What's Wrong With Wave?

Wave fail.

When Google Wave first came out there was a ton of excitement surrounding it. I was never fully persuaded by the hour-long introduction video, but I was still an early adopter (in the limited pre-beta beta release, or whatever Google called it) and a relatively enthusiastic user. I've tried playing Sudoku on Wave. I've tried having various group meetings on Wave. Some of these experiences were moderately successful, some of them led to failed projects (goodbye ISC 2010), and ultimately Wave seems to have failed to catch on.

Now, granted, it's still in a beta version, but I don't want to dwell on Wave's prospects. I'm simply surprised that its talented development team (former members of the Google Maps team) were responsible for its confused interface design, identity issues (the hour-long intro should have been a tip-off), and (improving, but still bad) performance and reliability.

I'm not sure what lessons to draw from the struggle of the Wave team. That even the best developers in the world still fail sometimes? That earlier validation of designs is necessary? That addressing problems in existing solutions is a disastrous method for innovation if you don't consider the new problems you're introducing? I could probably go on for a while but the thing is, as easy as it is to identify problems with hindsight, I don't know (and can't know) if I would have caught these things had I actually been involved in the design of Wave.

Companies are often happy to boast about their successful design strategies and crazy new innovation paradigms, but there is an obvious selection bias at play in the reporting of these strategies. Who's going to give a TED talk on "Industry-Standard-Centric Design"? Or "Useless But Totally Cool Sounding Paradigms"? Actually, that second one might be a real TED talk, but generally speaking, despite a trizillion books about it, there are not many trustworthy resources about software project management.

EDIT: This just in, other people are writing about interestingly related material! Specifically, lessons learned from failed software products.