Google’s “20% time,” which brought you Gmail and AdSense, is now as good as dead

And how many failed projects came out of 20% time?

They probably didn't lose as much money as they're gaining from the successful ones

Does this matter if the engineers and/or the company learn something from the failed project?

We under-value sincere-but-failed projects.

I'm sure the gains from AdSense and Gmail alone outweigh anything they lost. Also engineers can stay happy if they work on things they like more often. So there could be an implied cost (by losing employees) to not allowing 20% time. I'll also add that it's almost impossible to stay focus on 1 thing 8 hours+ a day for 5 days a week. Even if there is no 20% time I'm sure they still lose time to unproductive activities. Maybe 20% time actually captured some of the time that would otherwise be spent browsing the web or making pointless comments like this on HN.

It's not as if Google never would have thought to make a web-based mail if someone didn't do it in their 20% time...

Failure is a prerequisite for success, though. As long as those failures were inspected, digested, and understood, then the company gained more in terms of training than they lost in operational costs.

None, failures tend to be abandoned early in most cases.

Besides if you care about failure then you don't belong in the tech industry.

What is your definition of "failure?" A project that nobody uses and makes the company no money is a failure in terms of direct profitability, but may be a success for other reasons. Your engineers get to exercise their minds in unusual ways and keep themselves sharp. Your engineers are more willing to put up with boring tasks. You attract talented people who want to be self directed at some level.

At the time, Google was a search engine, period.

Expecting Google to build a webmail system would have been like expecting Twitter to release an accounting package.

Does that matter as long as you get a multi-billion dollar business out of it every now and then?

What a strange sentiment. We should all just not care about being successful?

How many successes you had matters. How many failures you had in addition is basically irrelevant.

How many failures gave you the knowledge and/or practice to allow a follow on (non-20%) project a better chance of success?

Search and email are closer than tweeting and accounting. Not even a good comparison.

You should care about successes, but you should ignore the failures that will follow.