Google's Waterloo: Release-And-Revise Fails Beyond The Web

The search giant needs to slow down, deliver on its promises and ensure that quality isn't trampled in the rush to market.
Google TV could have been great. It's based on Android, the presumptive heir to a mobile ecosystem defined by, and until recently dominated by, Apple. Google TV promised to unite television content and the Web with a search box, a friendly interface, and a keyboard--an appealing thought for anyone who has tried to search TV content using the sub-par tools provided by cable companies. And soon, Google says, it will let developers create Android apps for the living room, where viewers and their wallets are presumably waiting for engaging experiences.

Google TV is nothing if not creative. Consider this recent patent filing, "Displaying Advertisements on Blank Television Output," from Simon Rowe, Google TV tech lead in the U.K. Rowe sees an opportunity to deliver advertising during the gap of one to ten seconds when channels are changing--advertising that capitalizes on channel surfing.

But Google TV has been too much, too soon--and it's a warning for any company trying to convert its success in one market to another. Google mastered rapid innovation in the cloud. The company's executives, product managers, and engineers talk about speed and innovation constantly. They love to cite the number of features added to the company's online services--more than 130 just to Google Docs last year, for example. But Google still has something to learn about speed limits, quality control, and the cat herding that's otherwise known as working with hardware makers, content partners, and developers. Speed matters, but so too does execution. And when it comes to delivering the ecosystem needed around hardware like Google TV, Google's approach of "beta now and improve it later" hasn't cut it.

Google so far has failed to deliver the software that promises to make Google TV really interesting: Its Android SDK for Google TV. The SDK is likely to be introduced at the company's developer conference in May, but it should have been ready in October last year when Google TV launched.

Incompleteness is a problem that goes beyond the Google TV-specific Android SDK and afflicts all of Android. As Global Equities Research analyst Trip Chowdhry observed in a recent research report, the view among some 150 Android developers that his firm surveyed is that "Honeycomb," the tablet-specific version of Android, is "dead on arrival" because it's buggy, incomplete, and plagued by a poor user interface.

What's more, Google TV depends on hardware and, so far, Google's hardware partners haven't managed to make up for the unfinished software. Sony has said sales of its interpretation of Google TV, known as Sony Internet TV, haven't been as good as anticipated. Logitech has said as much about its Google TV companion box, the Logitech Revue. "To date the [Google TV] platform has not met widespread consumer acceptance, and our sales of Logitech Revue and related products have been below our expectations," the company said in a financial report in February, though it expects the situation to get better as features and functionality are improved.

Browse through the Logitech Revue support forums and the dissatisfaction is evident. Consumers don't like half-baked products. Google essentially acknowledged this in March when it decided to delay releasing the open-source code for Android 3.0, Honeycomb. Google explained its decision by noting that it has more work to before its code is ready. It doesn't need poor hardware reviews killing the Android buzz.

Google also brought Google TV to market without enough supportive content providers. Either spooked that Google TV might encourage cable account cancellation or resentful that they weren't been courted with sufficient cash, ABC, CBS, Fox, NBC, and Viacom blocked Google TV users from using the Google TV Chrome browser to access the shows they already were making available on the Internet. Google appears to have learned something from this; the company is said to be ready to announce a streaming video rental service through YouTube, with partners that will provide content rather than obstruct it.

"Content owners were not happy with the Google TV approach and held back their content," said Al Hilwa, program director of applications development software for IDC, via email. "I suspect this is at least in part what is behind taking YouTube to a higher value video store with rentals, for example. Google will need to work out deals with content providers as well as improve the user experience and the application development model to get a device like [Google TV] to the next level."

Google may have revolutionized software development with its online approach. Its release-early-and-often approach has forced Adobe, Microsoft, and Mozilla, among others, to re-evaluate software development cycles that used to drag on for years. But the search company has to think more carefully about balancing speed with the realities of marketing, particularly when physical products are involved.

Apple wasn't the first to make an MP3 player, a mobile smartphone, or a tablet computer. But its iPod, iPhone, and iPad succeeded nonetheless, because they were so much better than the competition. Google needs less speed and more excellence, internally and from its hardware partners.