Oracle announced partnerships last week with competitors Microsoft, Netsuite and, most surprising, Salesforce.com. These announcements serve the same purpose as the Argentine junta's invasion of the Falklands -- they're a riveting diversion from what's going wrong at home.
Oracle is missing its numbers, and I don't know when Oracle last missed three out of seven quarters. One drain on Oracle is its decline in the CRM market. Earlier this month, Gartner declared Salesforce.com, approaching a $4 billion run rate by the end of 2013, the market leader for the first time. Among other things, the new Oracle/Salesforce partnership signals that the CRM battle is over and Oracle has lost it. The next threat is the growing presence of Workday, which has conquered human resources and is seeking to become the Salesforce of online financial applications. And behind Workday is Intacct, ServiceNow, SugarCRM and 100 online software startups, all marching on Fortress Oracle.
Oracle wants to do something, anything, to slow the growth of Workday, which threatens to eat into two more of its revenue streams. Heretofore, Workday's biggest ally has been Salesforce, which has put on stage the large figure of Marc Benioff, brandishing his "No Software" sign while defending the effectiveness of online applications and providing thought leadership about how packaged software is being disrupted. Salesforce itself uses Workday financials and human capital management, or rather it did, until the Oracle partnership was announced June 25. Salesforce will start using Oracle products instead. Count one punch landed on Workday; don't worry, the effect will quickly wear off.
In a press and analyst teleconference Thursday, Ellison and Benioff combined to explain the new partnership. Benioff said the Salesforce cloud infrastructure was built in 1999 on the Oracle database and that will continue be so for next 12 years. Combined with Benioff's exultant tweet that Salesforce is buying 50 Exadata machines ("I'm super excited our new 12 year @oracle partnership includes 50 @Exadata to make our DB for our 1B customer transactions run faster!"), this pact looks more and more like a really big Salesforce purchase order, with a few "partnership" flourishes thrown in.
[ Want to learn more about how Larry Ellison reversed his opposition to cloud computing? See Ellison Threatens Cloud More As Friend Than Foe. ]
There had been rumors that Salesforce was experimenting with open source PostgreSQL, the database system that can match Oracle front-end features, run Oracle applications and take in wholesale transfers of Oracle data. Unlike MySQL, PostgreSQL is fully ANSI SQL compliant and is an open source competitor to a commercial relational system. It would have been an intolerable blow, on the eve of the Oracle 12C launch, if Salesforce converted to PostgreSQL.
If Salesforce wasn't testing PostgreSQL, it should have been. Three years ago, Oracle gave Salesforce an incentive to do so when Ellison went from database supplier to an attacker of Salesforce's ability to maintain its customers' data integrity.
At Oracle OpenWorld in 2010, Ellison warned that Salesforce's multi-tenant approach "co-mingled" its customers' data in one application system. It sounded like Salesforce mixed its customers' data in a shared memory pool; good luck on getting your transactions out. This, of course, is not how multi-tenant applications work. But with customers worried about data privacy, Ellison fastened onto the shared resources of multi-tenancy as the cloud's Achilles heel. Anything could happen with such a "weak security model," he said. In his address the following year, he carried the analogy a step further and said a Salesforce multi-tenant application may be co-mingling your data with that of your competitors on the same server, and you'd be lucky to escape them seeing it.
Salesforce had been a major Oracle customer for 11-12 years at the time these statements were made. At an all-Oracle event, there was little chance of rebuttal. Instead, Ellison used the CEO's keynote to attack the data integrity of one of Oracle's largest customers. As Salesforce's sole database supplier, he was in a position to know about its data architecture. Ellison didn't know -- or chose to ignore -- how multi-tenant applications actually work. They build virtual machine boundaries to wall off each customer's data. The walls can't be breached by application logic, and there's no case history of failure. Perhaps wisely, there was no outcry from Salesforce over these false charges. Nevertheless, it was still unusual to see a major technology vendor attack a major customer's operations.
Under other circumstances, Benioff (or someone) should have stood up for Salesforce CTO Parker Harris and his team. They had built the first multi-tenant, software-as-a-service product successfully used on a mass scale. Was it wise to shrug off an attack from such a prominent industry leader? Salesforce was probably spared the need to respond by Ellison's previous statements that the cloud was "complete gibberish" and an approach for "nitwits." That is, he had already established his reputation as a subjective and reckless critic.
Still, Benioff's previous relationship with Ellison, when Benioff was a young executive at Oracle, came into play as well. In his 2009 book Behind the Cloud, Benioff relayed how Ellison had been his mentor and how he had to be talked into resigning from Oracle -- he didn't want to leave -- to lead Salesforce full time. He named Ellison to the Salesforce board, then was furious and wanted to fire him when Ellison financed online applications supplier Netsuite. That, too, was an unusual move to have a board member launch a potential competitor.
For Benioff, the younger of the two, there appears to be an unrequited, cherished big brother aspect to his relationship with Ellison. For Ellison, there's the always difficult choice, where a young competitor is concerned, between big brother and Big Brother.