This just in—Larry doesn’t get the cloud and doesn’t want to.  Instead he is pushing Oracle to plow money into new features amidst a process innovation era.

He does get invited to headline Churchhill Club events with the likes of Ed Zander as the conversational curator, but this is exactly why he doesn’t get it.

Success can be the ultimate misinformation.

Some may wonder why I spend years of my life geeking out on innovation studies and industrial history—but Larry’s speech last night was a perfect of the predictive value of some of those models and studies. Here is why.

The Theory Behind His Mistake

He tried to frame cloud computing as lacking any product innovation saying:

“Cloud’s water vapor. It’s not new. Everything’s cloud. It’s nonsense. It’s computers, with memory, on nets running DBMS.”

He then clarified that unlike the hype-swamp of cloud computing, Oracle had real product innovation and engineering chops to create innovate features:

“We’re mischaracterized as a sales and marketing company. Per capita, we spend less on marketing and sales. We’re an engineering company, and engineering reports to me.”

(from @rhm2k, who providing great updates from the event)

Elision further espoused a big interest in solid state storage as a disruptive innovation, again clinging to any sign of a feature/performance driven market.

This is rock solid and factual tough talk from a venerable valley CEO who has been at the helm longer than I’ve been alive. But is he wrong?

He is investing in the wrong innovation curve

In my last post I took a pretty simple position. Cloud computing is part of a groundswell of process innovation in the way we deliver computing, as opposed to the traditional IT history of feature innovation. (See the somewhat vague Chart 1)

Ellison’s cloud rhetoric is powerful because it is factual. There are very few new technical features in cloud computing. It is computers, memory, DBMS all running on standard networks.

A few days ago I said as much myself: “There is nothing amazing about the over the network part of cloud,” and by this I meant that the architecture of web delivered applications is now over a decade old. Its not a net new feature.

Instead what is really interesting about SaaS is the way it incorporates features such as multi-tenancy which at first blush diminish control and customization. This is possible because generally what people are capturing in CRM systems now tends to be similar enough that multi-tenancy and operational/delivery standardization have a higher ROI than custom coding features over a 2 year systems integration project.

Simply expand that logic and you can see why I think Ellison is wrong about cloud computing being… vapor.

His rhetoric aligns him squarely with curve #1, but the value of feature innovations in this market is lower than the value of process and delivery innovations. He vaguely acknowledges the need for ‘process’ and integration based advances, but through an appliance model instead of a true assembled service. His frustration with the state of a familiar feature innovation based market was visible, when the forth richest man in the world suggested his industry was a bad place for entrepreneurs’ to try to make a difference.  If that’s the case why is Oracle spending so much on engineering and R&D?

He no doubt understands all of this in elegant detail.

He was a founding investor after all of the Netsuite SaaS platform for SMBs.  By targeting IBM he is staking out what looks to be a $100B market in highly custom, delivery process inefficient, IT technologies. He lives and breathes those customers and is still watching the massive PO’s come in. He is making a selling segment driven bet against the cloud.

A single Oracle project may be larger in scope than the estimated whole of Amazon’s cloud web services. Ad in an integrated Sun systems stack and mind blowing numbers like 3M DB writes a second and you have a provocative product innovation value proposition. But betting that enterprise IT buyers won’t be influenced by the increasing genius and efficiency of the public web is a yet risky call. Everyday we get more and more of our IT value from public compute resources, be they cloud based or otherwise, and less from silo’ed private and narrowly custom constructs.

More than anything it comes down to a question of growth, and the cloud will grow more in the next 10 years than Oracle will organically. It may be overhyped—but so was the internet in the first place and it hasn’t shrunk from relevance or impact.

Is this stock chart a preview of what’s to come?

Ellison made references to being the next IBM, but if I’m looking to invest do I find IBM’s 10 year stock performance chart that compelling?

I’m betting on the yet to be found nexus of scale, process innovation and developer mindshare to drive the next ten years of growth. I’d like to hear more from the venerable Mr. Ellison about why I’m wrong.