John Wookey, who headed up Oracle’s Fusion project, has moved on. I have good news and bad news on that for Fusion fans. The good news is that I doubt it changes much with respect to Fusion. The bad news is it is an admission that Fusion is badly off track. I’ve been talking with various Oracle folks at various levels for some time, and this news has been obvious for quite a while.
It takes a very long time to rearchitect enterprise software. SAP’s ByDesign was in the labs for five or six years, it is only recently available to “early customers“, and the target profile it is suited to is businesses smaller than SAP’s current power alley due to various limitations such as 4x slower performance. Fusion has a slightly easier job as it “only” amounts to an SOA integration of all the components in their suite. The rewrites, at least, need not be all the way to core. Nevertheless, they are chasing a moving target in the form of an ongoing acquisition strategy. They’re also having to retrofit a variety of different architectures with this capability, and they have to do that in the wake of having scaled back the various acquired entities to enhance profitability. Can you guess what the pecking order is between financial results, customer satisfaction, and a new technical SOA architecture? If you’re like me, you’ll see that Fusion is likely dead last in that pecking order. What this means is that Wookey may have moved on from a nearly impossible job.
As Vinnie Marchandani put it when he broke the story, Fusion in 2008 was always a moonshot. He says 2009 earliest, and I would not be surprised to see 2010. Oracle has a difficult set of numbers to keep going with, ongoing merger digestion, and frankly, great difficulty investing and mobilizing around software innovations of the kind Fusion represents. They traded in their innovation hats when they brought in a couple of M&A experts in the form of Chuck Phillipps from Morgan Stanley and Safra Catz to run the place under Larry Ellison.
Anyone who has worked at Oracle, as I have, will know that projects often proceed with considerable investment all the way to beta test and then get shunted into a new direction or cancelled entirely. I’m not suggesting that Fusion will be cancelled, rather, I’m suggesting there is a lot going on below the water level on the Oracle iceberg that’s completely invisible. Folks I talked to suggest that Fusion is isolated, at least until pretty recently, and that part of the problem is all the other products are moving on without it. If that’s true, and it’s not out of character for the culture, Oracle can demo Fusion all they like, but the lion’s share of the work will be below the surface getting these apps properly configured to talk to it. That’s a huge undertaking. To make matters worse, they’re targeting BEA now, which is theoretically a competitor to Fusion. Rationalizing that against work already done on Fusion is apt to set the schedule back further and could even prompt a restart. And, all that effort can’t start until BEA is acquired, a fate they seem to be resisting, at least for the moment.
This is not the best news for Oracle. SAP has their new toy out on the table. SaaS and Open Source are coming up from below. If the mission continues to be growth through acquisition, as it seems it must, Oracle will have to keep stepping up the pace there too. Meanwhile customers are going to see some reasons to wonder about buying anything new from Oracle. They’ll likely pay their maintenance and finish ongoing projects, but why add modules with such an uncertain future? At the least they should look out in the market and wonder whether a better module from a competitor today won’t wind up in Oracle’s stable tomorrow.
If the financial markets see a hint that Oracle’s installed base and businesses are cooling, it will not be a happy time for Oracle shareholders or management. Fortunately, that has yet to happen. Still, it makes me nervous when a software company ceases to be a software company and becomes an M&A aggregator. Those games eventually come to an end. We’ve seen it before several times in the industry with players like CA.