Reprinted with permission from Knowledge@Wharton
Last month, Motorola's Xoom tablet launched on the Verizon Wireless network. Soon after, complaints began to surface that the tablet's operating system — designed by Google — was prone to crashes when running third party applications. Experts say that while Google's typical strategy of launching new products and then perfecting them according to user feedback might work for Gmail and other online tools, it doesn't translate well to hardware-based devices like smartphones. In fact, they say, Google's penchant for experimentation by introducing so-called "beta" products with frequent updates may become a handicap as the search giant expands into new markets, such as software and set-top boxes for televisions, and tablets like Xoom.
Analysts have dubbed Google's approach "perpetual beta." Under this strategy, Google launches early versions of new products to see what sticks with consumers. The problem is that some of these experiments aren't sticking — especially when users have to pay for products. For example, last year Google launched Google TV, software designed to bring the Internet to television sets. After an initial splash and product launches with TV makers such as Samsung and Sony, the company stopped actively promoting the project. According to a March 24 statement by Best Buy CEO Brian Dunn, consumers have not yet shown a high interest in being interactive with their televisions. Meanwhile, in light of its problems with Xoom, Google is holding back on releasing its new tablet operating system, called Honeycomb, to open source developers, Bloomberg BusinessWeek reported.
Experts at Wharton applaud Google for being willing to experiment in new markets and quickly move on if the product doesn't succeed. For example, Google Buzz (a social networking project) and Google Wave (collaboration software) have both quietly fallen by the wayside. However, faculty say, Google may be due for a strategy change as it tackles new markets, such as consumer electronics, where the stakes and the expectations are higher. What works for free web services, such as Gmail, may lead to disasters for other products.
"The rise of web-based services and cloud computing are moving the entire information technology world to a perpetual beta mentality," says Kevin Werbach, a professor of legal studies and business ethics at Wharton. "One of the advantages of a centrally hosted service is that it can be iterated constantly, rather than on a staggered release schedule."
But while this model of frequent product updating may work for desktop software, Werbach and others point out, applying the perpetual beta approach to hardware isn't easy. "On the web, the launch-and-iterate cycle can be short. If your code is buggy, you gather the feedback and you can fix it in a few hours or in some cases a few days," notes Kartik Hosanagar, an operations and information management professor at Wharton. "The damage done from buggy code is not too acute. With hardware, the cycle time of product development, production and finally distribution is too long. You cannot just fix the issue and release new hardware in a couple of months, let alone hours or days."
Indeed, it may be that Google just has to get better at developing approaches for different markets, says Kendall Whitehouse, director of new media at Wharton. "Is this a Google story, or one about the characteristics of different markets?" Whitehouse asks. "If you think about software products, there are three tiers [in terms of upgradability]. Google is strongly entrenched in cloud computing. As soon as they update an online application, every customer receives the new version. The next tier is traditional desktop software, which can alert you to a new version, but typically requires the user to accept the update. The thorniest tier is when the software is tightly bundled into a device you don't control from a hardware partner."
In some respects, Google's beta conundrum boils down to expectations. When consumers spend nearly $700 on Motorola's Xoom, which competes with Apple's iPad, they don't expect the software to crash. "There is an expectation that the product will work well out of the box," according to Wharton legal studies and business ethics professor Andrea Matwyshyn. "It's one thing when an unknown company launches a new service. Google is at a different stage in its life."
Trip Chowdhry, an analyst at Global Equities Research, said in a research note that Google's tablet effort highlights how expectations have shifted for the company. Honeycomb is a tablet operating system that is too complicated for consumers, he pointed out, and Apple has set the bar very high. As a result, Android tablets will have a tough time playing catch up to Apple.
"In the past, if Google had an idea, it would work on it and launch it as beta for the end user to provide feedback," wrote Chowdhry. "The 'perpetual beta' on Gmail, Google Voice, etc., worked well as long as these products are free … even if the product is half-baked…." But when users pay top dollar for a product like Xoom, "expectations change dramatically. [They] don't expect to be guinea pigs trying out a beta product."
Experts at Wharton say the challenge for the company is navigating the two segments for its markets. One segment — web services like Gmail — can be tweaked over time. Google's Gmail was in beta for five years. When Gmail came out of beta in July 2009, Google acknowledged in a blog post that the beta tag would not suit everyone, such as "large enterprises that aren't keen to run their business on software that sounds like it's still in the trial phase."
But the other segment — consumer electronics — may require more software development rigor. "One has to separate the idea of on-going refinement from the notion of releasing a product before it's ready for prime time," says Karl Ulrich, an operations and information management professor at Wharton. "There is a trade-off between waiting too long for a perfect product and frustrating the user with a buggy product." When it comes to tablets, Google appears to be on the frustrating side of the equation for now. In the Bloomberg BusinessWeek article, Andy Rubin, head of Google's Android effort, acknowledged that design compromises were made to get Honeycomb to market as quickly as possible.
Hardware is much more challenging than online web applications, Whitehouse notes. "Google's mental framework is based around cloud computing. Hardware is different. Often, the device manufacturer is an external partner and your software is baked into the device. It's a whole new ballgame that puts more pressure on the software vendor to get it right the first time."
In other words, Google has a fine line to walk. On one side, "there's the appearance of progress and development on the fly that may indicate nimbleness — or [it could mean you're] shipping things too quickly," says Matwyshyn. On the other side "is having every aspect of the product planned out. It's a big conversation in the tech industry over what consumers expect and what is reasonable." As technology devices are increasingly woven into daily life, more is being expected of gadgets, Matwyshyn notes. "The market has shifted. People view their technology products like a car — they just want it to work."
As Google tackles new markets, it may also be necessary to be more selective, according to Wharton management professor David Hsu. Google has tried to keep up with Facebook through efforts like Google Buzz. In addition, the company reportedly tried to acquire Groupon, but now will compete with the company in offering group discounts. Google is also targeting Microsoft Office through its suite of online productivity tools. "Google is trying to do almost everything," Hsu says. "The company may have to be more strategic and place [fewer] bets."
In fact, Google may have to reorganize to address new markets, notes Saikat Chaudhuri, an operations and information management professor at Wharton. "It's an interesting organizational problem Google has. Largely everything Google does revolves around software, the web and advertising. When it expands into phones, tablets and PCs, Google has to think about how it organizes. It's a fundamental problem."
According to Chaudhuri, Google may have to take its trial-and-error phase farther before releasing a product even as a beta. In addition, Google may have to reorganize divisions based around different markets, much like Microsoft has for the Xbox. "Google will need different development processes because the cost of experimentation is high with hardware. There's a trade-off within technology. Companies can remain fairly nimble, but not at the expense of product quality."
While Google is the latest company that may have to change its stripes, it's far from the first one. Many tech companies that are operating in an ecosystem with multiple partners — hardware companies, component makers and sometimes wireless carriers — are finding it difficult to compete with more vertically integrated players like Apple.
With Apple's iPad and iPhone, the company controls both the hardware and the software. It makes the semiconductors and vetoes apps that don't meet its requirements. Google has an open source software model with Android and has multiple companies making hardware for both tablets and smartphones based on its operating system. In many ways, Google's current strategy for hardware resembles that of Microsoft during the PC era. Both companies supply operating systems that run on hardware from a wide range of manufacturers. And, like Microsoft with the PC, Google's Android platform has garnered the most smartphone market share. What's unclear is how the competition will play out this time. Against Microsoft, Apple's vertically integrated strategy with the Macintosh failed to gain a large market share. Today, Apple's approach is favored in the tablet market, which revolves around devices that, as Matwyshyn says, "just work."
"Obviously, you have less control over the full user experience with the supplier/partner approach," notes Hosanagar. "Many Android developers have complained that they do not know which hardware device they are developing for. So, their app may look great on one device but not on another. iPhone developers know the exact specifications of the hardware on which their app will be consumed. This obviously allows them to provide a consistent experience. Google's strategy allows diversity and ensures there is an Android for everyone. So there are pros to the partner strategy as well."
For now, the technology market appears to have shifted to a more integrated approach, say experts at Wharton. "Microsoft hasn't been as successful as Apple, in part because [Microsoft has] less control over the hardware," according to Werbach. "Google's challenge with Android is how to execute on the perpetual beta approach smoothly when it has many hardware partners that it doesn't control. There will be rough spots. On the other hand, Google built Android with continual updating in mind, and users today have more tolerance for glitches because they have been trained by the perpetual beta experience of the web."
"What you call the 'perpetual beta' has its origins in the spiral model of product development," Ulrich notes. "Instead of fully detailing in advance exactly what the product will do and then engineering to that specification, the developer rapidly iterates through define-build-test cycles in order to take advantage of the learning that occurs in the interaction with the user. Obviously this approach would not work well with a Boeing commercial airframe. [That type of] product has to be right from the start and you don't want to have to maintain many different versions of the product in the field."
Ultimately, Matwyshyn argues, the courts may also determine software standards. For instance, it's only a matter of time before there's a lawsuit because software that was rushed to market causes harm to someone. "As software becomes more integral to daily life, it will be held to a higher standard," says Matwyshyn. "Software has gotten a free ride for liability."
While Google's perpetual beta cycle may not succeed in all markets, experts at Wharton say it's too early to tell how well the company's strategy will fare in new device markets. "I will be really curious to see what the tablet market looks like in six months to a year from now," Chaudhuri says, noting that if Google — or some other vendor — can organize properly by aligning internal and external partners as if a product was vertically integrated and deliver a no-compromises tablet, the dynamics of the sector could change. "Apple's vertical integration has worked because it has this magical iPad device, but it hasn't seen any real competition yet."
Reprinted with permission from Knowledge@Wharton, the online research and business analysis journal of the Wharton School of the University of Pennsylvania