Aug 26, 2014 · 8 minutes

Amazon's surprise purchase of Twitch is fascinating for a few reasons. It confirms that live-streaming is no longer a fringe potential in the online-video market. It underscores how the M&A market is increasingly preferred to the IPO market as the exit of choice for the most promising startups. It shows how far the Internet giants like Google and Amazon are expanding onto each other's traditional businesses.

And, in what I find to be the most interesting angle – it illustrates how the hands-off approach Amazon has long taken to acquisitions is strongly preferred by the cream of startups to the more common approach: integration at all costs.

In May, Google entered talks to buy Twitch. In July, this deal was confirmed as a sure thing. Today, to lean on the tired metaphor of M&A as corporate marriage, Twitch left Google standing at the altar to run into the more welcoming arms of Amazon, accepting a slightly lesser price, no less.

This is quite the black eye for Google, and the venture investors that fostered Twitch through its hyperactive early years will have some 'splaining to do to Google in future closed-door meetings. And yet Google may also have some things to learn about what hot startups want, beyond cash or stock, from an acquirer.

On the one hand, Google is big enough to shrug this off. Just as it shrugged off the rejection of Groupon in 2012 (which remains valued below Google's $6 billion offer). And its rumored bids for WhatsApp and Snapchat (which have increased in value). On the other hand, the deal may well have fallen through because Google is big enough to shrug these things off. And if Google is big enough to shrug off a billion-dollar deal that doesn't work out, it may not be invested enough to make sure it does.

Even so, this jilting is especially embarrassing to Google for two reasons. First, this was seen as a done deal, even as the silence that lingered after the initial news stagnated into uncertainty. Second, Twitch is no mere rising star. It's emblematic of an important emerging business model – video platform as community. It's what YouTube wants to become as it enters middle age.

Twitch had 45 million monthly active users when Google was said to be the buyer. It now has 55 million. The company has its roots in the idea of lifecasting – the undying dream that a camera strapped to you would produce content as fascinating to others as it is to you. This idea proved to be very true in online gaming, the way televised sports is compelling to people terrible at sports, and Twitch was developed to exploit this idea. As its founder and CEO Emmet Shear explained in an interview last year:

It depends a lot on what kind of gamer you are. If you love Call of Duty multiplayer, Twitch offers you the chance to see some of the sickest kill streaks you’ve ever seen and watch tactics from the very best. If you love Minecraft worlds, Twitch lets you see all the mods you could be using and get inspiration for new creations. If you love League of Legends, Twitch is your chance to watch the top teams in the entire world battle for supremacy.
In time, it became clear this wasn't just a video platform that attracted a community of die-hard gaming spectators, there was a role to play for incumbent gaming companies as well. When Google was said to be buying Twitch last May, Citigroup analyst Mark May wrote:
Twitch is also a video platform that is used by publishers, developers and others (e.g., Sony’s PS4 and Microsoft’s XBOX) to incorporate community and interactivity into their services.
So let's get this straight. Gaming companies created consoles and high-def content that attracted gamers. Some of these gamers were better than others, and the better among them created their own content on Twitch that drew the attention of the not-(yet)-so-great. This aspirational culture led to classic network effects, helping gamers connect with each other, and the gaming companies profit as a result. All while Twitch grew like a mushroom.

And Google let this go?

Google has been trying to cultivate this type of experience within YouTube, with very mixed results. The hamfisted effort to force YouTube, along with all things Google, into the indifferently received Google+ may have persuaded some converts, but it alienated even more. Second, people still arrive at YouTube through search more than social discovery – when we think of YouTube as a community, we tend to think of the mind-numbingly insipid comments.

In this respect, Twitch has emerged as an anti-YouTube – a community of video enthusiasts that didn't have to be cultivated because it grew organically around a compelling spectacle. When reports emerged of a Google acquisition, ardent Twitch users fretted loudly that everything good about the service would slowly fray away as the company was grafted into the larger Google experience.

With Amazon buying the company, there may be a more cautious sense of hope. In the Internet industry, there have emerged, broadly speaking, two schools of acquisitions. One is to buy a startup, integrate it wholesale, and engage in wishful thinking that its magic sparks something big and new inside the parent company's aging operations. This almost never happens – Yahoo's history of deals is a good example – but sometimes it does, as with Priceline's purchase of overseas startups.

The other school of M&A thought has been recently exemplified by Facebook's purchases of Instagram, WhatsApp and Oculus VR – big-ticket deals that aren't integrated for a good long time but rather left to mature on their own, while still benefitting from the financial security of a deep-pocketed parent. For all the credit Facebook has earned on this front, Amazon has taken this approach for years. was bought by Amazon in 1998, yet many users still are surprised to learn of the relationship between the two. Visit, bought by Amazon for close to a billion dollars in 2009, and there's no hint in the site is an Amazon property. Devotees of Goodreads, bought last year, worried that Amazon would ruin the site, but it remains largely unchanged post acquisition.

All of these properties, of course, seamlessly and invisibly give Amazon one of its most precious resources – consumer-usage data.

It's not clear yet why Twitch dumped Google for Amazon. Forbes suggested that antitrust concerns prompted Google to walk away, an unlikely explanation when you consider Google's control of video streaming is far weaker than its grip on search. Bloomberg's early reporting on the Amazon purchase indicated that investment banker Frank Quattrone sought out other suitors for Twitch as the Google talks continued in an effort to spark a bidding war.

But Amazon says it paid $970 million in cash for Twitch, less than the $1 billion-plus price Google was rumored to have offered in May. So why did the bidding war result in a discount? In a letter to the “Twitch community,” Shear dropped hints that seemed to contain a few buried barbs for Google: “because [Amazon] believe[s] in our community, they share our values and long-term vision, and they want to help us get there faster.”

Faster than what? Google, which by inference shares fewer values and less faith in the Twitch community? Shear added, “We're keeping most everything the same,” underscoring the hands-off approach to acquisitions Amazon has favored.

It's not that Google is blind to the hands-off approach. When it bought YouTube in 2006, the company largely left it to grow on its own for years before mining it (profitably so) for ad revenue and later stuffing it in the Google+ envelope. But over the years, other startups (Dodgeball, Feedburner, Waze) have either lost their creative mojo or disappeared entirely once they entered the Google family. What's more, Google's track record in video has recently felt especially suffocating to creativity.

Amazon has an advantage over Google. Its revenue comes primarily from selling things, whether Prime subscriptions or consumer goods. Google relies heavily on ad revenue. Amazon doesn't have to worry about the battle between search and social that is a constant threat to Google's growth potential. What it needs is to understand the shopper better. It can not only afford to do this while leaving its acquisitions alone to do their thing. It can also benefit from doing so.

But again, Google can weather the occasional black eye, even the embarrassing jilting every year or so. If we journalists keep relying on the marriage metaphor to describe M&A's, then corporations are the most shameless of polygamists. Google has also made many acquisitions of late – DeepMind, Nest, Skybox, Dropcam – in markets that are only beginning to emerge.

The larger issue for Google's M&A strategy is that Google is big. Google is expanding the way a plate of spaghetti expands as you throw it at the wall. Google doesn't know what will stick when it finally gets there. And who wants to be one of the strands of pasta that ends up on the floor?

My guess is this is why Twitch went with Bezos, who keeps his pasta on his plate. Bezos and Page are both control freaks, both obsessed with chaos. Page, who sees chaos as a beast grazing in some future frontier, gives his acquisitions more free reign according to how far out their promise lies.

If you have near-term promise, which Twitch has in spades, you're better off with Bezos. For Bezos, chaos is the consumer – an animal understood, and therefore tamed, only through the endless points of nonsensical behavioral data it casts off. The Twitch community may believe it bought its freedom. But they are only free to be watched.

And Amazon will be watching.