Sun/IBM/Oracle

Just when it made sense for IBM to withdraw its offer to acquire Sun, now Oracle is repeating the mistake (or maybe not?). With Java and Solaris being open source, Sun is going towards a church based model. Given the separation of the church and the state, no one has made money buying out the church, has it? Given the industry-wide open source acquisition frenzy such as: SUSE, Jboss, XenSource, Zimbra.  sure..you get a good customer base and other religious followers, but how many of such acquisitions were actually profitable (see table below). The Anti-Trust needs to put in a First Amendment-like clause in its takeover rules else the entire value proposition of an open-source company becomes moot. The founder of open-source MySql quit Sun after its acquisition (a billion dollar deal) and founded yet another open-source database company and his manifesto here concurs with the open-source mentality I voice above.

Good were the days when Sun had smartly positioned itself as a hardware focused unix OEM. Leave it to Sun to internally destroy value. It acquired NetDynamics back in 1998 to enter the burgeoning middleware space and gained access to a quality customer base. It then boasted of a true end-to-end solution with its enterprise servers all the way to the client side security certificates (not to mention its relatively decent directory and mail server). It then acquired most of Netscape’s products (via its song-and-dance alliance with AOL) and boasted of an even superior app-server (Netscape’s app server was a hardcore C++ based product they acquired from Kiva). And just then they dismantle every thing with the excuse of merging the two products together and come up with a lower grade iPlanet App Server. By then, both IBM and BEA (now Oracle) surpassed Sun with their respective products and customer base leaving Sun with its hardware server fan following. Of course in parallel, Sun’s management did a bad job as well. Ed Zander, at the helm, pretty much wrecked up Sun and probably went on to do the same at Motorola. (I worked at Sun long enough to render the above philippic)

So it did not make sense for IBM to acquire Sun. Why? one word: Overlap. Sure, by all means, both deals (Oracle and IBM) would have been/will be financially disciplined – it will give the acquirer an opportunity to create value for its shareholders – and the deals would start to become accretive soon (Oracle’s boasting the deal is accretive from the first year on). Not a biggie. But did it make strategic sense for IBM to buy Sun? Surely not. In fact, it would be interestesting to find out whether the deal with IBM would get past the anti-trust authorities as it would make MergeCo (IBM/Sun) account for 42% of the total server market revenue. And with Oracle? Only time will tell? For starters, MySql will be dead for sure. The fact that Oracle has been mothering Linux gives an idea on what is going to happen to Solaris. Forte (now Netbeans, acquired by Sun for lack of an IDE) will surely not be Oracle’s forte. Only time will tell which products get to stay and which do not.

<snip – added a few days later>

One viewpoint for why the Oracle deal will make sense is under management arbitrage. Maybe the market thinks that the Oracle’s management is better equipped to handle Sun’s products. This is also evident by a quick re-org at Microsoft a few days after this Oracle announcement. Microsoft shuffled its Server Tools & Biz group to position most of its tools against SQLServer; it possibly perceives the merger as a threat to itself with Oracle having a full-on stacked solution vis-a-vis its Oracle database !

<end of snip>

Here are some deals in the enterprise software space during the last 4 yrs just to give you a perspective on deal size and success/failures. Use hindsight to judge which deal panned out favorably for the acquirer. (These numbers were handpicked and researched by me during the microsoft/yahoo deal. So take it for what its worth.)