Does context matter?
If you’re running a pay per click ad designed to support a cost-per-acquisition strategy, (Google AdWords, et.al.) then does it matter where your ad runs?
Remember, the point of the ad is to get someone to click (that’s what you’re charged for… the click) and then the goal of the site is to convert that click into permission and eventually a customer.
So, does it matter where the ad runs if it works?
Media buyers sure think so. Jason Klein at Special Ops Media says, "With Quigo, you know it’s on ESPN.com, not Joe Schmo’s sports blog."
I can understand why a media buyer would say this. I can understand why Jason Clement at Carat said, "We had essentially pulled all of those big advertisers off of the ad networks [Google, Yahoo] by the end of the year." After all, the media buyers need to demonstrate that they are using their hard-earned intuition to actually earn their commissions.
But if I were one of those ‘big advertisers,’ I’d think really hard about whether Jason is doing me a service. The hard work of running contextual ads is testing. Run an ad, test the landing page, see what works. If it works, do it more. If it doesn’t work, do it less.
Sure, you need to start with intuition. But my intuition tells me that Joe Schmo’s sports blog might actually perform better than a high-profile site. My intuition tells me that a click process that begins on a digital photo review site is more likely to lead to a purchase than one that begins on a fine art website.
In order to make this work, the big ad networks need to tell you where the traffic is coming from and they need to make it easy for you to choose where to run the ads next time so you can repeat and scale the process.
The funny thing is that this context argument was perfected by the big networks when cable TV came along. They used it to justify selling unmeasurable expensive ads on mass market network shows against the competition: unmeasurable ads on more focused cable shows. Then it happened again with banner ads–the big name sites always could charge more than the smaller ones.
This time, though, we’ve got numbers. Let’s use em.