The decline and fall of software empires

A few months back, Marc Andreessen posited that Oracle was in really big trouble, it just didn't know it yet (at least not publicly). His experiential reasoning: he invests in lots of companies, and is connected to many many more, and not one of them, without exception, uses Oracle. Everyone uses either MySQL/PostgreSQL (i.e. open source) SQL, or NoSQL Mongo/Couch/etc.

Yesterday, Oracle's earnings came out, revenue was basically flat, and the share plunged. As pointed out by Matt Rosoff, this may confirm Andreessen's prediction.

My take on it was the income from support and maintenance versus new license sales. When a software firm reaches the point that the revenue from maintenance equals or exceeds revenue from new sales, essentially the company is on the way down. I was a customer of IBM when VP at a large financial back in 2001-2005, and I clearly recall when IBM's Tivoli division crossed the 50% mark.

A software - and any - company's future depends on its ability to sell to new customers. If existing customers are paying the same or more as new ones, then you are having a hard time selling to new customers, and you are on your way down. You may milk the cow for a good few more years, in which, because of reduced R&D investment, your cash may be higher than before, but life is on the way down.

The real pity of it is that Sun was a great company (and great engineering shop with great talent), and likely will go down along with Oracle.