Analysis Open source and subscription-based pricing are taxing the best brains of software providers. Sun Microsystems became the latest to grapple with the issues this week, announcing it would no longer charge for its software.
Lucky developers can now download Sun's Java Enterprise System (JES) middleware stack, N1 grid engine and management software, C, C++ and Fortran tools and the SeeBeyond development and integration suite for free. Just don't expect to get any services or support, 'cos that's gonna cost you.
Sun's chief operating officer and prez Jonathan Schwartz validated the decision by citing JBoss and MySQL as examples of where the software market is headed.
JBoss's chief executive Marc Fleury returned the compliment, telling The Register Schwartz had made a "ballsy, bold and visionary" move that endorses services. Or was Fleury damning with faint praise?
Sun's business - and its software business in particular - has been on the comeback trail for years. After being treated as the illegitimate child of the product family, software has enjoyed a renaissance under Schwartz.
Sharing? How about "Revenue" as the new slogan
After three years - approximately the length of time Sun has been talking about subscription pricing and open sourcing its software to create a volume market among developers - you'd expect the green shoots of recovery to start popping up. And they are, only Sun is hell bent on ripping them up and planting new seeds.
Sun has had a tough fight convincing both skeptical Wall St analysts eager for profits and customers married to traditional notions of paying for software about the merits of per-user pricing on its JES and Java Desktop System (JDS). Proof was emerging that this market-changing business model was getting traction, with Sun having built a fledgling $100m revenue stream founded on JES.
Sun, though, is denting these hard-earned gains by forcing customers to reevaluate another new pricing model. In addition, Sun is throwing out $167m in annual revenue that it inherited this summer with the $387m SeeBeyond Technologies acquisition. We're willing to be that Sun will "change the software game" at least three or four more times before it has $387m in software services revenue to make up for the SeeBeyond buy..
The SeeBeyond aspect, in particular, points to a strategy that owes more to Sun's ideology of "sharing" than business sense.
Application tools vendors like IBM/Rational and Borland are taking pragmatic approaches to turning open source to their advantage. Invariably they are "giving away" or "opening" the basic, commodity features in their tools suites while holding back the real meat, or value-add features, that they - correctly - figure they can charge for.
Sun, though, is going balls out by giving it all away.
Sun also believes it can create a mass market for integration tools by "giving away" SeeBeyond. The company, should learn from Microsoft's example.
Last month, Microsoft launched Visual Studio 2005 Team System (VSTS), a suite that delivers a range of features for application development, modeling, testing and build. While Microsoft's vision is to "democratize" the application lifecycle management (ALM) market by offering low pricing and products that are relatively easy to use, it knows that not all of the VSTS features like application design and modeling will be used by its community of seven million plus developers. It does though, expect, overlap between development and testing.
Oh, and the price for democratization? Visual Studio starts at $49.
There are two key weakness in Sun's strategy. These are a belief that it can create a mass market that it can leverage with services, and - apparently - that it can create this market quickly.
Simply giving away software will not in itself create a mass market - and clearly not in the time frame Sun is apparently working on. Despite the decision to "give away" its rather poorly architected application server two years ago, for example, Sun's market share has declined. It has also declined despite a crafty rebranding exercise that turned Sun's application server into the reference implementation for all Java application servers.
Fleury, whose open source application server and middleware company cheerfully butts heads with IBM and BEA - numbers one and two in market share according to IDC - is blunt. "In the enterprise, Sun are still nowhere to be seen. They are not on our radar. Their software is not the best on the JES side."
Market building takes time and is predicated on the fact you have something that people actually find useful. JBoss has enjoyed success because it undercut BEA as a platform for developers while MySQL is approaching 50 per cent market share because it offers the same value proposition against IBM's DB2, Oracle and Microsoft's SQL Server.
It also took JBoss six years and MySQL 10 years to reach 39 per cent and 45 per cent market share, respectively.
So, no market share then. But that's ok if Sun has the services ready to go - all it needs to do is swap out product licensing and subscriptions for services, right? Problem is, and here's where Wall St will be pissed, it's unclear what services Sun will slip-in as alternatives. Sun promised services would be rolled out during the next year but also said some services are available internally. So which is it and what are they?
Also, Sun should be wary of betting its business on the siren song that is services. Even JBoss has had difficulty converting those who download its software into paying subscribers - BusinessWeek earlier this year reported just five percent of JBoss users are subscribers. Software, Sun seems keen to forget, is where the "easy" money is - and will be for some time. IBM made an 80 per cent margin on its software in 2004 compared to just 30 percent on hardware.
Sun, forever the come-back kid, is a giving kinda company. It likes to give things away. The challenge is converting the giving into something that profits the giver. By trading software licensing and subscriptions for set of unnamed services, Sun has re-set the clock on recovery and raised questions over its strategy.®
None of this is helped by the arrogance with which Sun puts out these "game changing" pricing announcements. Management attacks the press and customers with the same, tired pitches about Sun being ahead of the competition and revolutionary. It, however, always dodges the revenue questions like a stumbling politician or like a company with "share" as its slogan.
At the launch for Sun's new pricing plan, for example, President Schwartz doled out these gems.
"The challenge is to monetize the volume that is out there."
"This is not a way to get revenue to go down. This is a way to get the barriers to revenue to go down."
"The obstacles in front of us are not technical they are cultural."
"I don't see the application server as a business anymore."
The last item is especially telling. Schwartz seems determined not to find revenue in places where plenty of revenue still exists. Why is a company that is trying to make a comeback so set on sharing instead of making money?
Sun would be better served by using a more humble tone to launch these pricing experiments. It's clear that Sun's main goal is to wrap servers with items competitors can't match. Putting the pitch in these terms would help instead of pretending that the company is a revenue-producing software powerhouse that can sway the market as it pleases. Hardly the case.