To tell the story of the last 10 years at Google, it helps to look at the products that no longer exist.

The decade started with Wave (launched 9/09, deprecated 8/10), then Buzz (launched 2/10, deprecated 10/11). There was also a Wikipedia clone called Knol (launched 8/08, deprecated 3/12), the Groupon clone Google Offers (launched 5/11, deprecated 3/14), and of course, the beloved RSS client Google Reader (launched 10/05, deprecated 3/13), whose absence I feel to this very day. 

These are just the ones that came to mind. The official Killed By Google tracker counts 194 separate projects discontinued since 2006, although the site’s creators admit that they may have missed a few.

There’s a pattern to these shutdowns: Google will announce something cool at its I/O developer conference with a hazy release date, it will ship six to 12 months later, and it’ll last a few years before the company decides it’s not worth the upkeep. For anyone following the company closely, the cycle is relatively transparent, and it’s no less obvious from inside the company. Google likes new ideas, and it doesn’t like shutting down a project before it’s shipped. For most of the decade, the company was organized into small teams, with more resources available as the team grew, so it’s not that costly to spend a year or two launching something. Google will give it some runway to see if it takes off — but only so much. As anyone in the startup world will tell you, most things don’t take off.

The result is a kind of startup-y anarchy, housed within one of the largest companies on earth. There are real benefits. The company can take a lot of bets and never pay that much for betting wrong. Even when Google is on the wrong side of an industry trend (say, social networking or smart TVs), it rarely ends up with disasters on the scale of Windows Phone or Facebook’s stifled hardware ambitions. This is the “fail better” mantra, and Google has embraced it more than any other big tech company.

Despite the sorry endings, the churn has fueled some of Google’s most interesting ideas. Tiny AI-powered cameras and touch-sensitive jean jackets were never going to set the world on fire. But Google built them into not just prototypes, but actual products. Google Cardboard was never destined for long-term success, but it played a role in stoking interest in the technology. The Nexus line was never going to compete with Apple, but it really did push the boundaries for what an Android phone could be. These were worthwhile projects, and they were only possible because Google was so willing to fail.

But the main downside of failure is that sometimes you want to succeed. Google’s graveyard isn’t just full of well-intentioned lab goofs; there are also real, ambitious projects that had a chance and failed, in part because the company was so eager to pull the plug. Google had a chance to own messaging, and the company blew it, cycling through more than a dozen messaging systems (we counted!) without fully committing to any of them. Daydream VR could have been a true rival to Oculus, but after years of false starts, Google’s buy-in wasn’t enough to get Android manufacturers on board or convince developers that the platform was for real.

Other shutdowns have had strategic costs. The Google Reader shutdown was the death knell for RSS as a mainstream technology, cementing the Facebook News Feed’s dominant role in the spread of information. The News Feed model has become frightening in recent years (not least to Facebook itself), but competing products like Apple’s News Plus or Facebook News have struggled to gain traction. It would be immensely valuable for Google to have Reader still in place, incurring only marginal server costs — but that simply isn’t an option anymore.

Google is still competing in smart home tech, live-streaming, and augmented reality, and I’m sure they have a lot of exciting ideas still in the pipeline. But each launch comes with an uncomfortable question: is Google really committed to this? Or is this one more product slowly headed to the Google Graveyard?

There’s reason to think Google is already backing away from this scattershot approach to product launches. As the company struggles with its pressing political issues, it’s been focusing and consolidating. With co-founders Larry Page and Sergey Brin out of the picture, Google-watchers are predicting less entropy and more attention to the ways Google actually makes money. I hope it’s not true (it would make life at The Verge a lot less interesting), but I understand why it might be. The Google Graveyard is filling up — and with so many interesting ideas in the mix, it would be nice to see more of them survive.