Hi all, Would very much appreciate any feedback here in helping me work something out in my head. You buy some media banners on another site. You want to know you're spending the money wisely so you look at the reports (assume they are 100% accurate/honest). You can trace sales because markers (cookies) are left on people who click and land on your site and also people that are exposed to the banner. If your reports say that someone clicked on one of these banners, but doesn't buy initially, 2 weeks later (cookies are usually 30 days) they come directly to your site and purchase it is attributed to the banner you should be pretty happy. There's a possibility that they would have come to your site anyway but your rationality says it very unlikely this would would be a frequent occurrence. The agency selling you the media space also points out that people are coming to my site directly having previously been exposed (not clicked) (marked by a cookie). This is also valuable although perhaps less so if substantial amounts of people come to your website anyway. Depending on how you view these sales you probably don't see them as quite so valuable. What if the media you're buying is actually on the e-commerce site where the purchasers occur? Manufacturers do this and it is increasingly common. I don't think this makes a massive difference to the value of clicked purchases. It's clear the visitor is performing an action only because of the investment you've made. I think this potentially radically changes the value of exposed-only figures. Media is clearly influential but bear with me. People will be on that site buying products anyway and the only way to test the value of exposure to media is to A/B test exposure i.e. what sales occur when exposed to the media versus not. Your thoughts? Thanks!