[ap22213]

First thing I thought when reading the headline was 'so is Oracle'. Glad it's mentioned in the article.

AWS is absolutely killing it. Every release cycle, they put out quality products that directly attack Oracle. And, Oracle is in the bad position of having to eat their existing profitable business lines to compete.

Now, of course, cloud isn't here yet. There are certainly lots of issues around security, auditing, privacy, data ownership, etc. But, the trend is accelerating. Why? Because the cloud is so much cheaper. Why maintain and manage your own data center and infrastructure for millions of dollars when you can reliably outsource for a fraction?

I used to do a lot of work in the utility industry. I remember years ago giving a presentation on the cloud. The audience was passionate that it'd never, ever happen to them. They were extremely protective of their data. But, then I slowly saw it happening anyway. It crept in slowly. The first to go were the lowest risk systems. Then, the next lowest risk systems. Then, the next.

The incumbents will hold position for a long time to come, propped up mainly by government spending, hellishly long sales cycles, a corporate aversion to micro products, and FUD. They have plenty of time to catch up. But will they? That's the question. Unless oracle is actively researching new cloud platforms (and I'm guessing they are), they're going to be in pain in 5 years. Corporations that don't move to the cloud will be in a cost disadvantage.

On the other hand, serious work needs to be done to overcome the severe limitations of the current cloud. Lots of certification and auditing and risk assessments and SLAs need to happen.

[exelius]

Oracle is not quite as fucked as everyone else on the rest of the list. Oracle has the benefit of offering products that solve very real, very complex business problems: they aren't selling technology infrastructure. Oracle's core products are largely around financial and operational automation, and they are well understood by integrators across many industries. Their biggest threat is something like Salesforce -- but even then, the capital cycle for ERP systems is something stupid like 15-25 years. IBM is largely in the same boat as Oracle, albeit with slightly more government focus.

Oracle is ultimately a software company. Where their customers run that software is largely irrelevant to their core business. Is there margin to make by running an Oracle cloud? Sure, but the bulk of their profits are made just by licensing their LOB software packages and the associated annual support contracts. Oracle will be fine, and Larry Ellison will be able to buy an even bigger yacht in a decade.

[fredkbloggs]

The real reason Oracle is still making money (though never nearly as much as analysts forecast) is none of the above. It was mentioned in the article and has been discussed here many times in the past as well. For example, https://news.ycombinator.com/item?id=10207495 and https://news.ycombinator.com/item?id=9953657. Oracle's sales force is trained to use "license audits", threats of same, and various forms of bundling, such as MSAs and minimum sales commitments, to sell what Oracle wants to sell, not what the customer wants or needs. They are very effective in this, and it is the primary reason Oracle remains in business.

That said, the Oracle database product, while extremely expensive and obnoxiously licensed, is not without technical merit. There is a genuine market for it, just not a $50b one. One could say the same for what these other so-called dinosaurs are selling.

[exelius]

You're right about their sales practices -- and I've got more than one client who are slowly ripping out every Oracle product they have as a result. Oracle treats enterprise software sales like it's still the 90s, but in a lot of their markets there is no other serious vendor.

Oracle makes the majority of their money off of industry-function-specific ERP software that runs on top of their database platform. That's a market that's very hard to disrupt or displace, because you have to develop deep domain expertise in dozens of different industries and functions. Their technology platform licenses often sneak in the back door -- the customer wants the solution, and if they have to pay for the DB as part of the solution, then it's just part of the price.