One on One with Dan Keldsen on the demise of Google Wave

Google Wave came crashing down last week. The collaboration platform that was announced to much fanfare in May of 2009 came to a sudden and abrupt end. I asked Dan Keldsen, a principal at Information Architected, who worked with Wave quite a bit, why he thought it failed to catch on.

FCM: You were able to use Google (NASDAQ: GOOG) Wave in creative ways. Why do you think it failed?

DK: The earliest days of Wave were too limited to enable people to really use it. As I'd said in my early commentary, even though the buzz generated from the limited alpha/beta release made for great media fodder, it made it nearly impossible to actually collaborate, if you didn't know enough people with access to really use the system.

When I met Greg (aka Dr. Wave) at the Enterprise 2.0 Conference in San Francisco last year, I had the chance to speak to him and his team for a good 45 minutes, and I had to confess to him that even though I'm an extremely early adopter of tools like Wave, and a pretty geeky, self-taught, mostly reformed IT guy...even for me--a prototypical Wave user--it felt like work. Having real-time conversations in a wave--well, you could do that in IM, IRC, ICQ, etc. And yet, aside from IM, almost no business users that I've ever met have used those tools, and they had already been around for many years.

Actively collaborating on a "document" in Wave adds even more complexity. The current UI doesn't particularly feel like creating a document, and even for those who are Wiki-savvy, and are used to co-editing documents (suffering through "the lock" to wait for hand-off of editing privileges), while people found the demos of realtime co-editing to be stunning, the actual use is very odd. I don't believe we've developed the collaborative skills to really make that mode as effective as it could be, and will be, once this becomes the new norm for document creation.

FCM: It seemed Google was on the right track with Wave, but perhaps it was just too complex for the masses. Why do you think Google canceled it instead of trying to improve it or simplify it? 

DK: The current UI had always been seen inside Google as a bit of a throw-away, at least according to my discussions with the Wave team. It was an experiment, and from my standpoint, as the son of a chemist, I have to say that as a fan of the experimental method, if they actually learned what they were hoping to learn from this experiment, then it was high time to end the experiment and move on. As Michael Hammer would say, sometimes it's not worth the effort to fix something, blow it up and try again! Even though I frequently work through a continuous improvement lens (Lean, Agile, Six Sigma), there are plenty of times when you need to jump WAY out of the current approach and start fresh.

The other piece to consider is that people don't just do collaboration for no reason. They have some context for collaboration. When people say the "want to collaborate more effectively" - that statement by itself is really almost meaningless. Collaborate to solve a problem? That needs to be resolved in two years or in five minutes? That will involve two people or 20,000? Does that collaboration need to be kept and referred to continuously moving forward, or does it only need to be seen and understood by the participants in the moment?

I've offered some models of how to think about the appropriate tool sets for collaboration in a presentation called "Real-Time Working with Collaboration" (http://www.slideshare.net/dan.keldsen/realtime-working-with-collaboration), which I hope, hammers home the types of contexts to keep in mind.

What I found most interesting about Google Wave, were the implications and demonstrations of the "mash-ups" of Wave and, for example, SAP's Business Process Modeling environment. Anyone who has done modeling work understands that you don't normally have a single view of a process, particularly when it isn't yet automated to any great degree. By turning what is normally a very siloed and solo activity (modeling) into a group collaboration experience, in real-time, the odds of actually capturing a real representation of the model and refining it in advance of deploying it (or rapidly switching into a live, automated process and reflecting on what did and did not work), is a very interesting view of the near future.

The ability to have collaborative experiences directly in the context of a real work process or application is something that people had no idea they were waiting for, but is going to blow their minds and revolutionize their work.

Of course Wave isn't the first environment to offer this glimpse, it's been seen before with the OpenDoc movement around 10 years ago, and of course the rise of portals and more recently, with 2.0-style mashups. But Google Wave was the biggest and lowest cost experiment in this area, and the early partnerships/examples that they've shown is where I believe the future of the core of Wave is going to live on.

FCM: As someone who used Wave, what were its primary strengths as a platform?

DK: I don't think this is the end of the core technology behind Wave, and they've said as much. The distributed architecture for sync and federation, the ability to do real-time in a browser rather than a fat client ("pure" web vs. Java, Flash, Silverlight, Air, or native Windows/Mac/Linux app), the ability to include and embed interactive agents/bots--I expect those to all surface elsewhere.

Those are distinctive strengths of the platform, unfortunately, the ease of use of the current state of the UI made tapping into that power a bit too geeky for the masses in it's current iteration.

The decision to limit access to Wave, probably for too long, and then to soft re-launch to anyone with an email address to broaden the reach, making it possible to finally collaborate with a much larger crowd, is what ultimately caused this iteration of Wave to come to an end. We'll have to see what the Wave Phoenix holds when it is reborn--and my bet is that it's reborn in the context of much more specific and pointed collaboration applications or contexts, and not just as a raw capability.

Related Articles:
One on One with Dan Keldsen of Information Architected
The rise of the social in SharePoint 2010
Google waves bye bye to Wave
Google Wave: Is it useful, or just cool?