Overview
Ghost ads are not ads per se. What we do is take a 'sample' group to see how the campaign would work if we did not serve them any ads. We use Ghost Ads to measure Incremental Metrics in App Marketing campaigns and Drive-to-store campaigns in our platform.
Ghost Ads are bid requests that we have stored as “ghost impressions” or “placebo impressions” without us serving an actual ad (placing a bid). In other words, with Ghost ads, we do NOT serve ads to users that we determine should be placebo users, but we do not serve them any other ad instead.
How does this work
Campaigns running in mediasmart can use our algorithm to select placebo users locked from other campaigns running on the mediasmart platform.
When a user is identified as Placebo/Ghost, instead of serving them a placebo ad, we will "serve" them a Ghost ad, which means:
We do not place a bid.
Store the bid offer as a “placebo impression” with all the variables associated with the bid offer.
Generate a “placebo click” if it applies, based on the expected CTR calculated for that particular combination of variables in the bid offer.
We have a mathematical formula with which we measure how the campaign is running against that sample group. It's the same idea, that medical experiments use to test medicines, between control and placebo groups.
Advantages
Advertisers can gather more information at no price: no need to buy media to gather “placebo” results.
Placebo users will NOT be redirected anywhere, independently of whether the partner used by the advertiser allows for customization of the destination URL or not.
How to report Ghost ads
Impressions, Clicks, and Conversions will contain both regular and placebo information if not filtered, but by default, we will add the filter to ONLY REGULAR, which can be removed at any time by the user.