OpenWorld was kicked off Sunday with the customary CEO presentation. Larry Ellison, however, is not in the habit of doing anything in a customary way. For the fourth year in a row the CEO chose to focus on only one facet Oracle’s complex business –hardware/software appliances, now called “engineered systems”. Oracle’s software offerings were only mentioned in passing as examples of things that run well on Oracle hardware. Late in the presentation, an attendee behind me in the audience asked his associate if Oracle was still in the business of selling database software – something not apparent from anything the CEO said.

Database was not the only notable thing ignored. No reference was made to Fusion Applications at all. This could be an indication that progress continues to be slow on this front but it is dangerous to try to read too much into what Ellison chooses to talk about.

When Ellison first walked on stage, did a double take and commented “we have a lot of hardware up here!” He then spent an hour filling our heads with technical details, specs, and comparisons versus his hardware nemesis IBM. As usual, it was articulate, passionate and sounded very convincing. I am not enough of an expert on hardware to judge this facet of Oracle’s business, but Ellison sure can make it sound impressive. At the same time, IBM keeps sending me email messages bragging about how it is gaining market share in hardware from Oracle. Where the truth lies remains a mystery to me, but it is no mystery what Oracle management currently cares the most about. Winning the hardware battle and proving that acquiring Sun was a good idea is clearly the top priority.

As someone concerned about applications more than the hardware they run on, I found Ellison’s presentation very disappointing. A tiny bit of love sent in the direction of the application community would have been nice. I also find it strange that Ellison now pays so little attention to the parts of his business that make most of the revenue and profit.

With Ellison’s considerable help, computer hardware is moving faster toward becoming a low-cost, low-margin commodity. Competition from quality vendors has become increasingly fierce. Oracle, IBM, HP, Dell, EMC, Teradata and many others are racing to see how quickly hardware can be created that performs nearly perfectly and costs almost nothing. The movement toward cloud computing threatens to further speed up the pace of that race.

The hardware war is great for those worried about applications. Each year we can afford to be a little less concerned about hardware cost relative to creating the solutions that run on it. The arcane details of hardware design are generally only of concern to those job entails selecting and running it. I have to believe that 90% of the attendees at OpenWorld have little to do with hardware selection and that all the energy being spent here hyping Oracle appliances (sorry, engineered systems) is lost on the rest of us.


OpenWorld begins this coming Sunday and, as usual, I will be there from start to finish. Oracle never telegraphs what is coming, especially what CEO Larry Ellison plans to address in his annual presentation. Recent years have seen Ellison personally focus on database appliances, especially since the Sun acquisition and that will almost certainly continue to be the case. IBM has recently been bragging about its gains in market share versus Oracle in hardware so I expect a vigorous and entertaining counter-attack.

It is a relatively safe bet that Fusion Applications will be formally introduced in some way this week. If not, it will mean that something has gone terribly wrong. Every indication is that the opposite is true – pilot site testing seems to be proceeding well.

Principal Financial has already been named as the first large scale user of Fusion Applications so I expect that it will provide a compelling customer testimonial. Others that have tested portions of the application suite can also be expected to get lots of airtime. In the last few years, as the delivery date for Fusion Applications continued to slip, Oracle has taken a low key approach by refusing to make grand claims about the product line beyond assuring us that it is in the works, is going to some day be wonderful and that it will provide a smooth upgrade path for Oracle application customers.

A few weeks ago I talked to people from a company that is not yet using Fusion Applications but that has agreed to become an early adopter of the Payroll module later this year. They indicated that Oracle is making it incredibly attractive to become an early adopter for a group of carefully selected organizations. It will be interesting to see if the early adopters that are not yet using the software will be called out in any way in advance. I suspect not, but Oracle surprises me at every turn.

Not surprisingly, Oracle is trying to get early adaptors to run Fusion Applications on the Oracle Exadata and Exalogic appliances. Apparently the current list pricing is very attractive and obviously there will be lots of discounting. The servers being advocated for use with Fusion Applications offer a great deal of flash memory, often multiple terabytes. It will be interesting to see if Fusion Applications have been optimized to run on Exadata and if it will become normal for them to be sold together.

Fusion Applications will have to run on all the popular servers but the cost of running them elsewhere could turn out to be high. Complete adoption of Oracle’s “red stack” of hardware, database and middleware may be a de facto requirement for those that move to Fusion Applications. How Oracle attempts to both maintain its reputation for openness while also making the use of all of its products together compelling will be one of the more interesting aspects of the Fusion Application introduction.

With only a few days to go before the curtain is at long last lifted on Fusion Applications I can’t wait.


Oracle CEO Larry Ellison is famously elusive.  It is harder to get a private meeting with him than the Pope.  Outsiders, especially reporters and analysts, have almost no chance of ever meeting him.  Only the inner circle within Oracle itself ever has a real two-way dialog with Ellison. 

There are, however, two proven ways to meet Larry Ellison in person: spend tens of millions on an Americas Cup racing yacht; or show serious interest in buying an Exadata or Exalogic server.  For most of you the latter represents your best shot. 

A database expert at one of the top-end consulting firms confirmed to me recently that Ellison and President Mark Hurd continue to make personal sales calls on important hardware prospects.  It appears as if Ellison’s highest priority as CEO continues to be to promote the hardware appliances that emerged out of the Sun acquisition.  If that means sending the boss out on the road, then so be it. 

Acquiring Sun was one of Ellison’s boldest and riskiest moves.  Even though the new Exadata/Exalogic servers that emerged from Sun’s wreckage have been a huge successes, Oracle’s share of the hardware market continues to decline.  Some of the loss has gone to one time ally HP, but the larger part of it has ended up with IBM. 

At OpenWorld 2010 Ellison relentlessly mocked IBM and quoted numerous benchmarks to support his assertion about the superior performance of his to-of-the-line appliances.  To the best of my knowledge his characterization of them was accurate.  Like almost everything Oracle does, these new products are technological wonders with very impressive specifications. 

Across the whole hardware product line IBM appears to still have an advantage.  In addition to its formidable reputation as a hardware vendor, IBM has historically built customer loyalty by providing great service and support.  In a recent letter to partners IBM bragged that it now has 37% of the total hardware market and that it is the leading source of hardware to run Oracle software with a 20% share. 

A large JDE customer I talked to recently is planning to be an early adopter of a few of the Fusion application modules.   IBM has been its preferred hardware vendor for decades but Oracle is making a full-court press to get a foot in the door with Exadata.  The customer is highly impressed with the specs for the Oracle hardware and is taking great delight in having two giant vendors fight for its business. 

The war over hardware raging between Oracle and IBM (with HP often in the mix as well) is great for our community.  Prices are dropping at a faster than usual rate and cool new capabilities such as huge flash memories at bargain prices are becoming available every few months. 

You may not get a personal visit from Larry Ellison out of it, but many of you will want to at least kick the tires on Exadata.  At the very least, you might get a better deal from IBM the next time you need more hardware capacity.

Quest’s annual Northeast conference, held last week at the Mohegan Sun in Connecticut, is always a great source of news, information and gossip.  This year was no exception.  There was no earth-shaking news about either JD Edwards or PeopleSoft, but the era when dramatic things happen to either of them is long past.  Instead, executives from each of these product lines came and made the usual passionate plea to keep upgrading to the newer releases.  In that sense, nothing has changed since Collaborate in April where Lyle Ekdahl offered the JDE faithful three messages – Upgrade, upgrade, and upgrade! 

The PeopleSoft keynote did provide some additional insight into why Oracle seems to care so deeply about which release of its products you are using.  The 5,000+ PeopleSoft customers contribute $1.3 billion per year to Oracle’s revenues (out of $34 billion).  Over a billion of that comes from maintenance.  The rest comes largely from sales of additional modules to existing customers and from sales to 200 – 300 new customers. 

Oracle does not divulge profit by product line, but it is obviously spending far less than $1.3 billion on PeopleSoft.  The difference falls straight to the bottom line.  The numbers for JDE are roughly in the same range.  The JDE customer base is over 6,000, revenues are under $1 billion (but not far), and a few hundred new accounts offset occasional losses.  Between them, JDE and PeopleSoft appear to be contributing close to a billion dollars per year to Oracle’s margins.  Any speculation that Oracle does not care about our community is thus highly uninformed – we are a gift that keeps on giving. 

The party will continue for Oracle as long as it can keep us paying maintenance.  The way to do that is to get us to care about the future improvements that we get for our maintenance money.  In the near term, we get most of that value through upgrades.  Oracle tries to balance the need to provide us with improved capability on a regular basis with the requirement to limit how often we have to go through the upgrade process. 

With all this in mind, Oracle has settled on a schedule of doing a major release for JDE and PeopleSoft approximately every three years with more modest “feature pack” releases every year.  Tools releases also come once per year.  Vital problem fixes and things like tax table updates are done as needed. 

As time passes, a much greater incentive for continuing to pay maintenance will be the ability to upgrade to the equivalent Fusion application module at no cost.  After a very long wait, Fusion Applications will finally be formally announced at OpenWorld in early October baring a disaster during pilot testing between now and then. 

I was able to pick up some additional intelligence about how things are going with Fusion Applications at the conference as well which I will share in my next posting.

Oracle proclaims itself to be the leader in the Enterprise Performance Management (EPM) market – an assertion that is hard to dispute given the imprecise way in which that acronym is used. Like all of its competitors, Oracle has its own unique vocabulary and associated definitions of the jargon and acronyms it uses. For those trying to keep score at home, we will take a stab at making sense of what Oracle is trying to tell us.

A brief history lesson is needed to understand how Oracle arrived at its current usage of the EPM term. The Oracle acquisition of Siebel brought along two important Business Intelligence products: a suite of data analysis tools that Oracle gave the tongue twisting name Oracle Business Intelligence Suite Enterprise Edition (OBIEE); and a collection of analytic applications now called Oracle Business Intelligence Applications or OBIA.

These offerings transformed Oracle from a fringe player in BI into a possible contender. The Siebel offerings helped fill out a limited lineup of BI related data management tools that were little more than add-ons to Oracle’s 11g DBMS.

The acquisition of Hyperion pushed Oracle into the upper tier of BI vendors. It also created the challenge of integrating three separately created product lines and explaining how they fit together. The most important contribution was a suite of financial applications now called Hyperion Performance Applications that includes Strategy Management, Financial Close and Reporting, Planning Budgeting and Forecasting, and Profitability and Cost Management.

Oracle started to use the term Enterprise Performance Management (EPM) as an umbrella name for a now fairly comprehensive BI product line that featured the Hyperion Performance Applications. Unfortunately, in practice, Oracle people have fallen into the habit of using EPM as shorthand when talking about the Hyperion applications.

When EPM is used to mean the Hyperion applications Oracle can correctly claim to be the market leader. On the other hand, when Oracle claims that market leadership in EPM means that it leads the entire BI market, then an element of exaggeration is present.

Gartner is the closest thing we have to a referee when vendors make conflicting claims or use different jargon to describe the same thing. It uses the term Corporate Performance Management (CPM) to categorize the type of applications Hyperion offers and does indeed certify that Oracle is the clear market leader. For example, Hyperion’s flagship offering Financial Close has firmly established itself as the worldwide gold standard for large enterprise consolidation.

To be fair, at the same time Gartner has decided that SAP, with its Business Objects subsidiary, is the overall BI market share leader. Just to make things more confusing, SAP uses the term Enterprise Information Management (EIM) as its umbrella term for BI and the things that surround it including SAP’s own financial performance applications. SAP has understandable decided to use a term that plays to its greatest strength – its suite of data management tools.

If all that was not confusing enough, IBM (the other major BI vendor) prefers to use the simpler term Performance Management. It uses this term in a more narrow way to discourage unfavorable comparisons with the other vendors.

It is thus no wonder that IT decision makers can feel confused when trying to understand and compare the options available to them. All of the vendors make creative use of language, freely invent esoteric new acronyms, and stretch and bend commonly used terms to mean new things.

As you try to decide if your organization needs EPM, CPM, EIM or PM it is useful to know that these terms are slight variations of the same idea. By whatever name you choose to use, you will eventually need a strategy for collecting, organizing, managing and analyzing the massive amounts of data that passes through your organization every day. A number of excellent tools are available to help you do so, but the major vendors that sell them are certain to make it hard to sort out which ones are best for you by each using different terms to describe essentially the same thing.

Some retailers always rush things a bit by starting to sell Christmas items in September. In a similar way, Oracle likes to jump the gun by starting its Fiscal Year 2012 on June 1. This unusual schedule impacts the way Oracle interacts with its customers in a number of important ways. First, all energy in May is focused on customers that appear ready to buy something. Others have to wait for a while to get any attention. The last minute shoppers expect and usually get the largest discounts offered all year.

In June, the Oracle sales teams are often exhausted from the year end push and are either celebrating success, making excuses for missing targets, or are sending out resumes. Those that survive start planning for the next year. They also wait to find out what their new territory will be, their personal sales plan and quotas, and what sales promotions are planned. While all this is going on, contact with customers tends to be limited until July or later.

Those of you in a product evaluation cycle not geared for a May decision need to accept the reality of Oracle’s schedule. You should not feel slighted if your Oracle sales team has not paid much attention to you recently. Be patient, they will be back soon.

The new Oracle year also represents a good time to speculate about what changes the new fiscal year might bring in terms of products, sales campaigns, and messaging. I see few major shifts in direction occurring but can imagine many ways in which FY 2012 will differ from the past. The important changes that I believe are on the way include:

• More appliance offerings. Exadata and Exalogic are sure to be refreshed and expanded, and additional hardware/software combinations have to be in the pipeline.
• Cloud clarification. Ellison’s ad hoc comments about cloud computing at the last Open World were largely an appeal to buy his new Exalogic servers. This year Oracle has to make it clear what it really thinks about the cloud.
• Fusion Applications will at long last become something real this year. What the customers testing it out say will matter far more than the way Oracle describes it.
• Mark Hurd will finally emerge from the shadows at Open World. He will definitely put his own stamp on the event and on Oracle but exactly how is not yet clear.
• Oracle is overdue to acquire something sizeable. The next big thing will almost certainly aim to shore up the Sun franchise.

At the same time, some things won’t change very much in FY 2012:

• Solid financial results. Ellison’s flamboyant style sometimes overshadows his exceptional skill at making whatever he does profitable.
• Public attacks on HP and IBM will continue as long as they keep eroding Oracle’s share of the Unix server market.
• Oracle’s application products will keep relentlessly grinding out routine releases. All the real innovation will come through Fusion.
• Oracle’s low key marketing philosophy will change very slowly, even as Mark Hurd begins to play a more visible role running the company.

Since your Oracle sales team is really busy right now, let me be the first to wish all of you a Happy New Fiscal Year!

One of the more intriguing aspects of the IT industry is that businesses are often both great friends and bitter enemies at the same time. Oracle and IBM provide a great example. Oracle CEO Larry Ellison could hardly have been more vitriolic in his attacks on IBM’s hardware business at OpenWorld last September. Ellison used the introduction of Oracle’s Exalogic Elastic Cloud line of servers to make any number of caustic remarks about the IBM hardware it was developed to go after.

At the same time, the decades old relationship between IBM and Oracle’s JDE business remains as strong and cordial as ever. The latest example is the IBM i Solution Edition for JD Edwards .  Its purpose is to reduce the cost, complexity and risk of installing JDE applications on Power i servers. With it the effort needed for initial installation (or upgrading to a new release) can be reduced dramatically. The need for experts with deep CNC skill has also been reduced for those taking advantage of this program.

Continued cooperation between IBM and JDE is very important to the many customers that continue to run JDE applications using IBM hardware and middleware. Some still worry that the long period of détente will end but I see no signs of that happening. IBM continues to take overall market share from Oracle in the server space in spite of the tremendous success of Exadata and Exalogic at the top end of the market. If that trend continues, expect more barbs from Ellison again next October. Don’t, however, assume that it will derail the still strong IBM/Oracle relationship elsewhere.

The same line of thinking may not apply to the once rock solid relationship between Oracle and HP. Those close to the situation all seem to believe that Oracle’s decision to withdraw support the Intel Itanium family of processors is a direct attack on HP. Most HP servers use Itanium processors and a major portion of sales of them involve running applications over the Oracle database. Oracle’s move could thus badly hurt HP server sales and its HPUX operating system along with it. Oracle’s own servers could then take over some of the market share that HP loses.

Ironically, IBM will almost certainly gain part of whatever share HP loses. IBM servers run mostly on IBM Power processors and not Itanium. An intriguing question thus arises: if Oracle’s Intanium strategy succeeds in taking down HP, can it get away trying the same thing in the future by refusing to support IBM’s Power processor family? Doing so would end Oracle’s ability to claim leadership in openness so I don’t see it happening.