Subterranean proprietary software blues, Part 1

Every once in awhile I come across a story that seems too incredible to be true. This week, I heard the exact same story, from two different enterprises, about the same proprietary vendor. Unfortunately, I had to live the story, which I found unpleasant.

If you don't believe that the proprietary software model is inherently bad for customers, read on.

I've argued before that proprietary software pits vendors against would-be customers. At its most benign, the proprietary creates perverse disincentives to actually serve customers. And, as I've written, it dramatically increases the cost of failure for IT buyers and shifts nearly all risk to the buyer.

Talk about a raw deal.

This week I met with a prospect that spent millions of dollars on Proprietary Product X. For years, the proprietary vendor has happily counted its money, and each year asks for more in the form of an 18% maintenance fee. This wouldn't be so bad if the customer had actually been able to get the software to work.

Unfortunately, the software is so complex and clunky that its most useful purpose is to keep the shelves from gathering dust. The customer has asked for years (literally) for the vendor's assistance in making the product work. Surprise! Surprise! Once the vendor sold the license, its interest in the customer rapidly waned.

Think I'm giving you an edge case? An anomaly? I wish. Just ask Warner Music Group, as CIO Insight highlights:

Within of two years of deployment [of a proprietary solution, Warner] had found the software provider too rigid and unresponsive to meet the dynamic requirements of online entertainment and e-commerce. Warner software architects felt that fixing the system would require rewriting the vendor's code. Requests for changes in the licensed software would take months to be implemented...slowing Warner's ability to add and create new Web sites. [Warner] was stung that Warner was paying what [it] considered a premium price in licensing fees, and maintenance and support services, for a platform that wasn't getting the job done.

So Warner went with JBoss, and has been ecstatic ever since. Better software. Better value. More innovation. Less money.

In talking with customers, the Old World enterprise software problem is endemic to the enterprise software industry. It's a systemic problem that results from building on the wrong foundation: proprietary licenses, the crack cocaine of software finance.

Back to my story. Ironically, now that the vendor is about to lose a big new project, it's all charm and love. "We'll be right over to analyze the problem!" "Give us two months and we'll make our software look/feel/do what this upstart can do!" Where was this concern over the last three years that the customer has been trying to get the software to work?

The same thing happened with this vendor in another account that we won. In this other account, however, things got ugly. When the vendor discovered that the customer was looking elsewhere - dissatisfied, for some reason, with several million dollars in licenses and not a dime in working value - the vendor started to threaten our contact with retribution (I'm not making this up). Just the sort of vendor with which I'd want to do business.

I think this is shameful. I can't imagine treating a customer in such deprecatory fashion.

But then I can't, even if I wanted to, because I'm an open-source vendor. I'd be out of business tomorrow. If I disregard my customer, they can shut off the subscription (but keep 100% of the code). An open-source business model requires that I earn my invoice every day of the contract.

So, let's get this straight: open source costs dramatically less and yet delivers dramatically more value.

Still don't believe me? Ask the CTO of the Christian Science Monitor, Terry Barbounis:

"We're not really interested in going under the covers and changing source code, but we believe that with the community around open source we can be innovative without having to have the size of staff that a larger organization could afford."...
[What does it cost? Glad you asked.]...[T]he Monitor spent about a third the cost of a [proprietary] licensed platform in the first year of deployment, including new servers to handle the installation, and Barbounis estimates subsequent savings will be around 50 percent a year. Part of those savings, however, will be used to pay for high-level support.

Spend less. Get more. That's open source.

Are all proprietary vendors as terrible as the one to which I've been referring? Of course not. The problem, however, is that the proprietary-software model encourages the kind of sloth and disregard (though not the rancor) that I've detailed. Everyone wants to have happy customers, but a model that depends on an upfront, proprietary license fee, rather than one attuned to ongoing support/implementation concerns, is not one geared toward long-term customer happiness.

Vendors can go against the grain of their model to serve customers, but they don't have to do so and, indeed, proprietary lock-in makes it easy to discount customers' concerns because, after all, where can the customer go? This is why I urge vendors to "burn the boats" and completely leave behind this unhealthy model. Unhealthy to customers, that is.

Open source is not perfect by any stretch. And open-source vendors are even less so, as we continue to try to figure out the best ways to apply open source to business needs, as this CIO Insight article describes. But open source does align customer interests with vendor interests, and open source does give customers control of their IT.

In sum, I can't think of any way that a proprietary license is good for customers. Not one. Frankly, I don't think it's much better for vendors in focusing them on helping their customers. But that's another blog entry.