The new Google Home. Michael Short/Bloomberg

The war rages on over the connected home—an idea that home appliances and accessories will all be connected to the Internet. Soon, your fridge will send a text message when the milk is running low.

For most people, that prospect is still far from being today’s reality. But this Black Friday will spark the first battle between Amazon Echo and Google Home. Both products bill themselves as a Wi-Fi speaker and a sort of Siri-in-a-tube that answers users’ spoken questions. They weave together a search engine and artificial intelligence (AI) in the form of a female voice assistant. She can support music streaming, turns off lights, checks flight status, tracks packages, and much more. Google Home looks like a tabletop air freshener; Amazon Echo an aerosol spray can. So what’s the big deal?

An Amazon Echo. AP Photo/Mark Lennihan

Probably for the first time in the history of audio equipment, consumers pay little attention to the audio quality itself. Conventional specs, like strong bass and clear highs, don’t matter much. What does matters is how a device works with other gadgets around the house and how it integrates an ever-growing number of services into people’s lives. And this is where Jeff Bezos shines.

Taking a page from the iTunes playbook, the Amazon CEO has aggressively pushed his 1,000-employee team who work on Echo to speed up certifying third-party apps, aka “skills”. Back in June, the company announced that they’d reached more than 1,000 skills – which was up from 135 in January. Today, Echo masters 3,000, including Uber (hail a ride), Fitbit (review health stats), Mixologist (make a cocktail), Domino’s (order a pizza), Kayak (book a hotel), Capital One (check your account balance), CNN, NPR, and the like (get news), plus applications from other device makers (Philips, Samsung, and GE), and most recently, Twitter.

And how many third-party apps does Google Home currently have? One. Uber. In the funniest video showdown, Joanna Stern from the Wall Street Journal interviewed both speakers and asked them to remind her “to buy a monkey”. Amazon Echo did it with ease, as it has added Google Calendar support. Google Home, on the other hand, replied “Sorry, I can’t set reminders yet”.

“Google on paper is much better positioned to do something like this than Amazon,” said Jan Dawson, founder of Jackdaw Research. For all its financial success and world-leading technology, Google however, has been lagging behind for almost a year. Despite much fanfare at its launch event, Google Home is eerily reminiscent of last year’s Amazon Echo, in a growing market where it should have dominated. Why so slow?

When Decentralization Works

One of the most admirable traits of Google’s structure has been its decentralization. Product groups, from online search to mobile Android, are given the freedom to work independently. Google’s legendary 20% time perk to work on side projects one day a week has resulted in numerous great products like Gmail and Google Maps. And yet, the only reason such tremendous variations did not lead to utter chaos is because Google has a clearly defined strategy: Give away products for free, rapidly expand the customer base, drive consumer usage, mine user data, and then sell advertisements. Though there are no audio ads on Google Home at the moment, the voice assistant collects our queries just like the web search counterpart does.

In fact, advertising revenue has been Google’s mainstay since the company was founded. At $60 billion a year, Google is the de facto world’s largest advertising company by revenue, dwarfing News Corp ($6.9 billion), Hearst ($4 billion), and Time ($2.9 billion). Only $8 billion of Google’s 2015 revenue came from non-advertising activities. Anytime Google has veered from this successful formula, it gets itself into trouble.

Back in 2012, Google bought Motorola for $12.5 billion in an attempt to enter the hardware business, but the company ended up fire-selling Motorola at $2.9 billion to Lenovo. Its Nexus tablet never took off. Nest, a start-up that Google bought for $3.2 billion, has basically stuck to its initial product—the learning thermostat. And the much-hyped Google Glass, which heralded the arrival of augmented reality, turned out to be a flop.

Even its pioneering autonomous-car project, which Google started early, has failed to gain traction, only to be surpassed by Uber and Tesla. It’s as if there’s an invisible hand at Googleplex that derails all initiatives that don’t fit its advertisement model. “If you can’t immediately sell ads on this thing, we’ll kill it right away.” This is how decentralized innovation has slowed Google in building the smart connected home.

The Self-Appointed Adult at Amazon

To be fair, Amazon also has its long list of failed projects, including Amazon Destinations (hotel booking), Endless.com (high-end fashion), and WebPay (peer-to-peer payment). The chief fiasco is none other than Fire Phone, which employs a series of cameras to simulate a 3D screen, changing its image as the user moves the phone. This feature is at best a party trick, and all other features are designed around Amazon’s own services, with the phone’s biggest selling point being how easy it made shopping on Amazon. Adding to consumer skepticism was the small number of apps. One year in, CEO Jeff Bezos pulled the plug and took a $170 million hit, with some $83 million worth of unsold phones collecting dust.

What’s remarkable was that the CEO took personal responsibility to publicly justify those projects. “I’ve made billions of dollars of failures at Amazon.com. Literally,” Bezos recounted. “None of those things are fun, but also they don’t matter. What matters is companies that don’t continue to experiment or embrace failure eventually get in the position where the only thing they can do is make a Hail Mary bet at the end of their corporate existence. I don’t believe in bet-the-company bets.”

Unlike at Google, Jeff Bezos personally set the trajectory for his company’s myriad experiments. Bezos, insiders say, was “the product manager.” At Lab126, Amazon’s Silicon Valley–based R&D group that is best known for developing the Kindle e-reader, Bezos told employees not to feel bad about the Fire Phone’s performance, because the company had learned valuable lessons from it. Nonetheless, he set out to break down silos as various programs at Lab126 grew from mere skunkworks into a formidable hardware maker of more than 3,000 employees.

To seamlessly blend hardware and software demands an interdisciplinary approach to product design. Bezos stayed close to the development team. “I would see him brainstorming wild ideas with the industrial design team, or discussing font sizes and interaction flows with the UI team,” recalls a former top-flight designer. Bezos continued to play the role of the self-appointed adult, imposing ever-loftier demands on the smart creatives at his company. “Jeff had a vision of full integration into every part of the shopping experience,” said another employee at Lab126.

The Silo Buster

Internet companies are the fruit flies of the business world. Changes are rapid and life cycles short. In such an environment, there is little tolerance for inaction. Missing market trends can quickly lead to diminished earnings. Missing them a few times sends the company packing.

All this is not to suggest that Amazon will always dominate the connected-home market. But a defining characteristic of Amazon has been its ability to expand into new areas that diverge from the company’s past. From Amazon Web Services (AWS) to the Kindle e-reader to Fire TV, each business operates with different resources, working processes, and profit formulas to deliver a compelling value proposition to end users. Bezos is the mobilizing hand that breaks all the rules to keep corporate resources fluid.

Sometime back in 2012, the CEO began to insist that all services provided by AWS be built in a way that they could easily communicate with each other and with external parties over Web protocol. For cloud computing services, computers have to swap information with the servers in the cloud. That means Amazon must enable a common language, so that computers, regardless of different versions of operating systems and software, can talk to one another.

Bezo’s instructions on creating such a common interface—an application programming interface, or API—were written in an email that finished with a characteristic signature, “Anyone who doesn’t do this will be fired. Thank you; have a nice day!” Such authoritative tactics would seem unthinkable at Google, but it’s precisely the kind of intervention that breaks silos inside a large company and helps put the customer first.

This is also how Amazon Echo has developed its 3,000 third-party skills. Back in June, for example, Amazon released improved “Skill Kits” for developers. Third-party programmers can easily reuse common skills (e.g., navigate to the next item in a list, pause an action in-progress, go back to a previous item, or resume an action), and integrate these basic functions for more advanced features.

It may sound easy to open one’s system to third party developers, but building a set of easy-to-use tools and motivating people to use them are often less than straightforward. Obviously, Google has some work cut out for them in the coming months. “To do this well we really need to work with developers and third parties so we can provide these actions to our users,” conceded Sundar Pichai, Google’s chief executive.

The First Battle

Thanks to its ubiquitous search and an all-powerful algorithm, Google Home beats Amazon Echo hands down on answering unstructured and random questions. (E.g., what’s the average height of a monkey?) Rather than issuing voice commands the entire time, users can have a more intuitive and conversational interaction with the voice assistant.

But judging from utilities and functionalities, Amazon Echo is undoubtedly the winner for now. The competition, however, is far from over. Google Home may soon integrate many of the company’s existing services (Google Maps, Google Calendar, Google Hangouts, Gmail, and countless others) and make it so smart that Amazon Echo will look like an idiotic Chatbot.

Still, it would be crazy to assume that Jeff Bezos will just sit idly by and watch the world spin around. Who knows? Even Microsoft is reportedly developing a similar device. One thing’s for sure: the game is on.

Howard Yu is professor of strategy and innovation at Swiss business school IMD. Stay connected on LinkedIn and Twitter (@howardhyu).