Oracle announced an impressive cloud computing initiative this week; it was very nicely packaged. But the company initially did not even take the cloud seriously, and was very slow to realize that this cloud thing was going to be massively disruptive.
The most disruptive part is that, to play in the cloud, you must aggressively interoperate. IBM jumped on this by buying SoftLayer, which allowed it to scale quickly. Oracle is going to grow organically, and that is a far slower path — which is problematic when you are already well behind the leaders.
Amazon and Google came into the market on the wave of open source and have interoperability built into their founding DNA. Microsoft got dragged kicking and screaming into this market, found that this new concept was better, and now is one of its biggest advocates. And IBM, strangely, given it age, seemed to pivot surprisingly easily, likely due to its culture of adapting to change.
Oracle still largely operates under a sales-driven, “lock-in” model: hook the customers with whatever works, lock them in and then monetize the result. However, that really does not work well with cloud implementations. The workloads need to be mobile, often because they are shared across on-premise resources as well (in hybrid cloud scenarios) and can be sourced from a variety of vendors, including Oracle competitors. Oracle’s “lock-in” strategy would be in direct opposition to making the cloud work, not to mention the fact that Oracle has yet to build the ability to support the effort.
So, what is going on here?
Apple and Oracle
One of the interesting parts of Oracle's recent announcement is a service that would provide an AI back-end for digital assistants. It seems a stretch to think that Amazon, Google or Microsoft (who currently own the segment) would use Oracle’s cloud over their own services. But Apple is a new player in this space, and with its HomePod implementation, it currently isn't competitive. (As a side note, the HomePod has the best sound in a speaker under $1,000 but the thing is dumb as a brick.)
I thought that after its IBM alliance, Apple would hook up with that company and place Watson as its AI back-end. This would allow it to jump ahead of the others easily, given that IBM Watson is the most advanced AI at scale now. But that never happened. And that may have left the door open for Oracle.
Larry Ellison and Steve Jobs were very close, but Tim Cook does not seem to enjoy that same relationship. So, we will see, but this could be interesting.
One other aspect of the Apple/Oracle relationship is that Apple under Jobs was marketing-driven and Oracle under Ellison is sales-driven. These are very similar models and differ dramatically from the product/technology-driven strategies and polices of their peers. IBM, for a time under Louis Gerstner, was also marketing-driven. The difference, broadly, is that product/technology-driven companies fight with products; marketing/sales-driven companies fight with perceptions/ideas.
The world around us seems to be defined by fake news and false perceptions, so the power of a perception-driven strategy should be self-evident.
It is hard to fault Oracle’s execution, particularly lately. However, this time, because the elements of a successful cloud service would be in direct violation of Oracle’s lock-in strategy, the risk of not being able to execute is higher.
Perception vs. Reality
What I think is going to make this worth watching is that this will be a battle between perception and reality. When I say “this,” I mean the fight between the major existing cloud players and Oracle. Every one of them should be able to provide services in line with what Oracle has announced, but none of them have packaged them as well. Since Oracle hasn’t built out its data center plan yet, its ability to do what it is announcing wouldn’t seem to exist either yet — certainly not at the scale of the others, anyway.
But that might not matter if people believe Oracle is better and Oracle can execute before these customers realize Oracle’s current shortcomings.
I would normally end this by saying that, in general — and in this decade in particular — perception trumps reality. But Oracle still must execute. I do not think you can do a lock-in cloud computing service that is competitive given the extremely high interoperability requirements.
We will see, but Apple and Oracle have, over the years, showcased how powerful a perception-first approach can be, so this will be worth watching.
Photo courtesy of Shutterstock.