We all like to think that the crazy Wild West days of interactive experimentation are over, but the exact opposite is true: There are more harebrained, half-baked content schemes on the web than ever, and at least 90% of them will fail.

This is exactly as it should be: Interactive media is inherently Darwinian; barriers to entry are low or nonexistent and nobody's figured out what really works. So CMOs are faced with a double-edged sword: If they refuse to experiment and just hunker down with time-proven media, their interactive efforts will simply muddle along. But if they place too many bad bets, they'll be on a fast track to a pink slip.

Sorting out the I-rational from the I-ridiculous is a full-time job. But here are 10 content crazes that have enough lunacy in them to warrant serious CMO scrutiny: It's no secret that marketers are having many second thoughts about Second Life, which just a few months ago was the preferred virtual vehicle for many major brands to show their wares. Today, walking (or flying) through these branded areas is more chilling and depressing than walking through an abandoned amusement park. Do you really think IBM's brand is being helped by hosting a 3-D area that has tumbleweeds rolling through it? Do marketers seriously believe consumers are going to be gullible enough to believe product recommendations from people who are being paid to recommend the products? Of course not. And yet a sneaky, stealthy form of word-of-mouth advertising called "pay per post" is based on exactly this proposition. CMOs who dabble in pay per post or its deranged cousin, "astroturfing" (phony grass-roots marketing using blogs) are playing with fire and shouldn't be surprised when the brands they're supposed to be shepherding get toasted once the ruse is inevitably discovered. Ad units capable of displaying customized creative keyed to historical search behavior are a great idea in theory, but there are enough serious problems with this kind of advertising to give any CMO pause. Many users share computers, which means that Spouse A is going to be targeted with ads based on Spouse B's search behavior. What could be more irrelevant? Many CMOs have no problem authorizing multimillion-dollar TV and print campaigns, but their systematic neglect of search marketing borders on the criminally myopic. Search-engine marketing doesn't drive demand; it responds to it, which means that unless your brand is present to capture post-ad queries, you've failed to close the marketing loop. Plus, it's quite likely your competitors are already exploiting that failure with strategies to poach the awareness you've spent so much to generate. One of the main reasons why repurposing standard 30-second TV spots into video pre-roll ads is so ridiculous is that most users turn down or mute the sound to save their fellow workers exposure to the audio. True, most users will endure a (silent) pre-roll to catch a news clip, but unless your ad can stand on its visual elements alone, you're wasting your money. The reason search engines are much better places to find information than directories is because they leverage automation to do the grunt work that human editors used to do. Directories (especially vertical ones) are still useful, but the whole concept of "human-powered" search engines is so ill-conceived that it's a miracle venture capitalists, advertisers, the tech press and even some CMOs take it seriously. OK, Google runs a terrific search engine that's become an advertising powerhouse. But just because Google delivers results for search and contextually based media doesn't mean that it has an advantage in print and broadcast media, which are inherently less trackable. So far, Google's forays into non-search media have been failures, but it shows no sign of giving up. Some call this persistence, but the image that comes to my mind is Don Quixote attacking that famous windmill. Right now, only 30% of users regularly delete their cookies, but that percentage could soar if a widely publicized goof up (such as last summer's leak of AOL search data, which was detailed enough to identify individual users) starts the regulatory wheels turning in Washington. CMOs need to be cognizant of the dangers of behavioral targeting before committing their brands to what some will certainly view as excessive cyber-sneakiness. What could be more annoying and less useful than a site where thousands of people are given 140 characters to shout out about what they're doing at every moment of the day? The amazing thing is that enough people out there think this mindless stream of ephemera ("I'm eating a tangerine," "I'm waiting for a plane," "I want a Big Mac") is interesting enough to serve as the basis for a viable advertising platform. The mobile-marketing environment is qualitatively unlike interactive media, in which the user's basic situation (watching TV, surfing the web) can be presumed with some accuracy. Users of mobile devices are doing all kinds of things with their devices, some fanciful and some drop-dead serious. For that reason, they will never take kindly to advertising that interposes itself between them and a critical task.