Google first introduced smartphones of its own design nearly seven years ago, and hardly anyone has bought one. On Tuesday, it unveiled a revamped lineup, branded with Google's Pixel name.
The company's Android software already runs the vast majority of smartphones sold in the world, but Google wants to hedge its bets by having control over every feature and function in a selection of Android phones.
Google's aims only can pan out if it sells a large number of Pixel phones -- which it won't. And its strategic goal isn't a good enough reason for Google-branded smartphones to exist at all.
So why do it?
Ensure the use of Google apps
Android powers about 88 percent of the smartphones sold globally, according to IDC, but Google is losing control of the sprawl of manufacturers such as Samsung, Huawei and Oppo.
In China and India, local companies sell smartphones with the Android operating system but with local versions of app stores in place of Google Play. In China, people carrying Android smartphones may never use Google search, YouTube, or Google Maps at all -- undermining Google's strategy behind Android.
That's a big risk. If the company's internet options aren't front and center on people's phones, they may use Google less, and spend more time on Facebook, WhatsApp and WeChat. That in turn crimps Google's opportunities to turn consumer attention into ad sales. A home-grown line of phones has the benefit of ensuring Google's apps have prominent placement.
But the challenges remain unless Google can sell at least tens of millions of phones -- and it can't because of inherent business conflicts. If Pixels sell in huge numbers, Google risks annoying essential partners -- the companies like Samsung that make Android phones, plus the legions of mobile phone carriers likewise needed to push Android.
Showcase the best of Android and spur hardware innovation
There is a useful analogy here between Google and Microsoft. The latter company for decades designed the operating software for personal computers, and left the design and selling of PC hardware to partners such as Dell.
That arrangement changed in 2012 when Microsoft started selling its own line of Surface computers. Microsoft in part was trying to show some fresh ideas for PC designs, which had grown a bit dull and rote. It was a strategy I didn't understand at the time, but it has worked. Microsoft and its PC partners now make some clever personal computers and tablets that are being copied by others -- including Apple.
But unlike the PC business, companies that specialize in smartphones are still churning out new ideas. Samsung, Huawei and Xiaomi don't need Google to show them how to do it.
That's on the hardware side. In software, when Google makes its own smartphones it can ensure they run the best of Android software. Too many Android smartphones are clogged with clunky software made by the smartphone maker.
Google's own devices will be tidier and easier to use. And with Google's own phones, the company also can spur hardware makers and suppliers to build more of the cutting-edge features Google has designed for Android, such as virtual reality functions and special processors to manage crisp digital photos.
That's a worthy goal. So worthy that Google agreed to buy Motorola Mobility five years ago, also with the aim of expanding the boundaries of Android smartphones. The market has changed, but it remains unclear whether Google can convince hardware partners to follow its lead in smartphone technology.
Sell people on the merits of Google's wireless phone service
We deserve new ideas in cellphone service, and Google's 18-month-old, low-cost Project Fi wireless phone technology is among the best. Project Fi is only available on Google's own line of smartphones.
But Google hasn't pushed its wireless service very hard, and likely still won't, again because of potential business clashes. Google can't afford to anger the traditional cellphone companies like AT&T and Vodafone that pitch Android smartphones. Awareness of Project Fi, then, likely won't spread beyond a niche of tech nerds.
To be clear, making smartphones isn't a high-risk gamble for Google. The company's $70 billion annual advertising revenue -- a growing share of which is generated from smartphones -- provides a big cushion to absorb any failures.
But Google's strength is a weakness when it comes to its self-designed devices. Its business empire makes Google simply too conflicted to make good on the promise of its smartphones.