In 2006, AOL opened a website, research.aol.com, offering up 20 million search strings from over 600,000 AOL customers, for academic use by search engine researchers. Even though the data included no personally identifiable information, it was still a huge privacy violation. They quickly removed the data and apologized, and subsequently the chief technology officer resigned. The site no longer exists.

A few months ago, a company named Qtrax spent almost $1,000,000 to announce the world’s first free and legal peer-to-peer music download network. Twenty-five million songs and the complete backing of the record industry. Only they didn’t have that backing. They blew their launch. As I write this, the site says "we'll be enabling free and legal downloads soon." I'm not holding my breath.

Every day, new services launch on the Web, often earlier than they should with the thought that "just getting it out there" is a good idea. And many of these are in perpetual beta, like Google's Gmail.

Once upon a time the term "beta" meant that the software was very close to release, but the developers were still seeking feedback. Betas were typically only distributed to a limited audience of a few customers who understood the risks inherent in using software that might break, and who would report bugs. Those bugs would then be fixed and the final version released for sale.

Today, betas are instantly available to everyone worldwide; there are rarely limits. They are used to attract free publicity, or to drive user demand through "invite only"-style launches. And a beta is not a one-time event. The software changes frequently; new features are added constantly. If problems occur the answer is often not to fix them, but to state that "the product is only in beta". But people use the product as if it were final. At best, the service might not work exactly as advertised. At worst, your data could be lost, destroyed, or leaked.

This is understandable to a certain extent. Making software or services available on the Internet provides a kind of load testing that no quality assurance model ever could. And tens or hundreds of thousands of users will find bugs and suggest enhancements faster than a few hundred controlled users -- if you intend to fix the issues. Twitter has become a very popular service, but it suffers constant breakdowns.

Such a short release schedule also forces competitors to release their "me too" products much sooner than they should have, which leads to a downward spiral in overall quality of the products -- a race to the bottom, so to speak. And it forces users to quickly abandon what might otherwise be promising solutions as they try the next new thing rather than wait for the current service to get better.

And if they can't attract and keep customers while finding a way to generate revenue, software companies have a nasty habit of disappearing. It’s much better to put a little effort into getting it right the first time. Especially when you might only get one chance.

Larry Borsato has been a software developer, marketer, consultant, public speaker, and entrepreneur, among other things. For more of his unpredictable, yet often entertaining thoughts you can read his blog at larryborsato.com.

More news, commentary, and predictions from The Industry Standard:

• Opinion: 10 'Net services that will succeed (and 10 that will probably fail)
• Melissa Chang: Five reasons why a recession is a good time to start a company
• Melissa Chang: 10 reasons that start-ups 100% absolutely should outsource (almost) everything
• Melissa Chang: Facebook vs. MySpace: The battle for global social network dominance
• Ian Lamont: Is Twitter worth $75 million? $150 million? How about none of the above?
• Ian Lamont: The real problem with Google's Blogger service: Neglect
• Ian Lamont: Why podcasting is failing