First impressions of Google Wave

After spending a few hours using an early version of Google Wave today, it's clear that in its initial incarnation it won't be ejecting existing enterprise collaboration tools from the workplace any time soon. It's not that it isn't impressive, far from it, however Wave's complex interface and open-ended feature set provides an unexpectedly steep learning curve, particularly from a company that is famous for simple, powerful user experiences.

That said, Google Wave holds considerable potential for bringing next-generation Enterprise 2.0 capabilities to organizations looking for best-of-breed solutions.

After all, e-mail itself is decades old and even highly successful Web 2.0 communication tools like blogs and wikis have gotten somewhat long in the tooth, at least in their most common forms. With browsers capable of doing more than ever and tight integration with existing information assets becoming more and more critical to users, Google Wave attempts to up the ante by combining many of the features and capabilities we come to expect in modern Web applications.

These advancements include truly social conversation, simultaneous multi-user editing, connection to external Web/intranet apps through extensions and embedding, and much more. In fact, as we'll see, Google Wave has virtually all of the key ingredients to comply with my FLATNESSES mnemonic for identifying effective, Enterprise 2.0-capable applications.

The end result is something that comes across as a distinctly sophisticated Web application clearly made up of many elements that sometimes behave somewhat unpredictably precisely because it's designed to be highly extensible and freeform. Admittedly, my experience was with the developer sandbox for extensions, but this is exactly the intent of Google Wave: to be the center of integrated communication and collaboration in a dynamic and immersive yet safe experience.

Here are some of the observations I made during my use of Google Wave. Note that this is an early version of the software that will undoubtedly be richer and more complete upon release, though experience shows that Google rarely makes major changes to products once they are shown to early audiences.

Observations on Google Wave

1. The basic interface looks a lot like Gmail. This is generally good since Gmail is widely used and understood by millions of people. The biggest obvious difference is that the inbox/content area that takes up most of the page in Gmail is now split in half, with a list of waves on the left and an active wave on the right. The rest of the page is taken up with a Contacts pane, just like in Gmail, and some standard boilerplate links on the upper right. In fact, it's so consistent with the Google experience (including Google Accounts) that it seems quite likely -- to this author anyway -- that Google Wave capabilities will be added to Gmail at some point. Upshot: Other companies can and will make their own front end editors/viewers for waves and this user experience has few surprises. It is very much what you'd expect from Google with a user interface/navigation consistent with their other applications.
   
   
   Screenshot of Google Wave: Strong similarity to Gmail
2. Google Wave works better with groups of contacts.While this seems obvious, the issue is that online conversations tend to work better when they can involve a wider range of people than just those that you think of immediately. The tedium of starting a wave is that you have to add all the participants than you'd like to have in it. Auto-joining groups are supported at this time in a fairly interesting fashion (if slightly unexpected, see below in robot participants), but will be critical to create easily and quickly en masse in order to make Google Wave useful and time efficient. One potential issue: Supporting cross boundary waves and simultaneously supporting Google Accounts, Active Directory, and other user account databases. This will be a complex issue for enterprises that want to have waves that extend outside of their organization, as many will, at least until trusted extensions are created that deal with it.
3. Simultaneous live conversations create new collaborative patterns. In my conversations within Google Wave, the real-time capability of the tool changed the nature of the conversations themselves. Unlike the post/response pattern of classic and 2.0 tools both, including e-mail, IM, blogs, wikis, forums, etc, where you get to complete a thought privately and then dispatch it for others to view when you're done, the default in Google Wave is for others to see what you're typing, live. This can have a disconcerting effect and tends to change what is captured by the tool since respondents can start answering before you've even finished, altering what you've typed or making you inclined to abandon the thought completely. In contrast, one of the most important aspects of Enterprise 2.0 apps is that they can capture complete questions/inquiries and the subsequent answers so that the network learns as a whole from the distilled, uninterrupted interaction. Google Wave has the potential to disrupt this valuable pattern that builds collective intelligence, though the feature can certainly be turned off as well. Organizations are encouraged to monitor and remediate this (possibly through usage guidelines) to ensure they get the full value of the platform.
4. The notion of participant as either user or robot works well, making the social fabric of conversation both novel and broader. With Google Wave, participants aren't just people, they are often software extensions monitoring the wave that can then independently add to the conversation (such as performing real-time language translation, injecting views of Web/enterprise data, or even whole applications.) This is true of the initial group creator and Web publishing capabilities of Google Wave in particular, at least in its current form, and takes a little getting used to. This is also where the implicit assumptions of Google Wave may lose neophyte users. Whether they will be able to understand the significance of adding robot participants (non-human software extensions) to a conversation, including that they can cause unintended consequences including unexpected and significant security implications, remains to be seen. However, from a usability perspective it might actually make a lot of sense for users to add a person or a piece of software to a Wave in the exactly the same way, with largely the same meaning.
5. Waves are strongly conversation-oriented instead of result-oriented. Because each contribution has a visual boundary associated with it and cannot be made to blend into a finished work product, Google Wave cannot replace, for example, the classical wiki as a group editing tool. In the Google world, Google Docs is a better example of a service that can help multiple people create a unified artifact. With Google Wave, the default central artifact being created is a threaded conversation. Note that extensions and gadgets make it possible for 3rd parties to change this, further increasing the potential for multi-modal confusion as users try to understand what interaction model they're engaged in. This open endedness is a boon and a barrier both; Google will have to be careful to balance this aspect of Wave so that users can access the most value without cognitive dissonance setting it.
6. Public waves will make Google Wave easily distributable and viral. Unlike Gmail, Google Wave can expose conversations publicly if desired, allowing them to be moved to wherever the conversation/information needs to be exposed on the Web/intranet. It will also likely cause Google Wave to be adopted more virally than most of Google's other applications since users can encounter waves all over the network, wherever users want to put them. This leverages Jakob's Law nicely and is one of the more powerful aspects of modern Web application distribution that Google Wave clearly gets right.
7. Google Wave supports virtually all the key elements of Enterprise 2.0. Google Wave is very strong in all of the FLATNESSES components except possibly for signals. As you would expect from Google, search is extraordinarily good within Google Wave, making it possible to quickly find the waves containing the information you're looking for. Tagging, which has been particularly important for many Enterprise 2.0 deployments, is also present in Google Wave as well as extensions, social capabilities, and all the rest. The largely missing element signals (which you can actually argue is present in the form of the inbox) means that Google Wave doesn't seem to have RSS/ATOM feeds, e-mail notifications, etc. to let users know when conversations they care about are updated. In my opinion, however, this will almost certainly be remedied in the near future.

Google Wave is shaping up as a compelling and potent collaboration tool that promises to boost productivity, help increase levels of integration with existing IT systems, and capture/share greater amounts of institutional knowledge. So too, however, can most of the off-the-shelf Web 2.0 tools that most organizations already have today. One of the biggest differentiators is really the rich 3rd party market for high-value extensions that's clearly beginning to form around Google Wave, creating considerable potential to realize the vision for enterprise mashups. That and it's darn enjoyable to use after the initial learning curve.

The bottom line: If one can live with the potential for Google lock-in, Google Wave will be a compelling option for many organizations looking at adopting Enterprise 2.0.

Note that Google Wave is still only available as a developer preview and will start opening up to the public in late September, 2009.

Will you be using Google Wave when it becomes available? Why or why not?