I remember sitting in this marketing presentation by an auto company touting this amazing microsite they did. They had full video of the product, a message board, a contest, and of course their "viral" component -- "Email a friend." They walked us through the entire site, its promotion, the various funnels through SEM and email. When all was said and done, it cost them $1.2M. And then, the other shoe dropped. When someone asked how many cars it sold, the response was enthusiastic and excited. "We got more than 6,000 email addresses."

Bear with me here: 1,200,000 / 6,000 = $200 an email address. An email address does not translate into a car sale. You know how many cars they did sell? 14, for about $400,000 total, to people that were probably already predisposed to buy the car anyway from TV.

If you realize that the profit margin is probably only 10 percent on those vehicles, they spent $1.2M all outbound, for about $40,000 in profit. Of course, the poor man at this point was being ripped to shreds

"It's not about that! It's mainly branding!" OK, we'll go with you there.

"How many people -- uniques -- went to the site?" 57,000. So, they spent $21 per person just to take a look. When we dug deeper, only 9,000 spent more than three minutes on the site -- 9,000 people. Basically, you can make numbers look like whatever you want them to look like, but the truth of the matter is that immersion sites are often way more costly than the results that can be obtained by integrating that content into your main site.

The immersion everyone talks about online is extremely hard to obtain in a lean-forward, active medium like the internet, and it's much more prevalent in TV where the lean-back environment makes you receptive to it. You do not have to create a separate website. Eye focus, even on an ad, or a smaller page, tunes out the periphery.