Jonathan Rosenberg, Senior Vice President, Product Management at Google, believes "open systems win," according to this blog post "The Meaning of Open."
Of course, Rosenberg can believe anything he wants. But on the subject of enterprise software, openness certainly hasn't won in the past and is not winning today, 50 years after IT marketeers started throwing around the term "open" to try to break the IBM monopoly. (Rosenberg also believes open will "make the world a better place," but that's food for another type of blog, some "new age" thing with angels and flower petals.)
In 2009, the five largest enterprise software suppliers as measured by worldwide revenue all have based their past and current leadership on what most would call "closed" technologies. Rosenberg also admits there is a big debate, even within Google, about the meaning of open - and therefore of the meaning of closed - but for purposes of brevity, I'll stick with the conventional uses of the two terms.
The five leaders (Google, IBM, Microsoft, Oracle and SAP, in alphabetical order, and dependent on how you measure worldwide enterprise software revenue) have completely built their successes on closed enterprise software. Each grew to its leadership position in under 10 years of its founding because of the way you voted for it with your pocketbooks.
Two of the five - Microsoft and SAP - still have relatively little to say from a market-dynamics perspective about openness. The two are involved in various standards organizations and use openness, as they should, to reduce their research and development (R&D) budgets and to market commodity software functionality.
Two of the five - IBM and Oracle - actually market, as they should, a newly discovered fondness for openness, to various degrees. IBM does it because it is metamorphing into a management-services provider and openness is a key ingredient in holding down such professional services' delivery costs. Strategically, open standards are very important to IBM. Oracle will be as open as it has to be to keep the acquisition train on track. However, the concept of openness - no matter how you define it - has had no more to do with IBM's and Oracle's current successes than it had in their past successes.
The fifth leading enterprise software supplier, Google, has had no past successes. So far, it's just the one current runaway, extremely impressive closed Stanford-search-algorithm success. I've called Google a one-trick pony in the past but one-trick hippo would probably be more appropriate.
Google makes my top-five enterprise software listing because I believe, based on enterprise-software market dynamics research, that Google basically and primarily provides marketing and publishing software to enterprises. The fact that it monetizes that software via advertising revenue is incidental.
But if you think of Google as a consumer software provider or an advertising/media company instead, Google does not even qualify for my list. In that case, another closed technology provider you favor, Symantec, would move into fifth place. Only a few dozen other companies get over a billion dollars a year from you for software and software-related services. These less than 50 software suppliers account for more than 60 percent of software budget you spend. All would be considered closed in the same varying degrees applied to the leaders above.
So the successful history of the closed technologists - and lack of any winning open software company proof point for Mr. Rosenberg - continues right down the list. It is especially telling that most of these multiple dozen leaders rose quickly through the ranks of the thousands of software suppliers past and present while so-called open software companies grow slowly over decades if at all (remember I am using conventional wisdom definitions for brevity; I use different definitions in my research). More likely so-called open companies are designed by venture capitalists from the beginning to get acquired.
Maybe Rosenberg should not have defined "open." When he says "open systems win," maybe he should define the meaning of "win" instead.