This software brought to you by Ovaltine

Is an advertising-supported model the way forward for the software industry?

Are we having a recession yet? Ask and ye shall receive. If the bad news keeps rolling in from Wall Street, coded phrases like "economic downturn" won't stop companies from acting as if the recession is already here.

No wonder pundits are predicting boom times for SaaS, cloud computing, and other, subscription-oriented software offerings. Who wants to sink cash into software as a fixed asset when returns from short- and long-term investments are drifting into the red?

[ Which areas of IT spending will get funded when money is tight? InfoWorld rolls together projections from Forrester, Gartner, and IDC with answers from real-world CIOs in "Five top spending priorities for hard times" ]

To suggest that the death of traditional desktop software is imminent, however, would be premature. Too many of the current online offerings are too immature, too insecure, or too unreliable for mission-critical use. Still, that doesn't do much to dispel the grim but very real spectre now haunting the software industry: How in the hell are we going to make money?

Some claim the answer is that last refuge of scoundrels: advertising. A growing roster of top-ranked IT companies -- from Google to Microsoft, Mozilla to Symantec -- have been boosting the bottom lines of their products with cross-branding and marketing deals. But will it really work, or is it just another house of cards?

Sun's gambit
I've spent a lot of breath knocking Sun and its business model lately, but I have to give Jonathan Schwartz and company some credit. Viewed in hindsight, Sun's move toward a 100 percent open source, subscription-support model was a preemptive strike against just this kind of market eventuality.

If you don't want to pay for Sun's software, don't. Go ahead and use it anyway. But if you need world-class support to maximize the efficiency of your business infrastructure, Sun's operators are standing by.

But while Sun's open source business model -- like Red Hat's -- works well when the products are complex, mission-critical enterprise information systems, it's not a good match for consumer desktop software. In the consumer realm, user-friendliness is king. And when the app is user-friendly enough -- in other words, when the developers have done their jobs -- the customer doesn't need support. (Little wonder that Red Hat got out of the consumer desktop Linux business.)

That's OK, Sun's CEO says, because users who don't pay can be profitable, too. As Schwartz blogged earlier this month, Sun sees each free download as a channel opened to a user; and that channel, he says, can be monetized through co-branding and marketing deals with other companies. If he can be believed, the marketing channel to Java's audience is already the major revenue source for Sun's Java efforts, and he's currently auctioning off co-branding opportunities in the OpenOffice.org suite, too.

Sun isn't alone. The Mozilla Foundation, for example, derives some 88 percent of its revenue from the marketing deal that gives Google priority placement in the Firefox browser's search widget. Even Apple offers its Safari browser to QuickTime users as a kind of self-promotional ad campaign. As a trend, it can be compared to the early days of television: Your software will return after this brief word from our sponsors.

Mad Men or maddened customers?
While television viewers were accustomed to sponsorship from the even-earlier days of radio, however, the concept of software as a marketing channel is still brand new. Unfortunately, it doesn't necessarily translate the way radio ads translated to television.

Take Sun's approach with the JRE, for example. Just last night I was offered an upgrade to Java 6 Update 10. Before I could download the update, however, I had to uncheck a box that would have installed the Yahoo Toolbar into my browser, by default. This is wrong for two main reasons.

First, the JRE was software that I already had installed. I wasn't shopping for new software; in fact, I hadn't even been checking for new Java updates. The update had been marketed to me, via the system's auto-update feature. All I wanted was to have the most recent bugs and security flaws in the Java runtime fixed. Why, then, should I be forced to opt out of installation of a brand-new browser plug-in that I had never used before? Does the reliability and security of its software mean so little to Sun that it will put a marketing opportunity first?

Apple made a similar mistake when it made downloading and installing Safari the default for QuickTime updates. It's the equivalent of including a free pack of cigarettes with every oil change -- which, by the way, would be not just unwelcome but illegal. But I digress -- or do I?

Second, while it makes a certain amount of sense to use software installers to market other software products, it only makes sense when you're marketing to the people who are authorized to install software. In enterprise environments, that authority rests with the IT department, not end-users. You can switch off Java's auto-update feature, but it's on by default. Therefore, by default the Yahoo toolbar will be marketed directly to every single employee in an organization that uses Java.

Why is this a problem? Software licensing, for starters. As an organization, my company may have agreed to Sun's terms for Java, but we may never have even considered the Yahoo plug-in. Suppose it sends back user tracking information that would be forbidden under our IT policies? With a click of the "Agree" button, our users could potentially compromise site security. And who's to say that the software being offered isn't out-and-out adware/malware?

Sure, I can configure my firewall or local system settings to prohibit such installations, but it doesn't exactly engender goodwill for the vendor that's pushing the software. "Carpet bomb" marketing is fine for offshore pharmacies, but the Fortune 500 should know better.

Channel wars
Perhaps the most disquieting aspect of the software-as-channel model, however, is that it tends to consolidate control of software distribution into a few very large hands. To advertise effectively you need an audience. That's fine if you're an Adobe or a Microsoft, but what if you're a small software vendor trying to broaden your market? How do you gain access to an audience without coming under the thumb of the big vendors?

A company called OpenCandy offers an interesting solution. OpenCandy is building a network that allows software vendors to include advertising in their installers, just like the big boys do. Each vendor vouches for other applications that it thinks would be good fits for users of its software. OpenCandy's recommendation engine then chooses the best app to offer to each user who runs the installer. The result is a community-based version of what companies like Sun and Apple are doing for their own products.

It's too early to know whether OpenCandy (or a similar offering) will catch on. But in a world where the shrinkwrapped software market is rapidly becoming a thing of the past, new marketing and distribution mechanisms are needed badly -- not to mention new business models.

Still, I can't help but feel cynical toward the idea that advertising revenue can drive the software market over the long run. Maybe it's just my background. Ask a newspaper reporter how well advertising is working to support that medium. Will software be different?