There are really four distinct ad formats on YouTube, the online-video site with the lion’s share: display, text ads, InVideo ads, and video ads (10-30 seconds). Unlike most sites, YouTube forbids preroll, but does offer full-motion video ads if the user fires the player (in the home-page ads or in display areas).

So let’s talk about your advertiser’s goal for each, and then let’s give a good case study.

1. Display: Your display ad (a flat graphic) should primarily brand since 95 percent of people will not experience anymore than that. However if you want click-thrus of greater than a fraction of a percent, then have a good call-to-action. Nothing sells like more video content — especially if it’s non promotional.
2. Text: Text ads are dirt cheap, but largely ignored. However you can brand on these and run them CPM (cost per thousand), but the ad will eventually fall aside to CPC (cost-per-click) ads that are more profitable and maybe more relevant. So the smart thing to do is be extremely targeted and buy specific keywords, then have a specific call-to-action.
3. InVideo: InVideo ads remain novel, and give the advertiser the “hard to ignore” exposure of a banner that bursts into the video’s bottom 20%. Usually these are sold with accompanying display ads adjacent to the video. Here’s a good chance to brand (again- even a high click through will be in the low single digits). So focus on branding, but invite people to see more content.
4. Full-motion video ad: Here’s an opportunity to entertain, brand, and invite people (if necessary) to the next step of the sales cycle (for more content and commerce).

The bottom-line is that advertising against online-video means you have to engage the viewer in an entertaining format. Don’t expect many people to leave my CharlieCam video to buy a helmet cam (even though I am rather impressed with instant targeting last night by a commerce site). Instead, promote more content. ZipIt Wireless (which sells a device and service for instant messaging) used YouTube’s Fred in ads, and invited people back to a campaign site (FredOnZipIt) to see more videos. My guess is that about 2-4 percent of viewers (in the early part of the campaign) visited the campaign site, and of those maybe 5 percent purchased ZipIt devices (wild guess).

Did that math work for ZipIt? I notice ZipIt’s ads aren’t popping up on Fred’s videos or elsewhere, so maybe not… or perhaps the company is assessing results. Either way, this campaign was in the right direction — use content to lure the potential customer, and then expose them further to the product or service (in a web experience that can do that in ways television can’t).

Now look at the Burger King campaign for an even better case study. It’s an integrated campaign between a comedy show and a fast-food restaurant which might otherwise have a tough time getting people to engage in a branded experience.

Burger King is sponsoring Seth Macfarlane’s “Cavalcade of Cartoon Comedy.” Although the creator of Family Guy produces edgy stuff far more risque than much of the YouTube amateur Partner content, it’s instantly recognized.

Macfarlane also probably made a good choice to align with Burger King, because he sure wasn’t going to cover production costs of his short web series via ad-sharing through online video. And he’s also enjoying the publicity that Burger King is making for him.

It’s another page out of the Star Wars Happy Meal promotion playbook, but a good reminder that online-video is growing up.