Why Web Ads Don't Work

A long time ago, pundits started declaring the death of the banner ad. By and large they were right--banner ads now get clickthroughs that peak at .1%, the cost per thousand is close to zero and Google and others have demonstrated that contextual text ads are far more effective.

Yet, these banner ads (in a variety of shapes and sizes) persist. What stuns me is not that they continue to hang around, but that they're used so poorly, and (I'm hypothesizing here) so little testing is done.

Here's an ad for the revived go.com:

Click on it (yes, some people do) and it takes you to:

I'm just amazed that this is the best they could come up with. Where's the call to action? Why is this a better service? How does this answer the promise of the banner I clicked on? Have they tested different approaches?

The key to the process is simple: banners make a promise. Offer pages continue the promise and provide evidence that the solution is proven, legitimate and effective. If you test these steps, your yield has to go up. If it goes up, your costs go down.