With over 300 users now, it turns out they do want browser extensions and mobile app. So our product is far from perfect. You can have a look at it -
SearchTrack helps teams and individuals collaborate to save content, contacts, products and services around any topic. Best of all, you can share your research with the community so others in need can spend more time learning and implementing from your collection of rich resources (than they would, in finding and organizing those).
It was far from perfect when it launched, was still far from perfect when I took over, and I'm going to guess it is going to take a year or so before it gets to a point that I'm even happy with it.
The issues are plenty.
It looked horrible when I took over, and though we've done a redesign, it is still not a beautiful app.
Bottlenecks galore! Very slow page load times, parts of the app where we wanted to improve the UX are so duct-taped together we can barely touch it and are biding our time to do a re-write.
Dead code, confusing code, poorly thought out code. This has all of it. I even lost a developer because he was so frustrated with the code structure.
So, how bad can this all be?
We've got a large base of dedicated users that love the product, we're growing nicely (though growth is somewhat held back as we can't release features and improvements as quickly as we'd like).
Ignore the 'perfect', it's never perfect (but definitely aim for 'good enough'). Having users that love the product is much more important than it being perfect.
Also, what app do you think is or was perfect when it launched? Facebook sure as hell wasn't perfect. Twitter? Clearly not.
Nobody is perfect, give that up. The perfect time to launch an app is when it is perfect enough to get feedback.
Unless you are some kind of a wizard, waiting to clean everything up before making sure you are doing the right thing is probably a bad idea.
I think there is a legitimate caveat to this though. If you're app does something where failure leads to death or some other really bad outcome (might be medical software), then being super lean is probably not advised. But in the main, launch early, test, fix and repeat.