The tech industry loves to extol its ability to recover from failure. It regularly brags that, in contrast to older, more staid sectors of the economy, tech sees failure as a learning experience and bounces back with something different and better.

But that’s a bit of an exaggeration. Lots of companies try to improve on failed or disappointing products, but few overhaul them completely and start all over. Mostly they unveil more modest iterations they hope will do the trick.

Microsoft’s Windows 7 fixed the worst problems of the awful Windows Vista, but it was mainly a modern, sleeker evolution of the ancient Windows XP, rushed to market to expunge Vista from users’ minds. Apple’s iCloud cleared the low bar set by the embarrassing MobileMe, but, it wasn’t close to a total do-over, at least at first.

Now, Apple has undertaken an almost total software reboot of a struggling product with its latest operating system for the Apple Watch, called watchOS 3. In my opinion, it’s an even bolder overhaul of the Watch than the new Series 2 hardware model introduced a couple of weeks ago, which is heavily focused toward serious fitness-oriented users, with new features like built-in GPS and water resistance for swimming. (You can find a terrific review of that new hardware by my colleague Lauren Goode, here.)

There are two main reasons for this: first, the new OS dramatically improves the user experience for the entire installed base of Apple Watches, for free, without the need for a new Watch. Second, its design makes third-party apps easier and faster to use, thus raising the odds that developers will make more and better Watch apps.

I’ve been testing this new user experience, and I’m impressed.

The problem

Since its release about 18 months ago, Apple’s ambitious smartwatch — its first new hardware category under CEO Tim Cook — has been a disappointment, at least by the standards of the giant hardware maker. Unlike, say, the iPod, it has so far failed to take a struggling category of half-baked products from other companies and hit a home run by perfecting it and attracting a huge following.

Sure, Apple says it is now the world’s leading maker of smartwatches, but it doesn't represent very much, especially for a company of Apple’s size. Creative Strategies, a firm that follows tech trends closely, estimates that Apple has about 54 percent of the global smartwatch market, but it figures that only represents about 16 million watches sold out of a tiny total of around 30 million. Apple itself hasn’t released any sales numbers. (For perspective, Apple sold 13 million costlier iPhones in just their first weekend on the market last year.)

One reason might be price, which started out at a base of $399 and is now a still hefty $369 for the new model. Another might be that the Watch could do too little without being wirelessly tethered to the iPhone.

But I believe a big reason is that, although thousands of apps have been developed for the Watch, they were hard to find on the device and slow to launch. Thus, the Apple Watch just hasn’t been a good platform for third-party apps.

In fact, in an essay I wrote earlier this year, I noted that even though I’d worn the Apple Watch every day since its launch, I was still using it overwhelmingly for just three things, none of them third party apps: activity-tracking; notifications; and, to a lesser extent, Apple Pay. watchOS 3 aims directly at those issues.

The end of the scrambled egg

I still wear my Apple Watch daily, but I’ve upgraded that original model to watchOS 3. And, in my tests of the new software over the past couple of weeks, I’ve found myself regularly using other apps, such as The New York Times, The Washington Post, and my calendar — apps I’d given up on. I’ve changed watchfaces much more often. And I’ve generally been much happier with the Watch.

So, what made the difference?

Well, Apple has changed the interface so that favorite apps are easier and faster to find and they load much more quickly. (Greater speed was supposed to have been the aim of an interim software update last year, but it largely failed to materialize.)

Now, any apps you choose can live in a dock accessible by one click of the rectangular side button. Apps in that dock are easy to find, they are kept running, and they update their data so that they appear and are useful in a couple of seconds. You can put up to 10 apps in the dock, which is now, in effect, the home screen of the Watch.

The old home screen, a bewildering array of tiny, round, hard-to-tap icons, which I call "the scrambled egg" has been severely demoted. It’s still there, but will be rarely used, mainly to select which apps to dock.

An app dock is nothing new. A version of it is on the Mac, the iPhone, the iPad, and competing platforms. But it makes a huge difference on the Apple Watch.

The fast, docked apps made another confusing feature redundant: glances. These were quick peeks at app info you could see by swiping up from the bottom of the Watch screen. They were needed because launching apps was so slow. Now, they’re gone, replaced by a single control panel for things like checking battery life and setting airplane mode.

Another little-used feature was a standalone communications center from which you could select one of a few friends and then send them stickers, or other visual items like a representation of your heartbeat. Now, as in the new iPhone OS, iOS 10, these things and others are where they ought to be, integrated into the messaging features of the Watch.

Speaking of messaging, there’s even a feature called Scribble, which turns finger-drawn writing into text. In my tests, it worked surprisingly well. And dictating messages using Siri, which was reliable for me on the old OS, is still useful.

Also, you can now quickly switch watchfaces simply by simply swiping, instead of undergoing a cumbersome process involving pressing hard on the screen and then scrolling. (You still have to press on the screen to edit a watchface, as opposed to just changing them.)

I also found that raising my wrist to make the screen turn on so I could check the time worked more reliably, but that’s not been everyone’s experience.

The downsides

In my tests, the Maps app simply was unable to find nearby places such as gas stations and restaurants, even though they readily showed up in Apple Maps on the iPhone to which the Watch was paired. Apple couldn’t explain this.

Upgrading to watchOS 3 was ridiculously slow — it took over three hours in my case.

And battery life on my original Watch with the new OS has been a little shorter, possibly because I’ve been doing more with it (the new Watch model has a bigger battery), but it has still lasted through a day.

Bottom line

The Apple Watch still isn’t a must-have device. It still needs to find a purpose beyond fitness. But, with watchOS 3, it has a chance to become much more useful, and a bit more fun and satisfying, even for people who aren’t hardcore exercise fans. In ripping up the whole software design of its latest product, Apple showed — yes — courage.