Larry Ellison was at it again yesterday, making friends, influencing people and pissing off rivals. It was AWS in the keynote earlier in the week. Yesterday, it was Splunk, a seemingly innocuous logging software company, which somehow fell into Ellison's marketing cross-hairs. The company took serious exception.
Splunk is best known for logging all events related to IT. Ellison announced a new product -- a unified platform for SecOps, DevOps, and IT Ops -- or precisely Splunk's sweet spot. If you are the market leader in an area where Oracle wants a piece of the pie, you are going to feel the wrath of Ellison and that's what happened to Splunk.
Oracle didn't exactly try to hide they were going straight after Splunk with their new unified logging product. In fact, senior director of product marketing, Sridhar Karnam penned a blog post called Oracle's Larry Ellison takes on Splunk to announce the product on the company website.
But that was really the encore. The real show was at Ellison's announcement where he took aim at Splunk with his typical style, making great claims about what his product could do, and of course what his rival was incapable of.
Splunk fired back in an angry blog post you don't usually see from a software company, which runs behind the scenes and generally doesn't call attention to itself except when there are products to be announced. Splunk had a conference of its own last week, but there wasn't any name calling or corporate pissing matches that I heard about. This is the second such incident for Ellison this week, who thrives on controversy to get his company attention (and it's worked).
The company seemed quite flabbergasted by Ellison's attacks, yet felt compelled to answer his claims all the same. When Ellison said, " Splunk has no real entity model and leaves data in many disparate vendor silos," Splunk responded, "Splunk turns data into answers, applying schema on read to give structure to the data when you ask the question and not force entities when you write it (presumably to an Oracle database which is... again… convenient)," they wrote. What they are saying is Ellison was being misleading when he made that statement.
Then he took a swipe at their machine learning capabilities saying, "Splunk provides a machine learning toolkit that requires data scientists." To which Splunk replied, "Just wrong. We make machine data accessible, usable and valuable to everyone and we’re doing the same with machine learning." To be fair, he's not completely wrong, but again, he is being misleading.
When I spoke to the company just recently about its machine learning capabilities, they cited two paths. One was a toolkit that would allow that data scientist to build apps on top of the Splunk data, but the other was designed to provide machine learning capabilities in a more automated fashion for those companies who didn't have data scientists. So they offer both. Ellison just chose to focus on the path that made his company's product look better.
Just another day at Oracle OpenWorld with Larry Ellison making wild claims and his competitors taking exception. It's only Day Three. It feels like longer.