No sooner did I hit “publish” on my post about the new “beta” in today’s online application world did I come across this posting on the Google Operating System blog.
“Every new product from Google had a distinctive label: “Beta”. The dictionary definition for “beta” is: “a version of an application that is made available prior to the official release for the purposes of testing.” Google used the term with a more general meaning: an application that it’s almost complete, that may still have bugs, that’s not fully supported and that may be dropped if user don’t like it. Google has a principle: launch early, and “beta” was a term to illustrate that.
Now most Google products dropped the “beta” label and replaced it with “labs” to show they are still an experiment. You can see this change at most products listed at Google Labs. The only products that are still in beta are: Gmail, Google Toolbar 4 and Google Desktop 4.” 
Everyone reset his or her mental dictionaries. In Googleland “beta” now is “labs” - and yes, they mean basically the same thing. After you get your brain around this, ask yourself this question: Does the change from “beta” to “labs” give Google even more room to launch incomplete products for public testing?
The two terms carry different connotations in my mind, but I guess Google marketing can define them however they like. The explanation is interesting because I’ve rarely known a product at the beta stage to still be at risk of totally blowing up (ok, I lie) or being shot. If you have invested in the development of a software product sufficiently to get it packaged and available as a beta, the decision should have been made months prior if was a go/no go.
While the whole beta>labs change is probably a marketing gimmick, there is a larger evolution happening here that warrants a discussion. Google is one of many companies (perhaps the most visible) that are actively changing the way that applications are developed and deployed. There are changes happening to the development model, to the launch strategy, and to the way an application producer interacts with customers. Things are changing more slowly in the enterprise application world vis a vis online apps, so I’ll focus on the online application side.
In honor of Eric Schmidt, I’ll use a comparison between two of his companies.
Novell: During my tenure at Novell and the now defunct Volera I had great visibility into the Novell “machine”. Novell was (and probably still is) a company created “by engineers for engineers”. The company was paying a number of teams of developers lots of money to tinker and invent software and ideas. The flaw was the lack of strategy and focus, but that’s another story. What would happen is a group of developers would come up with something they thought was cool and interesting and compelling. Then there were many hoops to jump through before anyone outside of the company saw it. I saw silos inside the company working on stuff that had no commercial viability or even market need (my opinion, of course), and many good technological advancements that DID have viability never got to market. I think it was all in search of the next Netware or “MSFT killer”…but w/o timelines, requirements or waiting customers. There was no mechanism to test or tweak these potential products with customers without significant further investments and attention from multiple departments.
Google: I suppose there is some irony here in that Google operates like Novell in some ways, but under a totally different paradigm. Enterprise software and applications instead of web based apps and tools that can be marketed and packaged for consumers OR enterprises. Google encourages its developers to work on “pet projects” every week, and has created a culture where blank-canvas idea formation is part of the job. What is fundamentally different is in the execution and approach to bringing these products to market. In Google’s case, they can now get away with launching a mostly complete app as a beta “labs” release and seeing if it sticks. As Google marches toward total world domination (watch for a future post on this), they will continue to launch apps and utilities that may or may not become part of the product suite down the road. You just can’t do that as an ISV in enterprise software (not if you want to keep your installed base at least), and MSFT customers freak out on a regular basis when some component shows up in one release and then disappears or gets buried.
The net-net is a major evolution in how technology companies can get away with bringing new apps or ideas to market and testing them using the public, without having to commit to long term support or new features. If it sticks, awesome. If not, oh well.
I realize my comparison could be construed as apples and oranges as online apps operate under VERY different constraints than enterprise software platforms, but I believe it’s still relevant. Perhaps Eric’s old company can learn from his new company and bring the same paradigm shift to the enterprise? Will the non-committal public-beta approach trickle down in some ways to other software markets?
Sphere: Related Content
