It is vital to understand the concept of features versus bugs as it relates to the terms within a software contract, and not just within the software itself. A feature is beneficial; a bug is not. And so let’s examine some features and bugs in Oracle licensing contracts.
Price holds a feature? Hold on
Most consider the common practice of attaching a price hold to a license order a feature. A price hold is when Oracle will honor a highly negotiated discount on software for the length of the price hold, often two to three years. There is usually a minimum order amount of about $50,000, but otherwise the customer can execute license orders against the price hold in a no-haggle, drama-free manner. Sounds good, doesn’t it?
But here we should review support roll-ins. As you may recall, repricing is Oracle’s contractual right to take away original discounting if you seek to cut support fees by reducing software licenses. A roll-in is taking separate support streams and roping them together as if contained within a single Customer Support Identifier (a.k.a. CSI), Oracle’s unique identifier for an original license order. Mathematically speaking, the larger the support total is, the more software licenses must be cut to see licensing savings.
With that in mind, let’s revisit the price hold. Oracle’s technical support policies say that any software ordered pursuant to a price hold is effectively consolidated into a single order. In an important sense, then, price holds can and should be considered a bug, not a feature.
If the concepts of repricing and roll-ins seem confusing, then you aren’t alone. The important question in any license order is whether the terms will reduce an organization’s flexibility. The price hold, as innocuous as it may seem, reduces your flexibility to improve operations in the future by locking software into a single order. All too often, organizations forfeit contractual flexibility for an extra point or two of discount. The concessions they’ve negotiated for will actually hurt them in the long term.
Relatively speaking, price holds are flexible compared with Oracle’s unlimited license agreements (ULAs). While I’m not wholly against them, ULAs are akin to forsaking all others and marrying Oracle. I have written extensively on this topic, and suffice it to say that ULAs flip the features-versus-bugs equation on its head, stacking the deck in Oracle’s favor nine times out of 10. Don’t forget that the house always wins.
A dynamic most consider a bug -- but which is actually a feature -- is having dozens of renewals coming in throughout the year. It might seem easier for you if Oracle would consolidate all renewals into one easy payment at one point in time. But it would also be easier for Oracle, which would want to turn all your disparate support renewals into a guaranteed annuity year over year in one swipe.
Oracle: Master of long-term vendor lock-in
Anyone in business prefers high-margin, perpetual revenues. It’s not that Oracle is unique in this sense -- it’s that perpetual software is a relatively new and complex phenomenon. Where Oracle is unique is that it has mastered the long-term vendor lock-in. The combination of painfully nonfungible databases and inflexible contracts is the one-two punch that has brought Oracle to where it is today.
There are also strategies that balance up-front discounting with long-term flexibility, among other things. Keep in mind, most Oracle customers have been Oracle customers for many years, so decisions that made sense at the time may seem less wise now.
So how do you level the playing field with megavendors such as Oracle? You could use a vendor-independent facilitator with the processes, tools and experience necessary to merge best practices from software asset management, procurement and contracts, engineering and operations. The resulting plan must reduce cost without worsening service levels.
While short-term cost reduction may not be possible, your organization can start on a path toward flexible contracts, market-leading discounts, improved asset management and rationally engineered systems.
This was first published in October 2011