From gibberish to new product launches: Oracle's Ellison backtracks on cloud focus
Oracle's newly unveiled cloud offerings are intriguing, even if the numbers don't quite add up yet, argues Dennis Howlett
Yesterday, Larry Ellison, Oracle's CEO, pivoted on his past by declaring that: "We made a decision to rebuild all of our applications for the cloud almost seven years ago."
That would then mean his recent and past pronouncements about the cloud as 'gibberish' were....err...gibberish? But seriously, once you get past the inevitable rhetoric and hyperbole that accompanies Ellison's on-stage performance there is something there worth looking at in Oracle's public cloud offerings.
First, let's give credit where it is due.
- Early customer feedback and especially related to the Human Capital Management (HCM) applications has been positive. What's more interesting is that early customers are not opting for Fusion as an on-premise system but one they want to deploy directly into the cloud. Up until today, that has been restricted to Oracle's private cloud offering. The company claims 200 customers but the unknown question is which ones are live? My understanding is that the live customers are relatively low in numbers but Oracle expects to quickly ramp sales to live.
- The HCM applications currently slated on the Oracle public cloud website are all targeted at perceived sweet spots like workforce intelligence that helps predict which key employees are at risk of leaving and then allows managers to make compensation recommendations. The embedding of social components within HCM and CRM applications is a clear winner. Oracle positions this as 'more than Chatter' - a reference to Salesforce.com's social networking capabilities. It makes sense to embed social into business processes because then those same processes can take advantage of the context in which conversations occur.
- The early offering of developer and database assets in the public cloud is a logical thing to do. It is an easy way to attract today's generation of developers who don't think about downloading gigabytes of software but want to get up and running in two clicks. However, it will be interesting to see how Oracle competes with Amazon. Oracle makes a big deal about how you can run Fusion and its PaaS solutions on Oracle's infrastructure or move it to wherever you wish. Part of the cost has to include hosting and here, Amazon continues to aggressively push prices down. Will Oracle follow suit or play ancillary cards like its security model as a way of keeping its customers close to the mothership?
- I was especially pleased (and surprised) to see that Oracle is offering 30-day free trial licenses. IT decision makers would find it hard to make a firm buy decision in that period of time but line of business managers looking for componentised applications have no such problems. Oracle opens the door to both classes of user because even if IT managers are out of the initial buying loop, Oracle claims that IT will not have a problem integrating with what they already have on premises, assuming it is Oracle of course. That comfort factor alone will help prevent some of the infighting that occurs once IT discovers that business managers have been wielding their credit cards for technology assets.
So, at that fairly superficial level, there is plenty to admire about Fusion on the Oracle public cloud. Unfortunately, Ellison spoils it by over egging the story, bending the truth, hiding behind FUD and other entertaining tactics that only serve to confuse.
For example, he claims that Oracle 'will' have the most comprehensive set of enterprise applications in the cloud but then says it already has over 100 cloud apps. Notice the subtle shifting of the gears? But when you go to the public cloud website and sift through what they are offering there are only 13 Fusion CRM apps and nine Fusion HCM apps on display along with what looks like a solitary social network play.
In the Java section there are is little detail while in the Database area, the company talks about: "The Oracle Database Cloud Service includes a set of business productivity applications. All of these applications are easy to use, support mobile devices, and can be provisioned in seconds."
I don't see how this all adds up to 100 applications and none of them are available now. When you sign up for the free access trial, Oracle talks about database and Java availability sometime in the July/August time frame. CRM and HCM will come 'later' but there are no times set for those releases.
Past experience suggests this will be much later but I am prepared to be pleasantly surprised.
Right now, if you sign up you can expect a sales call so this is not about getting free access but about having a sales person schmooze you into making a decision. Even then it is not clear whether all comers will automatically get access because Oracle has said it will ramp up in a controlled manner. My sense is that a controlled rollout is needed because reading what Oracle says about virtualised instances, there is no true multi-tenancy. That means Oracle has to find ways of being super efficient at updating customers. It gets around this by saying it will give customers up to a year to upgrade rather than forcing the upgrade which is what often happens in multi-tenant environments. There's just one problem with that. I don't know of a single case where an upgrade wasn't welcomed by Salesforce.com and Workday customers - the obvious targets for this offering.
The flipside is that Oracle needs to be architected the way it is so that it can offer multiple methods of deployment. Some of my colleagues like that choice. I think it is road to nowhere because the multi-tenant benefits far outweigh multiple deployment methods.Oracle argues this is what customers want, but that only holds good for today.
Ellison's first and much anticipated tweet said: "Oracle's got 100+ enterprise applications live in the #cloud today, SAP's got nothin' but SuccessFactors until 2020." It's hard for me to know where to start with this one. I can't find 100 applications and analysts have been told they will be provided with a list. Why wasn't that list available for public scrutiny at the time of launch? It's such an obvious a question that Oracle should have anticipated it. Second, to claim SAP has 'nothing' is simply not true.
Last year, SAP sold Business ByDesign to more than 1,000 customers. Less than 400 are live but even that is more than Oracle is claiming today on relatively limited Fusion functionality. Then there is pricing. Nothing, nada, zippo. This has been a topic of much chatter (sic) in back channels. From what I understand, Oracle is remaining very tight lipped over this topic. That is unusual because Oracle has for years put its price book into the public domain. However, as any seasoned negotiator will tell you, that's meaningless because the company routinely discounts and bundles up its offerings. The cloud works differently and I don't see how Oracle can remain coy on this topic for very long.
Sooner or later, customers will reveal what they're being asked to pay and we'll all know whether Oracle is turning out competitive offerings or whether it is simply transferring its on premises pricing (or rather maintenance) model to a cloud equivalence.
Overall? Oracle is trying to find a middle road of transitioning from an on-premises vendor to one that can simultaneously play in the cloud space. There is plenty to admire but this is very much a first step which, I suspect, is designed to pave the way for more at Oracle Open World, later in the year.