Investment figures for the Java platform

Question

Are there any market research figures available calculating aggregate investment (worldwide, by region, by country) in the Java platform? This could include software (application servers, IDEs, profiling tools), hardware (Sun servers, bytecode executing processors), personnel (training, books), language development (Java, Clojure, Scala, ..), public/private (universities, governments, corporations), etc.

I'm working on a presentation on JVM languages and would like to illustrate—in some reliable manner—that existing investment in the Java platform is massive and not going anywhere anytime soon (think Cobol). I'm sure this point is clear for IT guys, but I'd like to put a dollar figure on it for the business folks to really drive the point home.

Before the comments start, I recognize that this question is on the outer edge of what most of us consider to be "programming related". Regardless, knowing this information can help many IT department discussions around the future of a platform, and may even assist developers make career decisions.

Answer 1

Oracle (previously Sun) would obviously be the first place to start. They'll be happy to evangelise the benefits of Java to you! In fact, that's how they see themselves:

Sun VP Rich Green has been quoted on countless occasions that Sun's ideal role with regards to Java is as an "evangelist." [O'Reilly Media, 2002]

I recommend asking them for those figures. If they exist, Oracle will have them.

Answer 2

Market research on this kind of topic is notoriously unreliable - mainly because nobody collects the data at a granular level so there is a lot of guesswork / extrapolation involved. Remember that money spent on the salaries of every coder and project manager in a bank working on Java applications is technically "investment in the Java platofrm" even though it goes nowhere near the books of Oracle or IBM or any other tech company.

I'd suggest trying to estimate the investment / level of activity from a number of different directions, and triangulate these to come to an informed view

Here's some approaches that you could use:

• Use something like the TIOBE index as a rough estimate of programming language share and multiply it by your best estimate of Global IT software development spending.
• Add up the revenues of Java-aligned tech companies (Oracle, IBM, Google, etc...), weighted by whatever proportion of their activity you guess to be related to Java-based products. Then multiply by 1/(1-X) where X is the percentage of software investment you estimate to be done in-house or with smaller companies not on your list
• Look at the growth rates and activity in new JVM languages like Clojure, Scala, Jython, JRuby etc. You can easily make the argument that these are ensuring an innovative future on the Java platform.