Solving the Issue of Reach for Retargeting

The IAB recently reported that spending on online display advertising in the UK is up 6.4% for the first half of 2010. Similarly, Kantar Media reported a 5.3% increase in display for the U.S. for H1. As more advertisers are shifting their ad spend towards display, the online display ecosystem continues to blossom, as Terence Kawaja’s familiar landscape map shows:

LUMA-Display-Ad-Tech-Landscape-for-AdExchanger

Because the Internet is growing more fragmented every day, advertisers will increasingly depend on media placement technologies to decide when and where their ads should be placed in order to deliver the best return. In an attempt to help marketers better understand some the wide range of media placement strategies, comScore and Valueclick recently published a study exploring the pros and cons of six different types of media placement strategies for online display. These strategies include the following:

  • Audience Targeting (aka “behavioral targeting”): serving ads to users based on their web behaviors
  • Contextual Targeting: serving ads based on the environment in which the ad is served (publisher context)
  • Efficiency Pricing: cost-per-click (CPC) ad pricing (as opposed to impression-based pricing, CPM)
  • Premium Pricing: high-visibility ad placement on premium publisher sites
  • Retargeting (aka “remarketing”): serving an advertiser’s ads to consumers that have previously visited their websites
  • Run-of-Network (RON): ads appear anywhere in the ad network

The primary question the study attempts to answer is how these various strategies could work, possibly together, to generate the strongest lift in website visitation and trademark search behavior for the advertiser’s brand. (Another comScore study explains why they chose these metrics as opposed to just the industry standard metric, click-through rate.) The results of the study showed that retargeting drove the highest lift for both of these metrics on both a short term and a long term basis:

‘Retargeting’ far surpassed every other strategy for delivering impacts, and it provides a moderate level of reach [compared to the other strategies], indicating this is a must-do for every marketer. (p. 13)

It’s for this reason that our targeting technology starts with retargeting as one of its foundations. It is a verified strategy to achieve the highest reach with the lowest amount of scattering losses when it comes to online display. (“Scattering losses” refers to the budget wasted for ads served to users who would not be potentials for the brand to begin with, which is more likely to happen with a loose placement strategy like RON.)

The downside to retargeting when compared to some of the other strategies, such as RON or contextual targeting, is that the reach of retargeting is limited to previous website visitors. This means that retargeting will probably have a smaller reach than these other two strategies – less people overall will be shown ads via this retargeting, including some people who might have potential for interest in the brand being advertised. On the one hand, the advertiser could use a combination of retargeting and one of these other methods to make sure that they are driving results without killing their reach, but there may be an even better way.

We’ve decided to address this issue by extending the reach of retargeting via “social targeting.” Put simply, social targeting takes the pool of retargeted users (i.e., the people who have previously visited the advertiser’s site), and we analyze that pool so we can grow the reach of the campaign to include other users with a high likelihood to have an interest in the advertiser. And how to we determine which users demonstrate that “high likelihood?” We combine several layers of data, but the two most important are user interest data (as is used in basic audience/behavioral targeting) and network pattern analysis data, which helps us determine pseudo-social networks online to see which users might be connected.

By combining these data sets, we can understand which users will be most likely to have an interest in the advertiser, enabling us to increase the advertiser’s reach while helping them to avoid the scattering losses associated with placement strategies like RON.

As advertisers increase their investment in display, this is just one way that they can make sure their display dollars are working as efficiently as possible to drive interested users to their website to convert.

Back to news overview