Oracle's bad behavior continues

Oracle is in hot water again over misleading advertising but is that all there is to it?
Written by Dennis Howlett, Contributor

Alex Williams recently had a crack at Oracle saying:

Oracle is stopping a marketing campaign attacking IBM in wake of a national advertising board‘s recomendation that the company made false claims when comparing its Exadata technology to competing IBM products.

It’s the second time in four months the National Advertising Division (NAD)  has taken action against Oracle for making false claims when comparing its Exadata products to IBM’s technology. The NAD is an advertising industry group that falls under the umbrella of the Better Business Bureau.

That was bad enough but as he opined on the topic Williams closed out with:

Okay. Yes, this is Oracle jockeying to some extent in its fight against IBM in the database market. It is part of the game. But Oracle has such a lousy reputation. It uses the language to make claims that are simply not true. I guess that’s the extent of it. When listening to what Oracle has to say, you always need to do a language check. Yes means no. Cloud means on-premise.  
You get my point.

Ouch. He now joins an increasingly raucous crowd of commentators (myself included) who are pointing out how Oracle routinely bends or embellishes the truth to suit its own marketing effort. What really happened?

To the casual eye, the advertisement appears to imply that Oracle Exadata runs 20x faster than IBM Power series systems. What the advert actually referred to was an anonymous 'giant European retailer.' The NAD concluded that despite Oracle's representations, the advert was sufficiently misleading to advise a takedown. It said:

In this case, NAD concluded that while the advertiser may have intended to convey the message that in one case study a particular Exadata system was up to 20 times faster when performing two particular functions than a particular IBM Power system, Oracle’s general references to “Exadata” and “IBM Power,” along with the bold unqualified headline “Exadata 20x Faster Replaces IBM Again,” conveyed a much broader message.

Oracle has been engaged in a war of words with IBM over performance claims for as long as I can remember. Airports around the world display adverts where Oracle claims to be number one in just about any market you can imagine. It almost never provides direct proof statements and even when it has, they have often been disputed. In fairness, they are not alone. 

Vendors of all kinds routinely make claims that on their face sound credible but take a different turn when examined in detail. In this case, the matter gets curiouser. 

In the NAD document, reference is made to the fact Oracle included a link to Oracle.com/EuroRetailer in the advert. This might lead you to conclude there must be a case study somewhere. Try the link above. This is what you see:


Slashdot claims the referring link was torn down. I'm struggling with this. Oracle said in a statement that it plans to appeal the decision although it has complied with the NAD request. That's fine. But surely if you are going to appeal something then you would go all out to maintain the case study and its underlying claims? If not then you open the door to skeptics (like me) who wonder what's really going on. 

I spoke with Debra Lilley, who is president Oracle UK User Group about the matter. (Disclosure: Debra and I are both advisory board members to Constellation Research.) She said that even though you always have to examine such claims carefully, it would be 'highly unusual' for Oracle to make such claims without there being a supporting case study. I asked if she knew the name of the retailer. She said no. Google search doesn't help. I am betting that anyone who does know is under NDA, making it almost impossible to test the veracity of the claims made in the advert. This is so often the case. Ms Lilley also said she would be 'extremely alarmed' if buyers were swayed by these kinds of advertising claim. Sadly that does happen. 

While talking I asked if she had seen a recent video featuring Larry Ellison, CEO Oracle where he repeats that 'lock in is a bad thing.' (see video above) This is a reference to Exadata and Oracle's claim that: vendor lock in is bad, Oracle always gives you choice. In the video, Ellison repeatedly makes the same claim. And it is true. On its face. 

You can run Oracle database on many platforms. But - as it relates to Exadata, the only database certified for those super powerful machines is (drum roll here please) Oracle 11g. This is not therefore lock in but lock out. Again, in talking to Ms Lilley, I asked whether it is true that Exadata outperforms (say) Fujistsu machines running 11g. (Fujitsu is her employer.) The answer was unsurprising: it depends on the workload. 

Update: Merv Adrian of Gartner reminds me that: "Oracle Times Ten on Exalytics also has special features that don't work on any other platform."

And this is where life gets very fuzzy. In recent weeks I have been critiquing a comparative TCO analysis between Exadata/Exalytics and SAP HANA running on a variety of hardware configurations. This has proven extremely difficult. Apart from the 'apples and oranges' problems one usually comes up against, the real difficulty is in establishing which configuration of Oracle engineered systems makes most sense in arriving at a reliable comparision. In some configurations, SAP looks like a clear winner. In other cases, it is neck and neck and in yet other cases Oracle wins. Which one is right? That's still under review. 

Contrary to what anyone will tell you, making these kinds of so called choices is far from simple. 

Now - before slagging off Oracle for having complex pricing models, it is worth referring to Oliver Marks recent post on licensing complexity where he said: 

The reality I've experienced first hand with clients is that Microsoft's aggressive auditing to find additional use revenues, coupled with baffling multiple ski's for products like Sharepoint are fresh in people's minds as we enter a new era of enterprise computing. This legacy of complexity and unanticipated expense is far more corrosive than I think most enterprise vendors realize; while the pace of innovation from inside 'MISO' has definitely picked up, their efforts to move with the times may yet be hamstrung by licensing and maintenance memories. Customer appetites are as much for less expensive, simpler, more transparent and agile agreements and support as they are for innovative more efficient new ways of working.

The same can be said for SAP. I regularly hear complaints about complex pricing that can only be properly understood by those with many years' experience in working through the models. 

In short, none of the big vendors is doing the right thing to make life simpler for buyers or for that matter, making their claims look credible when put under the microscope. They will bleat complexity while claiming simplicity. You can't have it both ways Mr Vendor. 

But as recent evidence shows, Oracle's tactics and marketing need to be viewed with special care. Always take care to listen for what's not said as well as what the carefully manicured expressions are really conveying. There are often disconnects.

In many cases, Oracle may indeed be the right decision. Early results and especially from those in financial services suggest that Exadata really is a great solution so let's not be writing off Ellison's baby as a case of overhyped, overpriced junk simply because the advertising doesn't stack up. If anything, I'd prefer to see Oracle play to its real strengths in a less-combatant manner so that the buyer really does feel that they are being given choice and that the claims can be substantiated. How about more from the Faces of Fusion and less of this kind of mindless baiting?

As always: caveat emptor.

Editorial standards