The Great Oracle Open Source Myth of 2006: It's the end of enterprise software as we know it...

There's been lots of discussion about Oracle and the not-so-secret plans for an open source M&A binge. The discussion started when BusinessWeek's Sarah Lacy wrote an article detailing Oracle's plan to acquire three of the better-known open source vendors:

• JBoss -- Open Source Application Server
• Zend -- Creators of PHP
• Sleepycat Software -- Creator of the popular embedded database, Berkeley DB

It didn't take long for Lacy's proclamation to start looking prescient as Oracle formally announced the acquisition of Sleepycat on Valentine's Day. Of the three rumored open source purchases, Sleepycat is the most synergistic with Oracle's already established focus. Recall, Oracle acquired TimesTen and  InnoDB last year.

Vinnie suggests that the key aspect of this acquisition will be how Oracle manages the dual licensing model that Sleepycat has propagated...

In effect, Sleepycat has learned to leverage the open source community for development, testing etc. It shares that "freeness" to development partners whose extensions it can itself harness. But if you want to make the resulting product proprietary (or just do not want to expose the resulting source code) you need to buy the commercial license.

Now imagine if Oracle can learn to do the same. Leverage a robust open source development community. Take advantage of free extensions in its community. Yet count on the privacy needs of its corporate customers to keep selling commercial licenses under the older model.

...while Jeff questions whether SAP is fighting yesterday's battle...

This is the thing that worries me about the acquisitions that Oracle is making in open source software, what if they came to the conclusion that they would never overtake SAP in license revenues and that the only way they could beat us was by not fighting our war but changing it. If that’s the case, maybe what Ellison’s crew is doing is buying their own LAMP stack and replacing the “M” with “O” and bundling in Fusion middleware with it? That would be a game changer and definitely not fighting yesterday’s war. In a recent Credit Suisse conference interview, Ellison spent a great deal of time talking about subscription revenue and open source, which makes me believe he may have come to the conclusion that the only way he can beat SAP in the application business is to do something he believes we won’t be willing to do, namely take away the license component of our economic model.

So now that Oracle has acquired Sleepycat, are the Zend and JBoss acquisitions truly a foregone conclusion?

Of the two, JBoss is the more noteworthy acquisition for two reasons. One, JBoss is directly competitive with Oracle's own proprietary application server and, as a result, there would be strategic value in bringing the product in-house [to protect its proprietary product and perhaps build out a dual licensing model not dissimilar to Sleepycat]. Two, JBoss has been on the block twice before, with BEA and IBM [the other major app server vendors] reportedly passing on the opportunity.

Charles Zedlewski has a comprehensive take on Oracle's open source plans and, while he understands the rationale behind the Sleepycat acquisition and potential sense of buying Zend, he thinks a JBoss acquisition would be "in-freaking-sane."

JBoss is, by objective standards, not a successful software business. Its application server is heavily downloaded, but JBoss never seems to make any money from it. We’ll never get objective figures on this because Mark Fleury’s reality distortion field is even more powerful than Larry’s, but the word is JBoss’ revenues from support contracts for their app server (i.e. revenue tied to product) are somewhere in the low single digit millions. JBoss’ total revenues are somewhere in the double digits, but the rest is just consulting and training revenues, the same as hundreds of other J2EE VAR’s in our industry typically valued at 1.5x sales.

While I think Charles' take is an interesting one, I'm afraid his assumptions about JBoss' revenues are materially understated. Several of my friends in the VC industry independently corroborated that JBoss is running closer to the $50mm revenue run rate. I'm not sure that would necessarily diminish Charles' objections, but it does suggest the proposed takeout price is much more reasonable than he suggests.

The story doesn't end here, as MySQL founder Marten Mickos confirmed that he was approached by Oracle but ultimately passed on becoming part of Ellison's OSS roll-up strategy. While Mickos was unwilling to sell out to Oracle, he was more than happy to grab a few million in funding from SAP.

So what's my take on Oracle's open source plans and  how it fits into their broader strategy?

Oracle needs to define its strategy, to its shareholders, customers and ecosystem partners. I can understand why the industry is worried that Oracle may be willing to slash and burn the traditional software license model; I won't believe that until I see it. Honestly, what's the point of buying Siebel, Peoplesoft, J.D. Edwards, Retek and a dozen other software companies if Oracle is REALLY resigned to commoditizing its core franchise business [i.e., databases]? Ellison, Chuck and Safra have spent the last 13+ months convincing the world that buying up the apps vendors would generate unprecedented levels of cash flow at optimum margins, while allowing Oracle to regain some of the footing its lost to SAP over the last few years. Now we're being asked to "hang in there" while Oracle evangelizes open source at the same time?

The ONLY conclusion I can accept is that Oracle is seeing what IBM is doing via Gluecode, while also understanding that it's already "game over" in the applications market, and has simply decided to ensure it can control the database infrastructure layer without the fear of massive commoditization. Oracle has the capital to make these and many more acquisitions, but at some point, I and other shareholders are going to need to see tangible return on our investment [how are these moves helping increase shareholder value?].

Note: At the time of this writing, I and/or funds I maintain discretionary control over, maintained long equity positions in SAP and ORCL, but did not maintain a position [long or short] in IBM.

sleepycat zend jboss open source oracle ORCL software M&A mySQL woodrow

 

Comments

[Jason Wood]

Hey Charles,

You bring up an excellent point, and one that in a future column I want to elaborate on because it speaks not just to the rumored JBoss transaction but all of the M&A activity we're seeing in software-land these days.

The most common metric being assigned to software takeouts these days is a multiple on future maintenance revenue streams; most notable deals have come off between 4x-10x that metric. But obviously there are so many other variables that go into an M&A price. If a software vendor is still growing fast [or faster than the acquirer], a la Wily Tech [which we both blogged about], you can and should pay a much higher multiple b/c not only is the license revenue super high margin but the lifecycle of future maintenance streams is extended by N years as a result.

[Charles Zedlewski]

Good synopsis and commentary Jason. From a sharholder perspective I imagine I'd want a more cogent story from Oracle too. A few clarifying points.

$50 mil is a little higher than I'd guess JBoss's total revenues at, but it's close to what I figured. My contention is more than 80% of that number is consulting revenue which has poor economics. Accenture is worth $18 billion and has $17 billion in revenues. So in fairness, let's buy JBoss for:

$1 for every $1 in consulting revenue and
$5 for every $1 in maintenance revenue

I think this puts the fair value at $90 million. I'll throw in an extra $20-30 million because it's "open source" and "hot," but we're still a ways away from $400 mil. This all assumes there's legitimate strategic value in having a second app server in the product portfolio, open source or otherwise.