deal is very different. We know exactly what we need," Osterloh says. "We want deeper engineering capabilities, and we happened to know this team very well at HTC."
Fair enough. The HTC acquisition is tiny compared to what Google tackled with Motorola. Nevertheless, it does underline how serious Google is about making its own phones — and perhaps Samsung will eventually chafe under that competition.
That's not really Osterloh's problem, since he’s in charge of Google’s hardware efforts. "That's probably a proper question for Hiroshi [Lockheimer, the head of Android]," he quips. He also points out that Google is doing more than just smartphones: it's making home speakers, headphones, a new kind of camera, and a new high-end laptop.
One area that Google's hardware division is steering clear of is Nest's domain, and Osterloh isn't worried about it. "The company has made the decision that they're independent businesses. … Our focus has been trying to get people connected to Google's services in the home." Nest and Google coordinate sales and supplier relationships, but when I ask if Nest is keeping his team from making hardware products they’d otherwise like to make, Osterloh's answer is: "No, not to date."
Google's smartphone efforts are frankly not likely to sell in numbers that will threaten Samsung anytime soon. "We're focused on the high end in a few markets," Osterloh says. "So while there might be some overlap with the ecosystem, there are 2 billion Android phones out there. And our business is obviously much more narrow than that."
Which, again, brings up one of those repeated questions: the seriousness of Google's hardware ambitions. Osterloh still thinks of the Pixel as "trying to deliver a benchmark experience," calling it "a very common thread from Nexus." But where the Nexus was never meant to sell in any real volume, the Pixel seems like it is.
While Osterloh expects the Pixel to "become big, meaningful business for the company over time," right now his benchmark isn't sales, it's "consumer satisfaction and user experience.” So I ask: what about five years out? "We don't want it to be a niche thing," Osterloh says. "We hope to be selling products in high volumes in five years."
For the time being, Google is focused on nailing high-end technologies. It wants to sell the phone with the best camera and the best OLED screens. Osterloh contends that Pixel phones will not only be competitive on quality, but even better than the competition in certain metrics.
More than anything, though, he believes Google's mission in hardware is to take advantage of Google's core strengths and apply them to computer products. It's common now for people at the company to say the same thing we've heard Apple say over and over: that the integration of hardware and software is key to making great products. But Osterloh and everybody else at Google are adding a new twist: “hardware and software and AI.”
When you talk about integrating software and hardware, the elephant in the room is silicon. Apple designs its own processors, and it is able to produce chips that, by many accounts, are years ahead of what Qualcomm can provide to Android manufacturers like Google.
"Do we have an advantage in hardware now? No," Osterloh says. But he does so in an interesting context: processors aren't actually getting faster at the same rate they used to, and they're also not getting more power efficient as they get smaller. So focusing explicitly on the CPU is a mistake.
"The Moore's law performance curve has really slowed down," Osterloh says. "So there's still performance gains, like raw performance gains, but it is definitely not doubling every 18 to 24 months on technology anymore." He also points to something called Dennard scaling, which usually means that processors get more power efficient as they get smaller. That's another "law" that no longer applies, he argues. "You can't just put in a new chip that's on the cutting edge in a laptop or a phone and expect that the thing is going to perform at the same level of power performance.”
"Core hardware technology is slowing down, changing, and the laws of the past are no longer are valid.”
Osterloh says that Google will attack some of these problems by working with Qualcomm. "What we can effectively do is work with them closely to move that in the direction that's important," he says. He continues, saying, "We are not developing chips ourselves."
Instead, the key to Google's performance strategy is AI. If the curve of processor performance is flattening, Osterloh’s plan is to differentiate Google's products by integrating machine learning and AI.
The entire company is directed toward improvements in AI and machine learning, and those improvements will come to Google's hardware in interesting ways, he argues. "Because core hardware technology is slowing down, changing, and the laws of the past are no longer are valid, we think that provides a future ability for us to develop hardware that can solve the problems we want to solve, and that can really move computing forward."
That advantage, Osterloh believes, is unique to Google. "I think the corporate-wide advantage is that we know where our software direction is going." As Google extends its lead in AI and machine learning, Osterloh will take that work and put it directly into Google's hardware. Sometimes — as with Google Assistant — it will talk to Google's servers. Sometimes it will just be integrated directly into the hardware with custom, local chips, as with the new Google Clips camera.
It's going to take some time before we can tell if Google's machine learning bet will play out. And who knows, it may eventually decide it needs to compete directly with Apple on chips. But what we can judge right now is the design of Google's hardware products. They focus much more on simple pragmatism than flashy design.
"You know, I don't think Google has ever been a 'flash' company, right?" Osterloh says. I remind him about the people parachuting out of a blimp once upon a time. "Let me restate that, then," he laughs. "The product design is attempting to reflect the brand. I think the company has always been about trying to solve real user problems and having the technology recede in the background."
Google's not trying to make edge-to-edge screens with cutouts, or even curved screens. ("Look, we don't think there's use in that.") Instead, Osterloh says, Google "wanted to try to keep it simple and nice and clean ... We're trying to be reminiscent of Google's main property: the big search box."
With Google, it always seems to come back to that search box. The company might be unveiling a large family of products. It might have hired a couple thousand new phone engineers. It might even be serious about creating products that are directly competitive with Apple, Samsung, Amazon, Bose, Sonos, and many others. It is doing all of those things. For the time being, however, it's not doing them at a massive scale — yet.
But after talking to Osterloh, it's very clear that Google wants to.