There has been quite a bit of talk in the online advertising community about attribution modeling which raises the question of how much credit the last ad should actually get when a user visits a site or performs a key purchase indicator?Â Many companies are looking at attribution modeling to determine the true effectiveness of an ad beyond just what the click-through and direct response metrics portray.Â The primary contention is that while a user may click on an ad to navigate to a site, previous ad views contributed to this action but are left unrecognized and uncounted.Â While clickthrough is an easily measurable metric, advanced marketers understand that it is not the most accurate way to measure a campaign’s effectiveness.
In an effort to measure the effect of display advertising on site visitation, we looked at marketing activity related to two of the largest retailers, Target and Walmart.Â We identified consumers exposed to either of these retailer’s online display campaigns using ad tags on Doubleclick.Â We then looked at site visitation with respect to display ad exposure.
In both cases, approximately 85% of visitors to the advertiser site were exposed to the retailer’s online display ad at least once, either before or after visiting the retailer site.Â We based these counts off of the user’s first visit to the retailer site in the month.Â Of those who were exposed, more than half were exposed before visiting the retailer site which means that about half of the visitors to either of the retailer sites saw an online display ad for the retailer before visiting the site.Â However, it is very likely that not all consumers who were exposed to a display ad were counted by the clickthrough rate metric for campaign success because not all consumers who see an ad respond immediately by clicking on the ad.Â The main question becomes how can campaign metrics close the gap between measuring direct clickthrough rates ( often around 0.1% ) and the broader impact of the display ads (this case suggests they could be as much as 50% of site traffic) driven by latent activity?
This is a question that is being addressed as many online metric measurement companies scramble to develop a "perfect solution."Â Such a general solution may not exist as each campaign is unique and the ideal campaign effectiveness measurement system may vary by campaign.Â Nevertheless, marketers need to start exploring alternative methods to direct campaign measurement in order to properly determine the true impact of their online advertising spend.Â Stay tune for a similar look at display attribution for search!